Content uploaded by Thomas Litschel
Author content
All content in this area was uploaded by Thomas Litschel on May 14, 2021
Content may be subject to copyright.
Annual Review of Biophysics
Protein Reconstitution Inside
Giant Unilamellar Vesicles
Thomas Litschel and Petra Schwille
Department of Cellular and Molecular Biophysics, Max Planck Institute of Biochemistry,
Martinsried 82152, Germany; email: litschel@biochem.mpg.de, schwille@biochem.mpg.de
Annu. Rev. Biophys. 2021. 50:525–48
First published as a Review in Advance on
March 5, 2021
The Annual Review of Biophysics is online at
biophys.annualreviews.org
https://doi.org/10.1146/annurev-biophys-100620-
114132
Copyright © 2021 by Annual Reviews.
All rights reserved
Keywords
model membranes, synthetic biology, bottom-up biology, in vitro
reconstitution, liposomes, phospholipids
Abstract
Giant unilamellar vesicles (GUVs) have gained great popularity as mimi-
cries for cellular membranes. As their sizes are comfortably above the op-
tical resolution limit, and their lipid composition is easily controlled, they
are ideal for quantitative light microscopic investigation of dynamic pro-
cesses in and on membranes. However, reconstitution of functional proteins
into the lumen or the GUV membrane itself has proven technically chal-
lenging. In recent years, a selection of techniques has been introduced that
tremendously improve GUV-assay development and enable the precise in-
vestigation of protein–membrane interactions under well-controlled condi-
tions. Moreover, due to these methodological advances, GUVs are consid-
ered important candidates as protocells in bottom-up synthetic biology. In
this review, we discuss the state of the art of the most important vesicle pro-
duction and protein encapsulation methods and highlight some key protein
systems whose functional reconstitution has advanced the eld.
525
Contents
1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
2. METHODS OF PROTEIN ENCAPSULATION IN VESICLES . . . . . . . . . . . . . . . 529
2.1. Lipid Film Hydration Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529
2.2. Inverted Emulsion Transfer Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
2.3. Microuidic Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
3. SUCCESSFULLY RECONSTITUTED PROTEINS AND PROTEIN
MACHINERIES .............................................................. 532
3.1. CytoskeletalProteins...................................................... 532
3.2. MembraneProteins....................................................... 535
3.3. Enzymes.................................................................. 539
4. CONCLUSIONSANDOUTLOOK .......................................... 541
1. INTRODUCTION
While biology has been an analytical science for most of its history, amazing progress in the life
sciences within the past decades has triggered ambitions to use the knowledge and methodol-
ogy acquired from dealing with living systems to actually build biological functionality from its
elemental building blocks. Among these building blocks into which biological systems may be
dissected, proteins are the most obvious. As the workhorses of the intricate chemical wiring of a
cell, they are tasked with regulating or carrying out every major cellular function through specic
interactions and enzymatic activities. In vitro protein reconstitution is a key practical challenge
for protein biochemistry, which has long been concerned with the production, purication, and
investigation of functional proteins under controlled settings.
Experiments with puried proteins have increased our understanding of life on the molecular
scale, making it possible to characterize the biochemical and physical properties of isolated
macromolecules. Through the combination of a limited number of components in simplied
experimental systems, it has been possible to gain detailed insight into the biological processes
and machines central to life. However, it has also become evident that purication and production
of particular proteins in aqueous buffer solution fall short of recapitulating the function of many
proteins, as proteins require specic geometries or biochemical environments to unfold their
physiological functionality. Therefore, molecules and structures that mimic cellular environments
are required for the functional reconstitution of puried proteins or protein systems. As these
reconstituted biological systems grew ever more complex over recent years, the term bottom-up
(synthetic) biology was coined, in contrast to the top-down approach applied to living cells, which
is analytical in nature (82). This rapidly growing and multidisciplinary eld combines molecular
and cell biology, biophysics, and biochemistry, as well as engineering disciplines such as micro-
fabrication and microuidics. The term bottom-up suggests that the eld aims to ascend from
low to high complexity, toward the assembly of a—however rudimentary—living organism. Some
researchers have taken on this challenge explicitly, with the daring goal of creating a functioning
minimal cell from biological building blocks (119, 133). While still not entirely within reach, this
goal will deepen our fundamental understanding of living organisms and the basic requirements
of life while also having exciting potential applications for a new generation of bioeconomy. Being
able to harness functional biological systems and subsystems with a modular synthetic approach
526 Litschel •Schwille
may yield benets comparable to the ones that were provided by synthetic chemistry a century
ago.
One key component at the center of many reconstitution experiments, and potentially essential
in the pursuit of creating articial cells, is the cell membrane. All living organisms known to date
have cell membranes made of lipid molecules. In addition to their roles as boundaries for cells
and cell organelles, membranes are also home to approximately one-third of cellular proteins (6)
and, thus, provide perhaps the most important reaction space in the cell. Crucial processes such as
inter- and intracellular signaling, photosynthesis, respiration, adhesion, motility, and division are
all centered around membranes.
The most obvious and fundamental feature of the cell membrane is its compartment-forming
role. In single-celled organisms—which are particularly important in the context of the origins
of life—the function of membranes as conning boundaries enables the existence of individual
entities whose personal genetic makeup is distinct from that of others and thereby forms the basis
for Darwinian evolution (27, 145). As such, a selective barrier for the exchange of material and
energy is one of the prerequisites for the emergence and existence of life as we know it (147). Fur-
thermore, in eukaryotes, intracellular membranes are essential to allowing local up-concentration
of proteins, metabolites and ions, and separate incompatible reactions in space to create optimized
microenvironments and establish concentration gradients within the cell. In addition to their ac-
tion as selective boundaries, the facile deformability of cell membranes is another critical aspect,
as membranes must support frequent division and subsequent growth, key prerequisites for the
proliferation of organisms. Moreover, in most organisms, membranes are frequently required to
adapt their shape to environmental cues and other structural components, including the internal
cytoskeleton or an external cell wall. Not surprisingly, the lipid composition of higher organisms
is very complex, often consisting of a large variety of amphiphilic phospholipids that readily form
bilayers. From a physical perspective, membranes can be considered two-dimensional uids for
lipids and immersed molecules: easy to bend, yet hard to stretch. Lastly, the spatial connement
in cell-sized compartments is important for the kinetics of biochemical reactions (70, 159, 165)
and the self-organization of structural elements and has thus been the subject of many biophysical
studies. Processes like cytoskeletal assembly and biological pattern formation, which have been
extensively studied in experiments and theory, heavily depend on 2D or 3D geometry.
Many properties and functions of the cell membrane and the associated proteins can be studied
independently, and in vitro reconstitution often calls for such reductionist approaches. Supported
lipid bilayers or liposome assays of dened composition are frequently used to study protein–
membrane interactions. In addition, water-in-oil emulsion microdroplets, which can be produced
in large numbers by microuidics, serve as excellent tools to investigate the effect of limited vol-
umes on biochemical reactions (159, 164). They are often useful in the context of chemical evolu-
tion (63, 149). Furthermore,microfabricated structures are used to observe the effects of geomet-
rical connement on protein assembly of cytoskeletal structures or biological pattern formation
(21, 31, 40, 91, 108, 138, 165), and free-standing bilayer structures, such as suspended membranes,
have been used to investigate membrane deformations (59).
These widely used experimental model systems are each perfectly adapted to the study of
one particular property of lipid membranes or membrane-associated proteins and have furthered
our understanding of membrane protein function signicantly in the past decades. However,
considering each property in isolation does not yield a comprehensive picture of the complex
and intertwined physiological roles of membranes in a living cell. A more generic model system,
which reects the fact that membranes serve as both containers and transformable spaces that
respond to the reactions within, is based on the 3D encapsulation of functional biomolecules
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 527
within unilamellar vesicles, made of phospholipids that resemble the composition of cellular
membranes. Until quite recently, systematic technical limitations hampered efcient membrane
protein encapsulation in lipid vesicles. Specialized protocols and equipment were required to
reconstitute proteins in these biomimetic compartments, and few labs with specic expertise
have succeeded in their study. In the past few years, methodological developments and technical
advancements in the eld of bottom-up biology have made it easier to generate vesicles as cell-like
compartments and opened up the eld to a much broader community of researchers.
In this review, we discuss various ways in which encapsulation of functional proteins into syn-
thetic lipid vesicles can be accomplished and demonstrate that this approach can be immensely
useful to reach a new level of complexity in reconstitution experiments, potentially paving the
way toward the creation of articial cells. We introduce some of the most frequently used meth-
ods supporting the transfer of complex solutions into intact membranous vesicles and highlight
relevant work that has led to new insights using these approaches to reconstitute a diverse array
of proteins inside vesicles. Figure 1 gives an overview of protein systems that are the subjects of
these studies.
Figure 1
Hypothetical giant unilamellar vesicle containing protein systems that are of special interest to reconstitution studies.
528 Litschel •Schwille
2. METHODS OF PROTEIN ENCAPSULATION IN VESICLES
Giant unilamellar vesicles (GUVs) are made from amphiphilic molecules that form a single se-
lective layer as an efcient boundary between an aqueous interior and an aqueous exterior. The
most common type of vesicles are phospholipid vesicles, i.e., vesicles that consist of a bilayer of
phospholipids and, therefore, best represent the cell membrane in structure and function. Phos-
pholipids, such as phosphatidylcholines and phosphatidylethanolamines, are organic molecules
that consist of a hydrophilic head group attached to two hydrophobic carbon chains. Above a
certain concentration threshold (the so-called critical bilayer concentration) (99), phospholipids
self-assemble into bilayers that spontaneously form spherical vesicles.
Such vesicles can be generated in a large range of sizes. Small unilamellar vesicles (SUVs) and
large unilamellar vesicles (LUVs), which are both often referred to simply as liposomes, have long
found applications as encapsulating compartments in areas such as pharmaceutics as drug delivery
systems (5) or in cosmetics and other industries. These smaller vesicles are also useful in many ways
in reconstitution experiments; for example, they can be used as intermediates when handling non-
soluble membrane proteins or to investigate protein–membrane interactions with more analytical
methods. GUVs are between 1 µm and 200 µm in size and thus roughly in the same size range
as biological cells. Due to their size, GUVs are compatible with light microscopy assays, making
them the preferred systems for biological reconstitution experiments using uorescence imaging
methods. The term unilamellar indicates that the membrane consists of a single bilayer. While
the differentiation between unilamellar vesicles and the usually undesired multilamellar vesicles
(MLVs) used to be critical with traditional methods for vesicle generation, modern encapsulation
techniques generate unilamellar vesicles by default. A more desirable variant are multicompart-
ment vesicles, either as vesicles that contain nested vesicles (vesosomes) (29) or with equally sized
compartments that share a single outer membrane leaet (28, 35).
The techniques used to generate lipid vesicles have evolved over the past decades, and recent
advances have greatly improved reproducibility and potential applications. In this review, we give
a brief overview of the most common techniques and discuss their suitability for encapsulating
protein solutions within GUVs.
2.1. Lipid Film Hydration Methods
The formation of GUVs by hydration (also called swelling) is one of the earliest techniques for
generating such vesicles (124). While some of the more recently developed methods discussed
below facilitate the targeted encapsulation into GUVs, hydration methods are still the most fre-
quently used techniques for the general preparation of GUVs as model membranes.
For the hydration method, phospholipids dissolved in chloroform are deposited as lipid lms
on a substrate such as glass. The lipid solution is then dried down to form lipid lms. Subsequently,
an aqueous solution is added, and the so-called swelling process starts. During this process, the
aqueous solution penetrates the lipid lm and forms membranous bubbles that eventually become
vesicles (Figure 2a). This type of hydration has certain restrictions regarding the lipid compo-
sitions (ideally charged lipids) and requires low salt concentrations (3). For the standard gentle
hydration method, the sample needs to be incubated above the phase transition temperature of
the lipids, and the process takes an extended amount of time, usually overnight. Many variations
have been developed that improve and extend the method according to various requirements (3,
61, 157), but each approach comes with its own drawbacks.
The most notable variant is the electroformation method. In this approach, swelling is assisted
by an externally applied electric eld (8). The lipid lms are deposited on conductive surfaces
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 529
WATER
OIL
WATER
WATE
R
a
c
OIL
WATE
R
WATER
b
Figure 2
Giant unilamellar vesicle encapsulation methods. (a) Hydration method. Dried lipids form layered lipid
lms. Upon addition of water, lipid lms hydrate, forming vesicular structures. If this process takes place on
an electrode with an applied alternating electric current, then the process can be much more efcient.
(b) Inverted emulsion transfer. A water-in-oil emulsion is prepared in which the lipid monolayer–lined
aqueous droplets contain the protein solution. These droplets then pass through a second, planar water–oil
interface where they become coated with the outer membrane leaet. (c)Microuidicmethodbasedon
double emulsions. Water-in-oil-in-water double emulsions are generated using microuidic
polydimethylsiloxane (PDMS) chips (or with glass capillaries). The oil contains solubilized lipids, so that
lipid monolayers assemble at the two interfaces of each droplet. In a second step (not shown), the oil phase
has to be removed for a bilayer to form.
(usually indium tin oxide–coated glass or platinum wires). Because it uses the alternating electri-
cal eld as an additional driver of the spontaneous rehydration process, this method is much faster
than the standard hydration (which is also called gentle hydration) and produces a higher yield
of larger, more homogenous vesicles. While the electroformation method is preferred by many
over gentle hydration, its major drawbacks come from being mostly incompatible with charged
molecules. As such, this method is largely restricted to phospholipids with a neutral net charge
and also requires solutions with low ionic strengths. Thus, the method is incompatible with most
buffer conditions required for protein functionality (100). Over the years, improved protocols
were proposed such that more complex solutions with physiologically relevant salt concentrations
(100, 118, 135) and even charged lipids (143) could be employed. Nevertheless, the basic principles
530 Litschel •Schwille
of hydration methods render it difcult for large and complex molecules to penetrate the dried
lipid lms during the swelling step. This often results in a low encapsulation efciency or overall
failure to encapsulate these molecules. Following vesicle generation, the surrounding phase needs
to be exchanged (i.e., highly diluted), and proteins that interact with membranes—which, natu-
rally, are often the subject of reconstitution studies with vesicles—complicate the process even
further. Other aspects, like the long incubation times or the low salt concentrations, might not be
problematic if proteins are to be added to the outside solution after vesicle generation but render
these methods problematic for encapsulation experiments.
2.2. Inverted Emulsion Transfer Methods
In recent years, inverted emulsion transfer methods (113) have proven to be among the most suc-
cessful strategies for encapsulating complex solutions into vesicles. While this technique has its
own challenges, many of the struggles that come with hydration methods are circumvented (en-
capsulation efciency, multilamellarity) or far less critical (incubation times, charged molecules).
Generally speaking, there are signicantly fewer limitations to what can be encapsulated with
emulsion transfer, as even large objects, close to the size of the nal vesicle, can be encapsulated.
Emulsion transfer methods, as well as the subsequent methods discussed in this review, are based
on the ability of lipids to assemble into lipid monolayers at water–oil interfaces. This robust pro-
cess can be used to convert lipid-layered water-in-oil droplets into lipid vesicles (Figure 2b). Even
asymmetric bilayers with leaets of different lipid compositions can be produced, as the two mono-
layers that eventually form the vesicle membrane assemble independently (112). While there are
fewer restrictions on the types of lipids that can be used, the nal composition in the vesicle
membrane can differ from the initial composition in the oil due to different interface adsorption
kinetics, which depend on the lipid species.
While this is a very reliable method once established in a laboratory setting, some of the steps
can initially be difcult to reproduce (152, 156). Details such as the composition of the (mineral)
oil used or the humidity of the ambient air can be critical. (For the latter, it can be benecial to use a
humidity-controlled glove box.) If a simple procedure with little specialized equipment is desired,
then the standard emulsion transfer technique developed by Pautot et al. (113) is an excellent
choice, facilitating encapsulation drastically better than hydration methods and with only a few
compromises compared to the methods discussed below, which allow for full or partial control
over vesicle size and other parameters by employing microuidics.
Some of these more sophisticated methods are directly based on the emulsion transfer princi-
ple but additionally utilize microuidic components (90, 102). One such method that has gained
popularity recently is the cDICE (continuous droplet interface crossing encapsulation) method
developed by Massiera and coworkers (2). Particularly in combination with a novel protocol to
prepare lipid-in-oil mixtures developed by the same group (22), this technique has proven suc-
cessful in reconstituting several more complex protein–membrane systems and has been adapted
by other groups (71, 79, 80, 84).
2.3. Microuidic Methods
Another category of methods is based on the microuidic generation of water-in-oil-in-water
double emulsions (107) (Figure 2c). Double emulsion droplets by default resemble vesicles in
that they are aqueous compartments in a surrounding aqueous environment but with a compart-
mentalizing shell made from bulk liquid oil. Double emulsions can be stabilized by lining the
two water–oil interfaces with lipid monolayers, in which case they only differ from vesicles due
to the large quantities of oil trapped between the two leaets of the bilayer. Various methods
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 531
were developed to remove the oil, either by solvent extraction (115, 150) or budding-off of the
oil through interfacial forces (30). In a very similar way, double emulsions can be made in glass
capillaries instead of silicone [polydimethylsiloxane (PDMS)] chips (153). The advantage of glass
is that it is resistant to solvents like toluene or chloroform that are used for the oil extraction
process (137). A method based on a different approach is the so-called jetting technique (42, 142),
which under certain circumstances can also allow for the generation of quasi-oil-free bilayers
(69). One strength shared by the microuidic methods discussed above is that they allow for the
generation of monodisperse vesicles.
The common thread between emulsion transfer and these microuidic methods is the use of
oil as a lipid solvent and water–oil interfaces as scaffolds for lipid monolayer assembly. Generally,
these methods simplify the encapsulation of complex solution and allow exibility regarding the
choice of lipids. However, the mechanism is prone to producing membranes with trace amounts
of oil between the bilayer leaets. The amount of oil is small enough to not measurably alter
the membrane thickness and does not affect most physiologically relevant properties (17, 98) and
biocompatibility [e.g., transmembrane proteins are still readily incorporated into the membrane
(2, 28, 30, 111)], but the minute presence of oil has been shown to result in vesicles with different
dynamical properties in membrane nanotube experiments (17).
Recently, Weiss et al. (158) introduced a new microuidic vesicle generation approach that is
most likely unaffected by these uncertainties. This vesicle generation process relies on the charge-
mediated fusion of SUVs inside a surfactant-stabilized droplet (55), a principle similar to tech-
niques for generating planar supported lipid bilayers (SLBs). While a simplied version of the
protocol was recently presented that does not require specialized equipment (52), the original
method beautifully demonstrated some of the advantages that microuidic techniques can offer:
The technique allows for the sequential addition of several components through pico-injection
(1) into a droplet-stabilized GUV. Weiss et al. showed that this temporal control can be crucial to
reconstituting different features and cellular functionalities within one vesicle (158).
In summary, there are many different approaches for generating GUVs. In many cases, gentle
hydration methods are still the standard for protein reconstitution experiments, in which free-
standing model membranes are required, and addition of protein from the outside is sufcient.
In particular, for quantitative membrane protein biochemistry and biophysics experiments, these
assays guarantee that results are not affected by additional chemicals, such as trace amounts of oil
in the bilayer. However, these methods often turn out to be incompatible with encapsulation of
complex protein solutions. Therefore, it is very promising that, with emulsion transfer methods
and microuidic methods becoming more widespread, practical alternatives are available.
3. SUCCESSFULLY RECONSTITUTED PROTEINS AND PROTEIN
MACHINERIES
3.1. Cytoskeletal Proteins
Phospholipids are not the only biomolecules that self-organize into large-scale assemblies. Cy-
toskeletal proteins polymerize into higher-order structures with emergent properties surpass-
ing those of their individual units, forming elaborate networks that span the entire cell. These
multipurpose structures are critical for myriad cellular processes, such as determining cell shape,
regulating division and migration, and guiding intracellular transport, on scales several orders of
magnitude larger than their building blocks. In this review, we focus on actin laments and micro-
tubules, which have been studied on and in model membranes for decades. Intermediate laments
and septins (now often referred to as the fourth type of eukaryotic cytoskeletal protein) have only
532 Litschel •Schwille
recently come into the spotlight, and in vitro studies are still limited in some respects. However,
we also discuss work involving prokaryotic cytoskeletal proteins, most of which are homologs of
eukaryotic proteins but can act in very different ways.
3.1.1. Actin. Actin not only was one of the rst proteins to be subjected to in vitro experiments
(146), but also remains one of the most prevalent candidates in current biophysical reconstitution
experiments (47, 82). In fact,many of the methodological papers introduced in the previous section
use actin as an example to demonstrate possible applications (2, 141, 157, 158). Thus, in this review,
we use actin as a rst example and as an introduction to protein encapsulation in GUVs.
Many current actin reconstitution studies focus on actin–membrane interactions, as many
actin-binding proteins are directly regulated via interactions with phospholipid bilayers (33). For
these studies, SLBs can be the most straightforward model membrane system owing to their ease of
preparation, stability, ability to form lipid patterns, and compatibility with most light microscopy
methods (74). However, their physical properties, such as lipid diffusivity (121, 140) and deforma-
bility, drastically differ from cellular membranes and also from the free-standing bilayers of lipid
vesicles, rendering the latter potentially more attractive to researchers in the eld. While encap-
sulation of functional actin into vesicles is technically challenging, any of the methods discussed
above can be used to generate vesicles as a membrane scaffold to bind proteins to their outside.
Extensive experiments with actin on the outside of GUVs have been performed to investigate in-
teractions between cytoskeletal proteins and membranes (18–20, 81, 83). The actin cytoskeleton
has been a great system to study the role of compartmentalization and boundaries in the self-
organization of biological systems. As mentioned above, microuidic compartments (31, 138) or
water-in-oil droplets (64, 65, 97) can also be used to study this aspect.
While work with proteins in vesicles has undergone a recent spike in popularity due to the
prospect of creating articial cells, encapsulation of actin in vesicles actually goes as far back as
the late 1980s and early 1990s (23, 78, 95, 96). With emulsion transfer methods and microuidics
not yet existing, these studies used hydration methods, or even older techniques, to encapsulate
actin inside vesicles. As described above, it can be challenging to encapsulate proteins such as
actin under physiological conditions with these methods, not only because of electrostatic inter-
actions, but also because polymerizing laments are unlikely to penetrate the swelling membranes.
To circumvent these issues, actin can initially be encapsulated under nonpolymerizing conditions
with low salt concentrations. Polymerization is then induced by ion transport through the mem-
brane, for example, through ion carriers like valinomycin or through electroporation. Actin in
combination with bundling protein was reconstituted in GUVs by Honda et al. (60) and later sys-
tematically characterized by Tsai & Koenderink (151) (Figure 3a), who observed that proteins
like fascin, which cross-links actin into thick bundles, cause spike-like membrane protrusions or
deform the entire vesicle into a rod-like shape, a process that is reversible when actin bundles
are subsequently disassembled (12, 79). Limozin & Sackmann (77) observed that actin bundles
can spontaneously form rings (and supercoiled rings) in vesicle connement (Figure 3b), and re-
cently, Litschel et al. (79) have come one step closer to reconstituting cell division by anchoring
actin rings to the membrane and showing how this can lead to membrane deformations.
One frequently reconstituted physiological feature is the cell cortex, which in vivo primarily
consists of cross-linked mesh-like actin networks. Early reports of a reconstituted actin cortex
showed the formation of shell-like actin structures that form spontaneously without the addition
of actin–membrane linkers (53, 76) (Figure 3c) and that have also been studied in detail for their
mechanical properties (114, 130, 131, 163). The interactions between actin and the membrane in
these vesicles are nonspecic and typically mediated by magnesium ions, a mechanism that has
not yet been shown to be relevant in vivo (132). Later work focused on reconstituting a thicker,
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 533
Prokaryotic cytoskeletonActin Microtubules
b
b
c
c
a
a
e
e
d
d
g
g
f
f
h
h
i
i
j
j
k
k
l
l
m
m
o
o
n
n
+/–
–/+
+
+
–
66 s 834 s603 s
Figure 3
Giant unilamellar vesicles (GUVs) containing cytoskeletal assemblies. (a) Actin bundles in osmotically deated vesicles can drastically
determine vesicle shape. Panel adapted with permission from Reference 151. (b) Bundled actin can form rings or even supercoiled rings
in vesicle connement. Panel adapted with permission from Reference 77. (c) Membrane-bound actin bundles form a cortex and align
in parallel. Panel adapted with permission from Reference 76. (d) Vesicle blebbing through contracting actomyosin cortex. Panel
adapted with permission from Reference 84. (e) Deformed active actin cortex upon vesicle adhesion to a surface. Panel adapted with
permission from Reference 86. ( f)Membranetensionbendsmicrotubule(MT)bundlesinthevesicle.Paneladaptedwithpermission
from Reference 45. (g) MT bending can be controlled through micropipette aspiration. Despite the appearance, MTs are fully
contained within the vesicle membrane. Panel adapted with permission from Reference 44. (h) Membrane stiffness determines
organization of encapsulated MTs. Panel adapted with permission from Reference 116. (i) Long membrane protrusions in the
MT-containing vesicle. Panel adapted with permission from Reference 38. (j)Lightactivationallowsswitchingbetweendifferent
states of the MT-containing vesicle. Panel adapted with permission from Reference 129. (k)DynamicvesiclecontainingMTsand
motor proteins. Panel adapted with permission from Reference 71. (l) The FtsZ-ring divides the GUV. Panel adapted with permission
from Reference 110. (m)Heavydeformationofavesiclecausedbymembrane-boundFtsZ.Paneladaptedwithpermissionfrom
Reference 66. (n) MreB controls vesicle shape. Panel adapted with permission from Reference 48. (o) Periodic vesicle deformations
caused by Min protein oscillations. Panel adapted with permission from Reference 80. All scale bars are 5 µm.
mesh-like cortex by using actin cross-linkers, such as Arp2/3, and by binding actin through specic
interactions to the vesicle membrane (93, 101, 117). Myosin activity in the cell cortex is crucial for
dynamic remodeling and force exertion and has been the subject of several recent reconstitution
approaches of actin cortices (20, 86, 152). Specically, the Bausch group has been using acto-
myosin systems within vesicles to reconstitute features like blebbing of vesicles (84) (Figure 3d),
adhesion (86) (Figure 3e), and cell protrusions (34) with physiologically relevant actin-binding
proteins.
3.1.2. Microtubules. While actin organization heavily depends on cross-linking proteins to
achieve the various properties required for cell functions, single microtubules (MTs) are often
spaced far apart and can act individually as transport tracks or structural elements (94). MTs are
able to achieve this because they have a much greater bending stiffness (exural rigidity) com-
pared to actin, such that even single MTs are strong enough to resist compressive forces without
buckling. When aligned into bundles, this rigidity naturally increases. Due to these mechanical
properties, visible membrane deformations were observed in early encapsulation experiments with
534 Litschel •Schwille
dark-eld microscopy (62), consistent with current studies. Specically, the Libchaber group fo-
cused on the interplay between vesicles and microtubules, i.e., deformations of the membrane (45)
(Figure 3f) and buckling of the microtubules (37, 44) (Figure 3g).
While actin encapsulation in vesicles has often been performed with the aim of reconstituting
dynamic processes, such as membrane remodeling and motility-related functions, the motivation
for encapsulating MTs is less obvious, as they play more indirect roles in concert with membranes
in vivo. However, encapsulating dynamic MTs is nevertheless illuminating, particularly when aim-
ing at a holistic model for articial cells or more sturdy mimicries of living systems. Some groups
have begun using the term molecular robots (54) for vesicles with encapsulated dynamic MTs
(58, 129) (Figure 3j). Dogic and coworkers developed an active nematic system based on MTs
with modied kinesin motors. Under connement, not only did these components show nematic
alignment on spherical surfaces, but the system was also shown to actively deform vesicles (71)
(Figure 3k) and autonomously move droplets through internal MT sliding (128).
3.1.3. Prokaryotic cytoskeleton. While work with actin and MTs has been popular in recon-
stitution experiments for decades, prokaryotic cytoskeletal systems, although generally lower in
complexity and thus simpler to recapitulate in vitro, can be more challenging, often due to the
smaller scale of the cytoskeletal structures. While a contractile actomyosin ring in a eukaryotic
cell is an easily discernible structure for contemporary uorescence microscopes, a bacterial Z
ring can be less than 1 micron in size and thus close to the microscope’s diffraction limit. This
factor also plays a role when attempting to reconstitute these protein systems within vesicles,
and, in fact, vesicles the size of bacterial cells barely fall under common denitions of GUVs (1–
200 µm). Nevertheless, exciting discoveries have been made through reconstitution of prokaryotic
cytoskeleton proteins in GUVs, either by using sufciently small vesicles or because the reconsti-
tuted features were scalable to larger sizes.
The most notable prokaryotic cytoskeletal protein might be the cell division protein FtsZ. FtsZ
is a tubulin homolog found in almost all bacterial cells and even in chloroplasts and some mito-
chondria. The rst reconstitution approaches of FtsZ in vesicles were performed by Erickson and
colleagues, initially in tubular multilamellar vesicles (109) and later in more cell-like unilamellar
vesicles (110). The major nding in these studies was that FtsZ assembles into rings in small GUVs
and can constrict or even divide vesicles, similar to the proposed function in vivo (Figure 3l).
Later studies demonstrated that larger vesicles can also be useful to investigate aspects like FtsZ-
mediated membrane deformation, even though FtsZ bundles are smaller in size than these vesicles
(16, 43, 46, 85, 122) (Figure 3m).
Different bacteria make use of a variety of different actin homologs, which were found to fulll
a diverse array of functions. MreB is such a protein and has been used to demonstrate protein
expression in vesicles (87, 89). Recent work showed that MreB bundles form a cortex and govern
vesicle shape driven by membrane crowding (48, 49) (Figure 3n).
3.2. Membrane Proteins
Membrane proteins have myriads of different functions and give each type of cell membrane its
characteristic properties. They account for about half of the mass of a typical plasma membrane
and even more in other types of membranes (4). Membrane proteins can be broadly classied
into two categories based on whether they are transiently or permanently associated with the
membrane: peripheral membrane proteins and intrinsic membrane proteins. The most common
type of intrinsic membrane proteins are transmembrane proteins.
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 535
3.2.1. Peripheral membrane proteins. Peripheral membrane proteins bind to membranes
transiently. They are cytosolic but can interact with or insert into a single leaet of the bi-
layer. These proteins often alternate between membrane-bound and unbound states, which can
correspond to physiologically active and inactive states of the protein. Peripheral membrane pro-
teins play important regulatory roles in many processes, including signaling, channel regulation,
and membrane remodeling.
The Min proteins from Escherichia coli are prime examples of transiently binding membrane
proteins. Like FtsZ (see Section 3.1.3.), the Min protein system is part of the bacterial cell division
machinery. In many bacteria, FtsZ is spatially positioned by the proteins MinD, MinE, and MinC
in the process of cell division. In E. coli, the Min proteins oscillate between the two cell poles,
based on a reaction–diffusion mechanism, and thereby direct the Z ring to the cell center. The
impressive self-organization dynamics of Min proteins, which is only unfolded in the presence
of membranes, makes them a highly attractive model system for studies with model membranes.
By repeatedly binding and unbinding from the membrane in a coordinated fashion, Min proteins
generate observable oscillatory reaction–diffusion patterns. Reaction–diffusion mechanisms are
susceptible to their connement geometry, and compartmentalization of the Min proteins leads
to in vivo–like standing wave oscillation (21, 164, 165), different from the traveling waves observed
in open systems. When reconstituted in GUVs, the membrane detachment and attachment of the
proteins periodically deformed the vesicles, resembling vesicle division (51, 80) (Figure 3o). As
such, encapsulating the Min system illustrates a great example of the benets of encapsulation
in lipid vesicles, both demonstrating an effect of limiting the reaction space by connement and
showing how protein–membrane interactions can lead to membrane deformations.
For Min proteins, the observation of membrane deformations is somewhat unintuitive, as these
proteins are not thought to be directly responsible for membrane remodeling in vivo. However,
many peripheral membrane proteins are primary players in such processes, and have therefore
been studied in vitro, particularly in the context of membrane remodeling. Figure 4bshows dif-
ferent mechanisms through which peripheral membrane proteins can determine membrane shape:
(a) insertion into the membrane, thereby inducing membrane bending; (b) imposition of their own
shape onto the membrane; (c) scaffold assembly; and (d) steric hindrance, whereby crowding of
membrane-bound proteins can cause the membrane to buckle. Oftentimes, peripheral membrane
proteins use a combination of these mechanisms to deform lipid bilayers and regulate membrane
shape.
GUVs, because they are deformable, free-standing membranes with negligible curvature on
the molecular scale, are ideally suited to work with peripheral membrane-transforming proteins,
as proteins can often just be added to the outside of vesicles. For example, vesicular trafcking pro-
teins like clathrin and COPII were reconstituted on GUVs, leading to membrane budding and
tubulation (10, 127). However, encapsulation inside vesicles has proven to be important to achiev-
ing certain cell-like membrane topologies. By combining protein encapsulation in vesicles with
membrane nanotube pulling using optical tweezers and micropipettes, Bassereau and coworkers
generated highly negative curvatures (120) and neck-like geometries (14, 26). These physiologi-
cally relevant topologies can be used to investigate curvature sensing and other properties of pe-
ripheral membrane proteins. With this technique, Prévost et al. (120) were recently able to show
that the BAR domain protein IRSp53 preferentially binds to negatively curved membranes due to
its intrinsic curvature.
In different work (14, 26), the same group encapsulated the ESCRT-III protein CHMP2B in
GUVs. In this case, the protein bound preferentially to membrane topologies resembling den-
dritic spines. Interestingly, membrane-bound CHMP2B formed a diffusion barrier for lipids and
other membrane-bound proteins. Additionally, in combination with other ESCRT-III complex
536 Litschel •Schwille
Insertion
a
c
iv
iii
i
ii
b
Peripheral
membrane
proteins
Transmembrane proteins
Direct
incorporation
Incorporation
after encapsulation
Articial
organelles
Protein shape
Scaolding
Crowding
Hydration
from
proteo-
liposomes
Hydration
from
proteo-
liposomes
Figure 4
Graphical illustrations of membrane proteins. (a) Different types of membrane proteins. (b)Mechanismsof
membrane remodeling by peripheral membrane proteins. Membrane curvature can be affected by insertion
of hydrophobic protein motifs; by large proteins with several binding sites imposing their shape onto the
membrane; through large-scale assemblies of proteins; or through protein crowding, where forces are
exerted on the membrane due to steric collisions between proteins. For many membrane remodeling
proteins, the exact mechanism is unknown and is likely a combination of some of the above. (c)Different
methods for incorporating transmembrane proteins into giant unilamellar vesicles (GUVs) for in vitro
reconstitution. From left to right: (i) GUVs can be generated from dried lipid lms containing the proteins.
By incorporating transmembrane proteins into pre-existing GUVs either (ii)fromtheoutsideor(iii)after
encapsulation, the orientation of the transmembrane proteins can be controlled. (iv) Protein function can
also be utilized by encapsulating transmembrane-containing small unilamellar vesicles (SUVs) as articial
organelles.
proteins, CHMP2B can deform pulled membrane nanotubes into corkscrew-like morphologies
(14). In archaeal cells, homologs of these eukaryotic ESCRT proteins are responsible for cell
division. Preliminary work with these archaeal homologs reconstituted inside GUVs showed
invaginations of the vesicle membrane (56). These experiments demonstrate the potential for
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 537
using encapsulation to study the cellular membrane remodeling machinery, and reconstitution
of such proteins offers another potential approach to achieving division of articial vesicle
compartments.
3.2.2. Transmembrane proteins. While GUVs are an ideal model system for directly ob-
serving membrane remodeling and higher-order protein assemblies on membranes, most trans-
membrane proteins do not exhibit activities directly observable by light microscopy. Despite this,
GUVs can be particularly valuable when studying transmembrane proteins, not only because free-
standing bilayers are a prerequisite for unfolding most of their functions, but also because having
compartmentalized volumes can be incredibly useful. Maintaining two separated aqueous phases
is key to studying membrane potentials and concentration gradients created by channels, carri-
ers, and pumps. Biophysicists have been studying membrane transport proteins for almost half
a century—both in cells, using methods like patch clamp, and in vitro, using black lipid mem-
branes or painted bilayers and smaller liposomes. Transmembrane proteins were incorporated
into membranes of giant vesicles early on using hydration methods and even older techniques
(24, 67). While some channel proteins are easily reconstituted,studies of other proteins have only
been possible owing to recent methodological advances.
Figure 4cillustrates different methods to incorporate transmembrane proteins into GUVs.
The rst step for the reconstitution of transmembrane proteins in GUVs is usually the formation
of SUVs or LUVs with incorporated proteins, which we refer to as proteoliposomes. As trans-
membrane proteins have large hydrophobic regions, they typically need to be solubilized using
detergents during protein purication. Proteoliposomes are formed by mixing solubilized pro-
tein with lipid-detergent micelles and subsequently removing the detergent. When dried down
into lipid lms, these proteoliposomes can be used to form GUVs with straightforward hydra-
tion methods (25, 50). However, the drying step can be harmful to some membrane proteins. To
reconstitute more delicate proteins, incorporation into preformed GUVs can be benecial by al-
lowing one to avoid the dehydration step. This is possible either by fusion of proteoliposomes
with preformed GUVs (32, 68) or by omitting the step of proteoliposome formation altogether
and directly incorporating detergent-solubilized proteins into GUVs (32).
Reconstitution of transmembrane channels into GUV membranes has a long history (67) and
can only be touched upon, as we predominantly focus on GUVs as containers. However, below,
we quickly highlight some examples that have proven useful as tools in contemporary in vitro
reconstitution studies. As mentioned above, ion carriers can be used to create physiological salt
concentrations within vesicles post–vesicle formation, which is especially useful when vesicles are
formed in low-salt conditions with hydration methods. Ion carriers do not have to be proteins, but
can instead be peptides (or even smaller molecules), such as the potassium carrier valinomycin,
which has often been used for these applications (23, 32, 95). Another commonly used channel
protein is the cytotoxic protein α-hemolysin, which, as a heptamer, forms large water-lled pores
that allow even small biomolecules to pass through the membrane. α-Hemolysin is a popular tool
in many reconstitution experiments. It can even be used to introduce ATP into vesicles, thereby
triggering biological reactions (155) such as actin polymerization. Often, α-hemolysin is used to
introduce a uorescent dye into the vesicle lumen as a control to demonstrate membrane func-
tionality (2, 28, 30, 35, 98, 103, 111, 142). Microuidic chips with vesicle traps can be particularly
useful for these kinds of assays, as they hold the vesicle in place while exchanging the surrounding
solution (to incorporate α-hemolysin or add a uorescent dye), so that the same vesicle can be im-
aged over the course of the experiment (126). Recently, transmembrane pores and channels have
again come into focus in the context of creating articial cells, as groups are trying to engineer
primitive means of communication between these compartments (9, 123).
538 Litschel •Schwille
Membrane proteins are generally asymmetric, possessing two hydrophilic parts with very dif-
ferent properties, sizes, and function. Even large pores like α-hemolysin transport asymmetrically
(92). Therefore, the orientation of the incorporated transmembrane proteins generally matters
for their functionality. The methods described above for membrane protein incorporation involv-
ing the use of proteoliposomes typically result in more or less randomly oriented transmembrane
proteins, which in most cases is a satisfactory outcome. Often, the wrongly incorporated proteins
are simply rendered dysfunctional due to their orientation and have a negligible effect on the out-
come of the experiment, leaving the membrane with sufcient amounts of functional proteins. For
example, even highly asymmetric transmembrane proteins, like the focal adhesion protein inte-
grin (41, 144) or the SNARE protein synaptobrevin (11, 148), have been reconstituted using such
approaches. For both of these examples, the physiologically accurate orientation has the larger
hydrophilic part outside of the vesicle; however, outside-in coincorporated proteins do not affect
these reconstitution experiments negatively.
For certain proteins, controlling orientation can be important—for example, for proton pumps
like bacteriorhodopsin, where a pH gradient will be generated across the vesicle membrane only
with a preferential asymmetric incorporation (32, 50, 68). This can be achieved with direct
incorporation of detergent-solubilized transmembrane proteins into preformed vesicles (32).
If membrane proteins are incorporated from the outside solution into the vesicle membrane,
then proteins are inserted with their most hydrophilic domain pointing outward. In the case of
bacteriorhodopsin, this results in cytoplasmic regions that face outward (inside-out orientation)
but that also create a proton surplus within the vesicles, as is potentially desired (32).
In cells, the more hydrophilic regions of transmembrane proteins usually face the cytosol [to
be more precise, generally, the more positively charged part faces inward (57)]. To preferentially
reconstitute this opposing orientation, encapsulation methods are used (7, 125, 160). Via encapsu-
lation of solubilized membrane proteins, proteins can be directed to orient their most hydrophilic
part toward the vesicle center. Using this approach, physiologically oriented incorporation of pro-
teins has been demonstrated for potassium channels (160); for photosynthetic reaction centers (7);
and, with a comparable approach, for SNARE proteins (125).
ATP synthases make up an important class of transmembrane proteins that use proton gradi-
ents to produce ATP; as such, researchers have long attempted to coreconstitute these proteins
with bacteriorhodopsin. With the goal of establishing photosynthesis of ATP, this system has been
of special interest in the context of synthetic cells. Again, asymmetric incorporation is crucial, as
a proton gradient is required to successfully generate ATP inside the vesicle. Interestingly, pho-
tosynthetic ATP generation has been achieved not by incorporating the proteins into the GUV
membrane itself, but by encapsulating proteoliposomes (SUVs) containing both proteins into the
GUVs without merging them with the GUV membrane (13, 75). The proteins are incorporated
inside-out into the SUVs, such that protons are pumped into the SUVs, and ATP is generated
in the GUV lumen. Remarkably, these synthetic organelles have been demonstrated to produce
sufcient ATP to power actin assembly (75) and synthesis of green uorescent protein (GFP) (13).
3.3. Enzymes
Enzymes are biological catalysts that guide networks of chemical transformations in cells. Acting
in organized sequences, enzymes coordinate the many stepwise reactions in metabolic pathways
by which nutrients are degraded, chemical energy is transformed and stored, and larger molecules
are synthesized from simple precursors. As many cellular reactions would not be possible or would
take years without catalysis, enzymes are ubiquitously required but, at the same time, highly spa-
tially targeted. Through spatially and temporally regulated expression of enzymes, the cell can
control which of the many possible chemical reactions actually take place.
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 539
3.3.1. Metabolic reactions. For biological reactions, spatial connement can be a dening fac-
tor.While the effect of space and scale is especially pronounced for reaction–diffusion mechanisms
like the Min system (see above), metabolic reactions are affected as well. Virk et al. (154) found
that, when encapsulated inside GUVs, an alcohol oxidase enzyme was 3.5 times more active and 20
times more stable than in bulk. They argued that encapsulation of enzymes in vesicles can protect
the enzyme from proteases or self-denaturation. Nonspecic interactions with the encapsulating
membrane have also been thought to have stabilizing effects on enzymes (161).
Generally, GUVs are relatively large compared to the molecular scale, and effects of conne-
ment are much more pronounced in smaller vesicle compartments like SUVs, where stochastic
effects come into play (72). However, for some complex biochemical reactions, GUV conne-
ment can be a determining factor for reaction dynamics. This is particularly the case for reaction–
diffusion systems, like the above-mentioned Min protein system (80), where biochemical reactions
are spatially coupled. Traveling waves can thus form with wavelengths on the scale of the GUVs
themselves. In this case, the presence of a conning membrane is of great relevance, as it induces
a symmetry break with respect to free 3D diffusion.
Membranes, as conning boundaries, further enable intracellular compartmentalization, a key
feature of all eukaryotic life forms. The separation of the intracellular space into membrane-bound
compartments allows crucial biochemical reactions to take place in optimized microenvironments.
Therefore, imitating cell organelles has been the subject of several in vitro studies in the recent
years (29, 52, 75). Elani et al. (35, 36) achieved remarkably complex spatial organization of enzy-
matic reactions within vesicles by creating multicompartment supergiant vesicles with multistep
enzymatic cascades. In this case, the different steps of the reaction pathway are isolated by bi-
layers in equally sized vesicle compartments but connected via membrane channels, allowing the
products of each step to traverse between the compartments (35). Elani et al. demonstrated the
conversion of lactose into the uorescent molecule resorun in a three-step signaling cascade,
with each of the steps taking place in a separate compartment. In another study, they encapsulated
eukaryotic cells in GUVs, which contribute, as organelle-like modules, to a similar metabolic path-
way (36). While the engineered eukaryotic cells only carry out the rst step by converting lactose
into glucose, a synthetic enzymatic cascade then processes glucose into the uorescent product.
Elani et al. also showed that the vesicle membrane can even help protect the encapsulated cells
from a surrounding toxic solution.
3.3.2. Transcription–translation systems. From the point of view of creating synthetic cells,
protein synthesis in articial lipid vesicles has been considered one of the greatest milestones. The
rst attempts at vesicle-encapsulated transcription were reported by the Luisi group, who synthe-
sized short polypeptides with essential transcription components in small vesicles (105). Shortly
after, Yu et al. (162) reported the rst full protein synthesis inside giant vesicles. By encapsulating
cell extract with nucleotides, amino acids, T7 RNA polymerase, and plasmids, they reconstituted
both transcription and translation inside GUVs and were able to observe the synthesis of GFP by
uorescence microscopy (104, 162). An impressive demonstration of the forthcoming advances in
the eld was given by Noireaux & Libchaber (103), who fused the genes of eGFP and α-hemolysin
and showed the successful synthesis and incorporation of the labeled transmembrane complex into
the membrane from within the vesicle.
In the past few years, in vitro transcription and translation (TXTL) has developed into a pow-
erful tool. This method provides a viable, convenient alternative to the traditional approach of
generating pure proteins via expression in cells and is thus often referred to as cell-free protein
synthesis. This success is the result of continuous improvements of the systems that are used.
Modern cell-free protein synthesis allows for the production of large quantities of protein in just
540 Litschel •Schwille
a few hours. One milestone in the eld was the development of the PURE (protein synthesis us-
ing recombinant elements) system, which is a protein expression system completely reconstituted
from puried proteins without the use of cell extract (136). New commercial TXTL kits (such as
new versions of the PURE system) are constantly being developed with improved efciency and
variability for cell-free protein expression.
The convenience and protein yield of these systems have improved greatly within the past
few years, and TXTL has become a true alternative to traditional protein expression and puri-
cation for reconstitution experiments, even within GUVs. In fact, some of the above-mentioned
examples of reconstitution within vesicles were conducted using TXTL in vesicles, rather than by
encapsulating puried proteins (43, 48, 49, 51, 87, 89). Since the same encapsulation procedure
can be used for the production of very different proteins, this could allow for streamlined proto-
cols for the reconstitution of proteins in vesicles. In the case of transmembrane proteins, TXTL
in vesicles can even simplify reconstitution experiments (88, 139). For example, expression of the
bacterial transporter protein EmrE has shown that some of the technical challenges that come
with the reconstitution of puried transmembrane proteins (see above) can be circumvented by
synthesizing transmembrane proteins directly inside GUVs (139). Although the mechanism of the
protein insertion is not fully understood (and must be different from both in vivo and conventional
transmembrane reconstitution), more than 20% of the synthesized proteins can be incorporated
into the vesicle membrane (139). A more cell-like approach is also being pursued by reconstituting
translocons entirely through in vitro TXTL in GUVs (106).
One of the greatest challenges of the eld of synthetic biology is lipid synthesis, as it provides
the basis for vesicle surface growth (39). Kuruma et al. (73) have taken on this task, demonstrating
an enzymatic cascade linking the synthesis of membrane proteins to synthesis and incorporation
of lipids into the membrane. Using in vitro TXTL, two acyltransferases are synthesized inside
small liposomes, which in turn produce phosphatic acids (the simplest forms of phospholipids)
and incorporate them into the liposome membrane (73, 134). So far, these experiments have only
been performed in SUVs; even if they are reproduced in GUVs, the yield of the reactions in the
current form would be too low to result in observable changes in membrane area (15). However,
reconstituting these two features—protein and membrane synthesis—within GUVs would be a
great leap toward achieving replication of synthetic vesicles, and thus toward creating life-like
articial cells (39).
4. CONCLUSIONS AND OUTLOOK
GUVs have long been a popular model system to study the functionality of proteins in dened
membrane environments. Through varying the membrane composition of GUVs, specic
protein–protein and protein–lipid interactions could be reproducibly studied in great detail, in
particular by uorescence microscopy and spectroscopy. GUVs were also considered ideal starting
points for a potential bottom-up assembly of functional biological modules into minimal proto-
cells. However, traditional GUV production methods often make the incorporation of functional
proteins into the membrane or the lumen of vesicles extremely difcult, which has hampered the
progress of many studies and restricted the use of the GUV model membrane system to a limited
number of expert labs in the biomembrane eld. Recent breakthroughs, mainly from microsystems
technology, have given us several simple and versatile methods that drastically simplify the produc-
tion of large membrane-enclosed protein reaction systems. The development toward GUV-based
protocells with complex functionalities will likely dramatically benet from these technical break-
throughs, promising an enormous acceleration of work with the aim to construct a minimal living
cell—a goal that only recently seemed decades away but may actually be closer than we think.
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 541
DISCLOSURE STATEMENT
The authors are not aware of any afliations, memberships, funding, or nancial holdings that
might be perceived as affecting the objectivity of this review.
ACKNOWLEDGMENTS
We thank Charlotte F. Kelley, Allen P. Liu, and Yashar Bashirzadeh for suggestions and proof-
reading. We thank Kerstin Göpfrich for helpful discussions. This work is part of the MaxSyn-
Bio consortium, which is jointly funded by the Federal Ministry of Education and Research of
Germany and the Max Planck Society.
LITERATURE CITED
1. Abate AR, Hung T, Mary P, Agresti JJ, Weitz DA. 2010. High-throughput injection with microuidics
using picoinjectors. PNAS 107:19163–66
2. Abkarian M, Loiseau E, Massiera G. 2011. Continuous droplet interface crossing encapsulation (cDICE)
for high throughput monodisperse vesicle design. Soft Matter 7:4610–14
3. Akashi K, Miyata H, Itoh H, Kinosita K. 1996. Preparation of giant liposomes in physiological conditions
and their characterization under an optical microscope. Biophys. J. 71:3242–50
4. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, et al. 2002. Molecular Biology of the Cell. New York:
Garland Sci. 4th ed.
5. Allen TM, Cullis PR. 2004. Drug delivery systems: entering the mainstream. Science 303:1818–22
6. Almén MS, Nordström KJV, Fredriksson R, Schiöth HB. 2009. Mapping the human membrane pro-
teome: A majority of the human membrane proteins can be classied according to function and evolu-
tionary origin. BMC Biol.7:50
7. Altamura E, Milano F, Tangorra RR, Trotta M, Omar OH, et al. 2017. Highly oriented photosynthetic
reaction centers generate a proton gradient in synthetic protocells. PNAS 114:3837–42
8. Angelova MI, Dimitrov DS. 1986. Liposome electroformation. Faraday Discuss. Chem. Soc. 81:303–11
9. Aunger L, Simmel FC. 2019. Establishing communication between articial cells. Chemistry 25:12659–
70
10. Bacia K, Futai E, Prinz S, Meister A, Daum S, et al. 2011. Multibudded tubules formed by COPII on
articial liposomes. Sci. Rep. 1:17
11. Bacia K, Schuette CG, Kahya N, Jahn R, Schwille P. 2004. SNAREs prefer liquid-disordered over “raft”
(liquid-ordered) domains when reconstituted into giant unilamellar vesicles. J. Biol. Chem. 279:37951–55
12. Bashirzadeh Y, Wubshet NH, Liu AP. 2020. Connement geometry tunes fascin-actin bundle structures
and consequently the shape of a lipid bilayer vesicle. Front. Mol. Biosci. 7:337
13. Berhanu S, Ueda T, Kuruma Y. 2019. Articial photosynthetic cell producing energy for protein syn-
thesis. Nat. Commun. 10:1325
14. Bertin A, de Franceschi N, de la Mora E, Maiti S, Alqabandi M, et al. 2020. Human ESCRT-III polymers
assemble on positively curved membranes and induce helical membrane tube formation. Nat. Commun.
11:2663
15. Blanken D, Foschepoth D, Serrão AC, Danelon C. 2020. Genetically controlled membrane synthesis in
liposomes. Nat. Commun. 11:4317
16. Cabre EJ, Sanchez-Gorostiaga A, Carrara P, Ropero N, Casanova M, et al. 2013. Bacterial division pro-
teins FtsZ and ZipA induce vesicle shrinkage and cell membrane invagination. J. Biol. Chem. 288:26625–
34
17. Campillo C, Sens P, Köster D, Pontani L-L, Lévy D, et al. 2013. Unexpected membrane dynamics
unveiled by membrane nanotube extrusion. Biophys. J. 104:1248–56
18. Caorsi V, Lemière J, Campillo C, Bussonnier M, Manzi J, et al. 2016. Cell-sized liposome doublets reveal
active tension build-up driven by acto-myosin dynamics. Soft Matter 12:6223–31
542 Litschel •Schwille
19. Carvalho K, Lemiere J, Faqir F, Manzi J, Blanchoin L, et al. 2013. Actin polymerization or myosin
contraction: two ways to build up cortical tension for symmetry breaking. Philos. Trans. R. Soc. Lond. B
368:20130005
20. Carvalho K, Tsai FC, Lees E, Voituriez R, Koenderink GH, Sykes C. 2013. Cell-sized liposomes reveal
how actomyosin cortical tension drives shape change. PNAS 110:16456–61
21. Caspi Y, Dekker C. 2016. Mapping out Min protein patterns in fully conned uidic chambers. eLife
5:e19271
22. Claudet C, In M, Massiera G. 2016. Method to disperse lipids as aggregates in oil for bilayers production.
Eur. Phys. J. E 39:9
23. Cortese JD, Schwab B 3rd, Frieden C, Elson EL. 1989. Actin polymerization induces a shape change in
actin-containing vesicles. PNAS 86:5773–77
24. Criado M, Keller BU. 1987. A membrane fusion strategy for single-channel recordings of membranes
usually non-accessible to patch-clamp pipette electrodes. FEBS Lett. 224:172–76
25. Darszon A, Vandenberg CA, Schönfeld M, Ellisman MH, Spitzer NC, Montal M. 1980. Reassembly
of protein-lipid complexes into large bilayer vesicles: perspectives for membrane reconstitution. PNAS
77:239–43
26. De Franceschi N, Alqabandi M, Miguet N, Caillat C, Mangenot S, et al. 2019. The ESCRT protein
CHMP2B acts as a diffusion barrier on reconstituted membrane necks. J. Cell Sci. 132:jcs217968
27. Deamer D, Dworkin JP, Sandford SA, Bernstein MP, Allamandola LJ. 2002. The rst cell membranes.
Astrobiology 2:371–81
28. Deng N-N, Yelleswarapu M, Huck WTS. 2016. Monodisperse uni- and multicompartment liposomes.
J. Am. Chem. Soc. 138:7584–91
29. Deng N-N, Yelleswarapu M, Zheng L, Huck WTS. 2017. Microuidic assembly of monodisperse veso-
somes as articial cell models. J. Am. Chem. Soc. 139:587–90
30. Deshpande S, Caspi Y, Meijering AEC, Dekker C. 2016. Octanol-assisted liposome assembly on chip.
Nat. Commun. 7:10447
31. Deshpande S, Pfohl T. 2015. Real-time dynamics of emerging actin networks in cell-mimicking com-
partments. PLOS ONE 10:e0116521
32. Dezi M, Di Cicco A, Bassereau P, Lévy D. 2013. Detergent-mediated incorporation of transmembrane
proteins in giant unilamellar vesicles with controlled physiological contents. PNAS 110:7276–81
33. Doherty GJ, McMahon HT. 2008. Mediation, modulation, and consequences of membrane-
cytoskeleton interactions. Annu. Rev. Biophys. 37:65–95
34. Dürre K, Keber FC, Bleicher P, Brauns F, Cyron CJ, et al. 2018. Capping protein-controlled actin
polymerization shapes lipid membranes. Nat. Commun. 9:1630
35. Elani Y, Gee A, Law RV, Ces O. 2013. Engineering multi-compartment vesicle networks. Chem. Sci.
4:3332–38
36. Elani Y, Trantidou T, Wylie D, Dekker L, Polizzi K, et al. 2018. Constructing vesicle-based articial
cells with embedded living cells as organelle-like modules. Sci. Rep. 8:4564
37. Elbaum M, Fygenson DK, Libchaber A. 1996. Buckling microtubules in vesicles. Phys.Rev. Lett. 76:4078–
81
38. Emsellem V, Cardoso O, Tabeling P. 1998. Vesicle deformation by microtubules: a phase diagram. Phys.
Rev. E 58:4807–10
39. Exterkate M, Driessen AJM. 2019. Synthetic minimal cell: self-reproduction of the boundary layer. ACS
Omega 4:5293–303
40. Faivre-Moskalenko C, Dogterom M. 2002. Dynamics of microtubule asters in microfabricated cham-
bers: the role of catastrophes. PNAS 99:16788–93
41. Frohnmayer JP, Brüggemann D, Eberhard C, Neubauer S, Mollenhauer C, et al. 2015. Minimal syn-
thetic cells to study integrin-mediated adhesion. Angew. Chem. Int. Ed. 54:12472–78
42. Funakoshi K, Suzuki H, Takeuchi S. 2007. Formation of giant lipid vesiclelike compartments from a
planar lipid membrane by a pulsed jet ow. J. Am. Chem. Soc. 129:12608–9
43. Furusato T, Horie F, Matsubayashi HT, Amikura K, Kuruma Y, Ueda T. 2018. De novo synthesis of
basal bacterial cell division proteins FtsZ, FtsA, and ZipA inside giant vesicles. ACS Synth. Biol. 7:953–61
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 543
44. Fygenson DK, Elbaum M, Shraiman B, Libchaber A. 1997. Microtubules and vesicles under controlled
tension. Phys. Rev. E 55:850–59
45. Fygenson DK, Marko JF, Libchaber A. 1997. Mechanics of microtubule-based membrane extension.
Phys. Rev. Lett. 79:4497–500
46. Ganzinger KA, Merino-Salomón A, García-Soriano DA, Buttereld AN, Litschel T, et al. 2020. FtsZ
reorganization facilitates deformation of giant vesicles in microuidic traps. Angew. Chem. Int. Ed.
59:21372–76
47. Ganzinger KA, Schwille P. 2019. More from less—bottom-up reconstitution of cell biology. J. Cell Sci.
132:jcs227488
48. Garenne D, Libchaber A, Noireaux V. 2020. Membrane molecular crowding enhances MreB polymer-
ization to shape synthetic cells from spheres to rods. PNAS 117:1902–9
49. Garenne D, Noireaux V. 2020. Analysis of cytoplasmic and membrane molecular crowding in genetically
programmed synthetic cells. Biomacromolecules 21:2808–17
50. Girard P, Pécréaux J, Lenoir G, Falson P, Rigaud J-L, Bassereau P. 2004. A new method for the recon-
stitution of membrane proteins into giant unilamellar vesicles. Biophys. J. 87:419–29
51. Godino E, López JN, Foschepoth D, Cleij C, Doerr A, et al. 2019. De novo synthesized Min pro-
teins drive oscillatory liposome deformation and regulate FtsA-FtsZ cytoskeletal patterns. Nat. Commun.
10:4969
52. Göpfrich K, Haller B, Staufer O, Dreher Y, Mersdorf U,et al. 2019. One-pot assembly of complex giant
unilamellar vesicle-based synthetic cells. ACS Synth. Biol. 8:937–47
53. Häckl W, Bärmann M, Sackmann E. 1998. Shape changes of self-assembled actin bilayer composite
membranes. Phys. Rev. Lett. 80:1786–89
54. Hagiya M, Konagaya A, Kobayashi S, Saito H, Murata S. 2014. Molecular robots with sensors and in-
telligence. Acc. Chem. Res. 47:1681–90
55. Haller B, Göpfrich K, Schröter M, Janiesch J-W, Platzman I, Spatz JP. 2018. Charge-controlled mi-
crouidic formation of lipid-based single- and multicompartment systems. Lab Chip 18:2665–74
56. Härtel T, Schwille P. 2014. ESCRT-III mediated cell division in Sulfolobus acidocaldarius—a reconstitu-
tion perspective. Front. Microbiol. 5:257
57. Hartmann E, Rapoport TA, Lodish HF. 1989. Predicting the orientation of eukaryotic membrane-
spanning proteins. PNAS 86:5786–90
58. Hayashi M, Nishiyama M, Kazayama Y, Toyota T, Harada Y, Takiguchi K. 2016. Reversible morpho-
logical control of tubulin-encapsulating giant liposomes by hydrostatic pressure. Langmuir 32:3794–802
59. Heinemann F, Vogel SK, Schwille P. 2013. Lateral membrane diffusion modulated by a minimal actin
cortex. Biophys. J. 104:1465–75
60. Honda M, Takiguchi K, Ishikawa S, Hotani H. 1999. Morphogenesis of liposomes encapsulating actin
depends on the type of actin-crosslinking. J. Mol. Biol. 287:293–300
61. Horger KS, Estes DJ, Capone R, Mayer M. 2009. Films of agarose enable rapid formation of giant
liposomes in solutions of physiologic ionic strength. J. Am. Chem. Soc. 131:1810–19
62. Hotani H, Miyamoto H. 1990. Dynamic features of microtubules as visualized by dark-eld microscopy.
Adv. Biophys. 26:135–56
63. Ichihashi N, Usui K, Kazuta Y, Sunami T, Matsuura T, Yomo T. 2013. Darwinian evolution in a
translation-coupled RNA replication system within a cell-like compartment. Nat. Commun. 4:2494
64. Ierushalmi N, Malik-Garbi M, Manhart A, Abu-Shah E, Goode BL, et al. 2019. Centering and symmetry
breaking in conned contracting actomyosin networks. arXiv:1907.10642 [physics.bio-ph]
65. Ito H, Nishigami Y, Sonobe S, Ichikawa M. 2015. Wrinkling of a spherical lipid interface induced by
actomyosin cortex. Phys. Rev. E 92:062711
66. Jiménez M, Martos A, Cabré EJ, Raso A, Rivas G. 2013. Giant vesicles: a powerful tool to reconstruct
bacterial division assemblies in cell-like compartments. Environ. Microbiol. 15:3158–68
67. Jørgensen IL, Kemmer GC, Pomorski TG. 2017.Membrane protein reconstitution into giant unilamel-
lar vesicles: a review on current techniques. Eur. Biophys. J. 46:103–19
68. Kahya N, Pécheur E-I, de Boeij WP, Wiersma DA, Hoekstra D. 2001. Reconstitution of membrane
proteins into giant unilamellar vesicles via peptide-induced fusion. Biophys. J. 81:1464–74
544 Litschel •Schwille
69. Kamiya K, Kawano R, Osaki T, Akiyoshi K, Takeuchi S. 2016. Cell-sized asymmetric lipid vesicles fa-
cilitate the investigation of asymmetric membranes. Nat. Chem. 8:881–89
70. Kato A, Yanagisawa M, Sato YT, Fujiwara K, Yoshikawa K. 2012. Cell-sized connement in microspheres
accelerates the reaction of gene expression. Sci. Rep. 2:283
71. Keber FC, Loiseau E, Sanchez T, DeCamp SJ, Giomi L, et al. 2014. Topology and dynamics of active
nematic vesicles. Science 345:1135–39
72. Küchler A, Yoshimoto M, Luginbühl S, Mavelli F, Walde P. 2016. Enzymatic reactions in conned en-
vironments. Nat. Nanotechnol. 11:409–20
73. Kuruma Y, Stano P, Ueda T, Luisi PL. 2009.A synthetic biology approach to the construction of mem-
brane proteins in semi-synthetic minimal cells. Biochim. Biophys. Acta Biomembr. 1788:567–74
74. Lagny TJ, Bassereau P. 2015. Bioinspired membrane-based systems for a physical approach of cell or-
ganization and dynamics: usefulness and limitations. Interface Focus 5:20150038
75. Lee KY, Park S-J, Lee KA, Kim S-H, Kim H, et al. 2018. Photosynthetic articial organelles sustain and
control ATP-dependent reactions in a protocellular system. Nat. Biotechnol. 36:530–35
76. Limozin L, Bärmann M, Sackmann E. 2003. On the organization of self-assembled actin networks in
giant vesicles. Eur. Phys. J. E 10:319–30
77. Limozin L, Sackmann E. 2002. Polymorphism of cross-linked actin networks in giant vesicles.Phys. Rev.
Lett. 89:168103
78. Lipowsky R, Richter D, Kremer K. 1992. The structure and conformation of amphiphilic membranes:
overview. In The Structure and Conformation of Amphiphilic Membranes: Proceedings of the International
Worksh o p on Am p hi p hl i c Membrane s , Jü l ich , G er ma n y , S ep t em b er 1 6 –1 8 , 19 9 1,pp.1–6.Berlin:Springer
79. Litschel T, Kelley CF, Holz D, Koudehi MA, Vogel SK, et al. 2021. Reconstitution of contractile acto-
myosin rings in vesicles. Nat. Commun. 12: In press. https://doi.org/10.1038/s41467-021- 22422-7
80. Litschel T, Ramm B, Maas R, Heymann M, Schwille P. 2018. Beating vesicles: encapsulated protein
oscillations cause dynamic membrane deformations. Angew. Chem. Int. Ed. 57:16286–90
81. Liu AP, Fletcher DA. 2006. Actin polymerization serves as a membrane domain switch in model lipid
bilayers. Biophys. J. 91:4064–70
82. Liu AP, Fletcher DA. 2009. Biology under construction: in vitro reconstitution of cellular function. Nat.
Rev. Mol. Cell Biol. 10:644–50
83. Liu AP, Richmond DL, Maibaum L, Pronk S, Geissler PL, Fletcher DA. 2008. Membrane-induced
bundling of actin laments. Nat. Phys. 4:789–93
84. Loiseau E, Schneider JA, Keber FC, Pelzl C, Massiera G, et al. 2016. Shape remodeling and blebbing of
active cytoskeletal vesicles. Sci. Adv. 2:e1500465
85. López-Montero I, López-Navajas P, Mingorance J, Vélez M, Vicente M, Monroy F. 2013. Membrane
reconstitution of FtsZ-ZipA complex inside giant spherical vesicles made of E. coli lipids: large membrane
dilation and analysis of membrane plasticity. Biochim. Biophys. Acta Biomembr. 1828:687–98
86. Maan R, Loiseau E, Bausch AR. 2018. Adhesion of active cytoskeletal vesicles. Biophys. J. 115:2395–402
87. Maeda YT, Nakadai T, Shin J, Uryu K, Noireaux V, Libchaber A. 2012. Assembly of MreB laments
on liposome membranes: a synthetic biology approach. ACS Synth. Biol. 1:53–59
88. Majumder S, Garamella J, Wang Y-L, DeNies M, Noireaux V, Liu AP. 2017. Cell-sized mechanosensitive
and biosensing compartment programmed with DNA. Chem. Commun. 53:7349–52
89. Martino C, Kim S-H, Horsfall L, Abbaspourrad A, Rosser SJ, et al. 2012. Protein expression, aggregation,
and triggered release from polymersomes as articial cell-like structures. Angew. Chem. Int. Ed. 51:6416–
20
90. Matosevic S, Paegel BM. 2013. Layer-by-layer cell membrane assembly. Nat. Chem. 5:958–63
91. Mellouli S, Monterroso B, Vutukuri HR, te Brinke E, Chokkalingam V, et al. 2013. Self-organization of
the bacterial cell-division protein FtsZ in conned environments. Soft Matter 9:10493–500
92. Menestrina G. 1986. Ionic channels formed by Staphylococcus aureus alpha-toxin: voltage-dependent in-
hibition by divalent and trivalent cations. J. Membr. Biol. 90:177–90
93. Merkle D, Kahya N, Schwille P. 2008. Reconstitution and anchoring of cytoskeleton inside giant unil-
amellar vesicles. ChemBioChem 9:2673–81
94. Mitchison TJ. 1992. Compare and contrast actin laments and microtubules. Mol. Biol. Cell 3:1309–15
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 545
95. Miyata H, Hotani H. 1992. Morphological changes in liposomes caused by polymerization of encapsu-
lated actin and spontaneous formation of actin bundles. PNAS 89:11547–51
96. Miyata H, Nishiyama S, Akashi K, Kinosita K. 1999.Protrusive growth from giant liposomes driven by
actin polymerization. PNAS 96:2048–53
97. Miyazaki M, Chiba M, Eguchi H, Ohki T, Ishiwata S. 2015. Cell-sized spherical connement induces
the spontaneous formation of contractile actomyosin rings in vitro. Nat. Cell Biol. 17:480–89
98. Moga A, Yandrapalli N, Dimova R, Robinson T. 2019. Optimization of the inverted emulsion method
for high-yield production of biomimetic giant unilamellar vesicles. ChemBioChem 20:2674–82
99. Monnard P-A, Deamer DW. 2002. Membrane self-assembly processes: steps toward the rst cellular
life. Anat. Rec. 268:196–207
100. Montes LR, Alonso A, Goñi FM, Bagatolli LA. 2007. Giant unilamellar vesicles electroformed from
native membranes and organic lipid mixtures under physiological conditions. Biophys. J. 93:3548–54
101. Murrell M, Pontani L-L, Guevorkian K, Cuvelier D, Nassoy P, Sykes C. 2011. Spreading dynamics of
biomimetic actin cortices. Biophys. J. 100:1400–9
102. Nishimura K, Suzuki H, Toyota T, Yomo T. 2012. Size control of giant unilamellar vesicles prepared
from inverted emulsion droplets. J. Colloid Interface Sci. 376:119–25
103. Noireaux V, Libchaber A. 2004. A vesicle bioreactor as a step toward an articial cell assembly. PNAS
101:17669–74
104. Nomura SM, Tsumoto K, Hamada T, Akiyoshi K, Nakatani Y, Yoshikawa K. 2003. Gene expression
within cell-sized lipid vesicles. ChemBioChem 4:1172–75
105. Oberholzer T, Nierhaus KH, Luisi PL. 1999. Protein expression in liposomes. Biochem. Biophys. Res.
Commun. 261:238–41
106. Ohta N, Kato Y, Watanabe H, Mori H, Matsuura T. 2016. In vitro membrane protein synthesis inside
Sec translocon-reconstituted cell-sized liposomes. Sci. Rep. 6:36466
107. Okushima S, Nisisako T, Torii T, Higuchi T. 2004. Controlled production of monodisperse double
emulsions by two-step droplet breakup in microuidic devices. Langmuir 20:9905–8
108. Opathalage A, Norton MM, Juniper MPN, Langeslay B, Aghvami SA, et al. 2019. Self-organized dy-
namics and the transition to turbulence of conned active nematics. PNAS 116:4788–97
109. Osawa M, Anderson DE, Erickson HP. 2008. Reconstitution of contractile FtsZ rings in liposomes.
Science 320:792–94
110. Osawa M, Erickson HP. 2013. Liposome division by a simple bacterial division machinery. PNAS
110:11000–4
111. Ota S, Yoshizawa S, Takeuchi S. 2009. Microuidic formation of monodisperse, cell-sized, and unilamel-
lar vesicles. Angew. Chem. Int. Ed. 48:6533–37
112. Pautot S, Frisken BJ, Weitz DA. 2003. Engineering asymmetric vesicles. PNAS 100:10718–21
113. Pautot S, Frisken BJ, Weitz DA. 2003. Production of unilamellar vesicles using an inverted emulsion.
Langmuir 19:2870–79
114. Perrier DL, Vahid A, Kathavi V, Stam L, Rems L, et al. 2019. Response of an actin network in vesicles
under electric pulses. Sci. Rep. 9:8151
115. Petit J, Polenz I, Baret J-C, Herminghaus S, Bäumchen O. 2016. Vesicles-on-a-chip: a universal mi-
crouidic platform for the assembly of liposomes and polyersomes. Eur. Phys. J. E 39:59
116. Pinot M, Chesnel F, Kubiak JZ, Arnal I, Nedelec FJ, Gueroui Z. 2009. Effects of connement on the
self-organization of microtubules and motors. Curr. Biol. 19:954–60
117. Pontani L-L, van der Gucht J, Salbreux G, Heuvingh J, Joanny J-F, Sykes C. 2009. Reconstitution of an
actin cortex inside a liposome. Biophys. J. 96:192–98
118. Pott T, Bouvrais H, Méléard P. 2008. Giant unilamellar vesicle formation under physiologically relevant
conditions. Chem. Phys. Lipids 154:115–19
119. Powell K. 2018. How biologists are creating life-like cells from scratch. Nature 563:172–75
120. Prévost C, Zhao H, Manzi J, Lemichez E, Lappalainen P, et al. 2015. IRSp53 senses negative membrane
curvature and phase separates along membrane tubules. Nat. Commun. 6:8529
121. Przybylo M, Sýkora J, Humpolí ˇ
cková J, Benda A, Zan A,Hof M. 2006. Lipid diffusion in giant unilamel-
lar vesicles is more than 2 times faster than in supported phospholipid bilayers under identical conditions.
Langmuir 22:9096–99
546 Litschel •Schwille
122. Ramirez-Diaz DA, Merino-Salomon A, Heymann M, Schwille P. 2019. Bidirectional FtsZ lament
treadmilling promotes membrane constriction via torsional stress. bioRxiv 587790. https://doi.org/10.
1101/587790
123. Rampioni G, D’Angelo F, Leoni L, Stano P. 2019. Gene-expressing liposomes as synthetic cells for
molecular communication studies. Front. Bioeng. Biotechnol. 7:1
124. Reeves JP, Dowben RM. 1969. Formation and properties of thin-walled phospholipid vesicles. J. Cell
Physiol. 73:49–60
125. Richmond DL, Schmid EM, Martens S, Stachowiak JC, Liska N, Fletcher DA. 2011. Forming giant
vesicles with controlled membrane composition, asymmetry, and contents. PNAS 108:9431–36
126. Robinson T, Kuhn P, Eyer K, Dittrich PS. 2013. Microuidic trapping of giant unilamellar vesicles to
study transport through a membrane pore. Biomicrouidics 7:044105
127. Saleem M, Morlot S, Hohendahl A, Manzi J, Lenz M, Roux A. 2015. A balance between membrane
elasticity and polymerization energy sets the shape of spherical clathrin coats. Nat. Commun. 6:6249
128. Sanchez T, Chen DTN, DeCamp SJ, Heymann M, Dogic Z. 2012. Spontaneous motion in hierarchically
assembled active matter. Nature 491:431–34
129. Sato Y, Hiratsuka Y, Kawamata I, Murata S, Nomura SM. 2017. Micrometer-sized molecular robot
changes its shape in response to signal molecules. Sci. Robot. 2:eaal3735
130. Schäfer E, Kliesch T-T, Janshoff A. 2013. Mechanical properties of giant liposomes compressed between
two parallel plates: impact of articial actin shells. Langmuir 29:10463–74
131. Schäfer E, Vache M, Kliesch TT, Janshoff A. 2015. Mechanical response of adherent giant liposomes to
indentation with a conical AFM-tip. Soft Matter 11:4487–95
132. Schroer FEC, Baldauf L, Buren L, Wassenaar TA, Melo MN, et al. 2020. Charge-dependent interactions
of monomeric and lamentous actin with lipid bilayers. PNAS 117:5861–72
133. Schwille P, Spatz J, Landfester K, Bodenschatz E, Herminghaus S, et al. 2018. MaxSynBio: avenues
towards creating cells from the bottom up. Angew. Chem. Int. Ed. 57:13382–92
134. Scott A, Noga MJ, de Graaf P, Westerlaken I, Yildirim E, Danelon C. 2016. Cell-free phospholipid
biosynthesis by gene-encoded enzymes reconstituted in liposomes. PLOS ONE 11:e0163058
135. Shaklee PM, Semrau S, Malkus M, Kubick S, Dogterom M, Schmidt T. 2010. Protein incorporation in
giant lipid vesicles under physiological conditions. ChemBioChem 11:175–79
136. Shimizu Y, Inoue A, Tomari Y, Suzuki T, Yokogawa T, et al. 2001. Cell-free translation reconstituted
with puried components. Nat. Biotechnol. 19:751–55
137. Shum HC, Lee D, Yoon I, Kodger T, Weitz DA. 2008. Double emulsion template monodisperse phos-
pholipid vesicles. Langmuir 24:7651–53
138. Soares e Silva M, Alvarado J, Nguyen J, Georgoulia N, Mulder BM, Koenderink GH. 2011. Self-
organized patterns of actin laments in cell-sized connement. Soft Matter 7:10631–41
139. Soga H, Fujii S, Yomo T, Kato Y, Watanabe H, Matsuura T. 2014. In vitro membrane protein synthesis
inside cell-sized vesicles reveals the dependence of membrane protein integration on vesicle volume.
ACS Synth. Biol. 3:372–79
140. Sonnleitner A, Schütz GJ, Schmidt T. 1999. Free Brownian motion of individual lipid molecules in
biomembranes. Biophys. J. 77:2638–42
141. Stachowiak JC, Richmond DL, Li TH, Brochard-Wyart F, Fletcher DA. 2009. Inkjet formation of unil-
amellar lipid vesicles for cell-like encapsulation. Lab Chip 9:2003–9
142. Stachowiak JC, Richmond DL, Li TH, Liu AP, Parekh SH, Fletcher DA. 2008. Unilamellar vesicle
formation and encapsulation by microuidic jetting. PNAS 105:4697–702
143. Steinkühler J, De Tillieux P, Knorr RL, Lipowsky R, Dimova R. 2018. Charged giant unilamellar vesicles
prepared by electroformation exhibit nanotubes and transbilayer lipid asymmetry. Sci. Rep. 8:11838
144. Streicher P, Nassoy P, Bärmann M, Dif A, Marchi-Artzner V, et al. 2009. Integrin reconstituted in
GUVs: a biomimetic system to study initial steps of cell spreading. Biochim. Biophys. Acta Biomembr.
1788:2291–300
145. Szathmáry E, Santos M, Fernando C. 2005. Evolutionary potential and requirements for minimal pro-
tocells. In Prebiotic Chemistry,ed.PWalde,pp.167–211.Berlin:Springer
146. Szent-Györgyi AG. 2004. The early history of the biochemistry of muscle contraction. J. Gen. Physiol.
123:631–41
www.annualreviews.org •Protein Reconstitution in Giant Vesicles 547
147. Szostak JW, Bartel DP, Luisi PL. 2001. Synthesizing life. Nature 409:387–90
148. Tareste D, Shen J, Melia TJ, Rothman JE. 2008. SNAREpin/Munc18 promotes adhesion and fusion of
large vesicles to giant membranes. PNAS 105:2380–85
149. Tawk DS, Grifths AD.1998. Man-made cell-like compartments for molecular evolution. Nat. Biotech-
nol. 16:652–56
150. Teh S-Y, Khnouf R, Fan H, Lee AP. 2011. Stable, biocompatible lipid vesicle generation by solvent
extraction-based droplet microuidics. Biomicrouidics 5:044113
151. Tsai FC, Koenderink GH. 2015. Shape control of lipid bilayer membranes by conned actin bundles.
Soft Matter 11:8834–47
152. Tsai F-C, Stuhrmann B, Koenderink GH. 2011. Encapsulation of active cytoskeletal protein networks
in cell-sized liposomes. Langmuir 27:10061–71
153. Utada AS, Lorenceau E, Link DR, Kaplan PD, Stone HA, Weitz DA. 2005. Monodisperse double emul-
sions generated from a microcapillary device. Science 308:537–41
154. Virk SS, Baruah VJ, Goswami P. 2013. Giant vesicles as encapsulating matrix for stabilizing alcohol
oxidase and as container for coupled enzymatic reactions. Artif. Cells Nanomed. Biotechnol. 41:255–58
155. Visco I, Hoege C, Hyman AA, Schwille P. 2016. In vitro reconstitution of a membrane switch mechanism
for the polarity protein LGL. J. Mol. Biol. 428:4828–42
156. Walde P, Cosentino K, Engel H, Stano P. 2010. Giant vesicles: preparations and applications.
ChemBioChem 11:848–65
157. Weinberger A, Tsai F-C, Koenderink GH, Schmidt TF, Itri R, et al. 2013. Gel-assisted formation of
giant unilamellar vesicles. Biophys. J. 105:154–64
158. Weiss M, Frohnmayer JP, Benk LT, Haller B, Janiesch J-W, et al. 2018. Sequential bottom-up assembly
of mechanically stabilized synthetic cells by microuidics. Nat. Mater. 17:89–96
159. Weitz M, Kim J, Kapsner K, Winfree E, Franco E, Simmel FC. 2014. Diversity in the dynamical behavior
of a compartmentalized programmable biochemical oscillator. Nat. Chem. 6:295–302
160. Yanagisawa M, Iwamoto M, Kato A, Yoshikawa K, Oiki S. 2011. Oriented reconstitution of a membrane
protein in a giant unilamellar vesicle: experimental verication with the potassium channel KcsA. J. Am.
Chem. Soc. 133:11774–79
161. Yoshimoto M. 2011. Stabilization of enzymes through encapsulation in liposomes. In Enzyme Stabiliza-
tion and Immobilization: Methods and Protocols, ed. SD Minteer, pp. 9–18. Totowa, NJ: Humana Press
162. Yu W, Sato K, Wakabayashi M, Nakaishi T, Ko-Mitamura EP, et al. 2001. Synthesis of functional protein
in liposome. J. Biosci. Bioeng. 92:590–93
163. Zhang Y, Cheng C, Cusick B, LeDuc PR. 2007. Chemically encapsulated structural elements for probing
the mechanical responses of biologically inspired systems. Langmuir 23:8129–34
164. Zieske K, Chwastek G, Schwille P. 2016. Protein patterns and oscillations on lipid monolayers and in
microdroplets. Angew. Chem. Int. Ed. 55:13455–59
165. Zieske K, Schwille P. 2013. Reconstitution of pole-to-pole oscillations of min proteins in microengi-
neered polydimethylsiloxane compartments. Angew. Chem. Int. Ed. 52:459–62
548 Litschel •Schwille
