Modulation of Reactivity in the Cavity of Liposomes Promotes the Formation of Peptide Bonds

Article abstract of DOI:10.1021/jacs.5b06207

In living cells, reactions take place in membrane-bound compartments, often in response to changes in the environment. Learning how the reactions are influenced by this compartmentalization will help us gain an optimal understanding of living organisms at the molecular level and, at the same time, will offer vital clues on the behavior of simple compartmentalized systems, such as prebiotic precursors of cells and cell-inspired artificial systems. In this work we show that a reactive building block (an activated amino acid derivative) trapped in the cavity of a liposome is protected against hydrolysis and reacts nearly quantitatively with another building block, which is membrane-permeable and free in solution, to form the dipeptide. By contrast, when the activated amino acid is found outside the liposome, hydrolysis is the prevalent reaction, showing that the cavity of the liposomes promotes the formation of peptide bonds. We attribute this result to the large lipid concentration in small compartments from the point of view of a membrane-impermeable molecule. Based on this result, we show how the outcome of the reaction can be predicted as a function of the size of the compartment. The implications of these results on the behavior of biomolecules in cell compartments, abiogenesis, and the design of artificial cell-inspired systems are considered.

Full text of DOI:10.1021/jacs.5b06207

Subscriber access provided by Stockholm University Library
Article
Modulation of reactivity in the cavity of liposomes
promotes the formation of peptide bonds
Anna Grochmal, Luba Prout, Robert Makin-Taylor, Rafel Prohens, and Salvador Tomas
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.5b06207 • Publication Date (Web): 10 Sep 2015
Downloaded from http://pubs.acs.org on September 13, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Modulation of reactivity in the cavity of liposomes promotes the
formation of peptide bonds.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Anna Grochmal,1 Luba Prout,1 Robert Makin-Taylor,1 Rafel Prohens2,3 and Salvador Tomas1*
1. Institute of Structural and Molecular Biology and Department of Biological Sciences, School of Science, Birkbeck
University of London, Malet Street, London WC1E 7HX, UK.
2. CIRCE Crystal Engineering, 07121 Palma de Mallorca, Spain.
3. Unitat de Polimorfisme i Calorimetria, CCiT, Universitat de Barcelona, 08028 Barcelona, Spain
ABSTRACT: In living cells reactions take place in membrane bound compartments, often in response to changes in the
environment. Learning how the reactions are influenced by this compartmentalization will help us gain an optimal un-
derstanding of living organisms at the molecular level and, at the same time, will offer vital clues on the behavior of sim-
ple compartmentalized systems, such as prebiotic precursors of cells and cell-inspired artificial systems. In this work we
show that a reactive building block (an activated amino acid derivative) trapped in the cavity of a liposome is protected
against hydrolysis and reacts nearly quantitatively with another building block, which is membrane-permeable and free in
solution, to form the dipeptide. By contrast, when found outside the liposome, the hydrolysis of the activated amino acid
is the prevalent reaction, showing that the cavity of the liposomes promotes the formation of peptide bonds. We attribute
this result to the large lipid concentration in small compartments from the point of view of a membrane-impermeable
molecule. Based on this result we show how the outcome of the reaction can be predicted as a function of the size of the
compartment. The implications of these results on the behavior of biomolecules in cell compartments, abiogenesis and
the design of artificial cell-inspired systems are considered.
In living cells, most reactions take place within com-
partments bound by a lipid membrane. Clearly, the com-
partmentalization can have various effects on the reac-
tions. For instance, compartmentalization can lead to
larger local concentration of reactants, speeding up reac-
tions, or may prevent or slow down reactions between
reactants found in different compartments. Moreover, the
lipid membrane itself can enhance the reaction between
hydrophobic molecules that dissolve preferentially within
the membrane.1,2 Some of the effects in compartmental-
ized reactivity are less obvious however, and, when look-
ing at a living cell, may remain hidden by the complex
regulatory processes typical of biomolecular systems.
These additional effects can be revealed by the study of
simpler compartmentalized system in the form of lipid
vesicles. For example, studies have shown that compart-
mentalization assists some enzymatic reactions by pre-
venting parasitic reactions that slow them down.3,4 Also,
some enzymatic reactions that lead to the formation of
multimeric enzyme assemblies are favored by compart-
mentalization.5 These are important effects that are diffi-
cult to observe and quantify when studying the cell, but
become all too apparent in simpler systems. The implica-
tions are that, whatever the relative importance that they
may have within complex biological systems, they will
dictate, to a large extent, the behavior of these simpler
systems. In other words, it is important to be able to
quantify and predict these effects in order to (i) improve
our ability to design cell-inspired materials (such as nano-
and microreactors for sensing and drug delivery,6-10 or the
bottom up design and assembly of minimal cells with
tailored functionality11-18) and (ii) gain a better under-
standing of abiogenesis,11, 19-22 a process that involved the
chemical evolution of non-living matters to cell ancestors
(termed proto-cells) and then to living cells. Clearly, the
formation of proto-cells is nature’s version of the assem-
bly of a minimal cell, and as such, understanding it
should improve our ability to create cell-like de novo sys-
tems.
Current research in the effects of compartmentaliza-
tion focusses on enzymatic reactions.4 In the context of
the assembly of nanoreactors and artificial minimal cells,
effects of compartmentalization on enzyme kinetics are
extremely important. Nevertheless, the effect of com-
partmentalization in non-enzymatic reactions may also
be important, in particular for the assembly of very simple
functional systems with truly bottom up design , i.e., in
the absence of any complex biomolecules. Also, the fate of
non-enzymatic reactions in compartmentalized systems is
particularly important in the context of abiogenesis (i.e.
the assembly of the initial proto-cells), especially if we
focus our attention on the initial steps of proto-cell evolu-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 8
tion, where enzymes as we know them would not have Here we use a model system to investigate the modu-
1
2
3
4
5
6
7
8
been present.
lation of reactivity within the cavity of lipid-bound com-
partments, in the form of lipid vesicles. The model reac-
tants are suitably modified amino acids, one of which is
membrane-impermeable (to ensure that the reaction
takes place either outside or in the cavity of the liposome)
and the other membrane-permeable. We demonstrate
that, while the lipid membrane is shown to catalyze the
condensation reaction to some extent in all cases, the
residence within the liposomal cavity of one of the build-
ing blocks leads to an almost quantitative dipeptide for-
mation. The results allow us to simulate the observed
effect with the size of the compartment and to describe a
plausible sequence of events that may have led to the as-
sembly of the first vesicle-trapped genetic oligomers.
In the current understanding of proto-cell formation it
is widely accepted that it involved the assembly of lipid
vesicles (formed of fatty acids and alcohols) that would
entrap short oligonucleotides within their cavity.23,24 The
formation of these oligonucleotide-containing vesicles is
attributed to the catalytic action of some minerals’ surfac-
es.25,26 Oligonucleotides are membrane-impermeable,
while nucleotides can permeate fatty acid membranes.
This simple fact - the differential permeability of the lipid
membrane - is enough to start a process of growth and
division that may lead to a natural selection process bene-
fiting oligonucleotide-containing vesicles over empty
ones, thus starting the long road towards cellular life.27-31
It is believed that the minimum size of RNA with ribo-
zyme polymerase functionality requires in excess of 200
oligonucleotides.32 Therefore, in the initial steps of abio-
genesis, a small trapped oligonucleotide would have been
unable to catalyze its own self-replication with enzymatic
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results and discussion
Tryptophan derivatives 1P and 2 and dipeptide 12 (Fig.
1) were synthesized using standard peptide chemistry
methods and purified by preparative HPLC (see Methods
for details). These amino acid derivatives were chosen for
convenience: they are simpler and easier to handle than,
for example, nucleotide derivatives. Also, the presence of
the pyranine moiety in 1P allows the straightforward
monitoring of reactions of peptide bond formation and
hydrolysis using optical spectroscopic methods. Their
behavior can however be extrapolated to other systems
where condensation reactions compete with hydrolysis,
such as phosphodiester bond formation leading to nucleic
acid oligomerization.
efficiency, but could still promote the synthesis of com-
plementary strands by a simple template effect.29
The
27
implications are that, in the absence of enzymatic cataly-
sis, any modulation of reactivity brought about by the
compartmentalization would have a clear impact on the
evolution of the system. For example, any role that com-
partmentalization may play on the synthesis of oligonu-
cleotides or other biopolymers, such as oligopeptides, will
play a central role in the behavior of the system.
Lipid membranes are known to catalyze condensation
reactions of amino acids,1,2 and have been seen to enhance
this type of reaction when the catalysts are simple oligo-
33
peptides.
This effect is attributed to the preferential
association of hydrophobic amino acids with the mem-
brane (leading to an increase in local concentration) cou-
pled with the fact that, on the surface of the membrane,
the concentration of water is lower than within the bulk
solution, favoring condensation reactions such as peptide
bond formation (or phosphodiester bond formation in
nucleic acids). This ability is however displayed on the
surface of the membrane and it is unclear how it may af-
fect the composition or behavior of chemical species
within the cavity of the proto-cell, unless all reactants are,
perchance, found within the proto-cell and are both
membrane-impermeable. In such a case however reac-
tions in the compartment of the proto-cell would be dic-
tated solely by their initial composition. It is clear that
evolution requires a system that is separated from the
environment by the membrane which defines the com-
partment and allows the transit of some molecules, thus
making it responsive to changes in the chemical envi-
ronment. This ability is displayed by oligonucleotide-
containing fatty acid vesicles, in which the trapped oligo-
nucleotide is impermeable yet individual nucleotides can
cross the membrane.28 In this context it becomes essential
to understand how the reactivity of cavity-confined mole-
cules versus molecules that can cross the compartment
membrane is modulated by the compartment.
Figure 1. Chemical compounds used in this work. The nam-
ing and their cartoon representation is also shown.
1P is very water-soluble and undergoes hydrolysis in
buffered aqueous solutions. The rate of hydrolysis was
determined by following changes in the UV-vis spectrum
of 1P over time readily attributed to the release of the
pyranine moiety, P, and fitting them to a first order kinet-
ics equation of the form:
ꢎ
ꢀ ꢁ ꢂꢃꢄꢅꢃꢆꢇ ꢈ ∆ꢂꢄꢅꢃꢆꢇꢉꢊꢋ
(1)
ꢌꢍꢍ
where A is the absorbance at the wavelength under
study, εP is the molar extinction coefficient of 1P at the
same wavelength, ꢀε the difference in molar extinction
coefficient between 1P and P, [1P]0 the initial concentra-
ACS Paragon Plus Environment

Page 3 of 8
Journal of the American Chemical Society
bringing the reagents together and thus enhancing the
tion of 1P and kapp the apparent rate constant of reaction.
1
2
3
4
5
6
7
8
In this experiment, kapp is the constant of hydrolysis in the
bulk solvent, khb. Analysis of the UV data showed that the
half-time of hydrolysis in the experimental conditions is
around 8 hours (Fig. 2A, Table 1). Tryptophan 1P and 2
may, in principle, react with each other to form dipeptide
12. When 2 was added to a solution of 1P in water the rate
of 1P consumption, followed by UV spectroscopy, was
similar to that seen in the absence of 2, with the half-time
of reaction being the same within the error (Fig. 2A, Table
1). This result suggests that in the presence of 2 the hy-
drolysis of 1P is still the main reaction taking place in the
experimental conditions (Fig. 2A). HPLC analysis after
complete reaction of 1P confirmed that both in the ab-
sence and presence of 2 the only detectable amino acid
derivative produced in the reaction is boc-tryptophan 1
(Supplementary Fig. 4, Table 1).
formation of 12 (Fig. 2B), a phenomenon that has been
described in both liposomal and micellar systems.1,2 The
apparent rate constant, shown in Table 1, is in these
conditions a function of the rate constants of hydrolysis
in the bulk and in the membrane (khb and khl) and that of
coupling with 2, k12:
ꢏꢐꢑꢑ ꢁ ꢒꢅꢃ∙ꢓꢏꢔꢕ ꢖ ꢒꢅꢃ∙ꢓꢏꢅꢚ ꢖ ꢗ1 ꢈ ꢒꢅꢃ∙ꢓꢘꢏꢔꢙ
(3)
9
where k12 is a pseudo-first order kinetic constant that
depends on K2•L and the concentration of 2. We can use
this constant given the fact that 2 is always in large excess
in relation to 1P. The value of k12 can be derived by using
the known values of khb and khl (see Supplementary Table
2)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1P is an amphiphilic molecule that can insert into lipid
membranes using the side-chain of the tryptophan moiety
as a lipid anchor. We evaluated the binding affinity of 1P
for lipid membranes of vesicles composed of EYPC by
means of a UV titration method. The dissociation con-
stant derived from the titration data, K1P•L, is 0.7 mM (see
Supplementary Information and Supplementary Fig. 1 for
details). In the presence of EYPC liposomes, the rate of
hydrolysis of 1P is somewhat reduced, as shown by the
analysis of the changes in the UV–vis spectrum (Fig. 2B,
Table 1). This result is consistent with the fact that a part
of 1P is anchored in the lipid membrane where it is pro-
tected from hydrolysis to some extent. The apparent rate
constant is therefore a combination of the rate constant
on the bulk, khb, and the rate constant when anchored in
the lipid membrane, khl:
ꢏꢐꢑꢑ ꢁ ꢒꢅꢃ∙ꢓꢏꢔꢕ ꢖ ꢗ1 ꢈ ꢒꢅꢃ∙ꢓꢘꢏꢔꢙ
(2)
Where x1P•L is the fraction of 1P bound to the mem-
brane which, in presence of an excess of lipids, may be
assumed to remain constant during the course of the re-
action (see Supplementary Information for details)
The value displayed in Table 1 for this reaction is kapp
obtained from the fitting of the data. This value is con-
sistent (within the error) with the value obtained by using
the values of khb (calculated in the experiment in the ab-
sence of liposomes) and khl (calculated in the experiment
with confined 1P, see below). Tryptophan derivative 2 is
also an amphiphilic molecule able to insert into lipid
membranes. The dissociation constant, derived from fluo-
rescence spectroscopy titration methods, for membranes
of EYPC vesicles, K2•L, is 1.3 mM (see Supplementary In-
formation and Supplementary Fig. 1 for details). When 2
was added to 1P in the presence of EYPC liposomes, a
small but clear acceleration of the rate of 1P reaction was
observed (Fig. 2B, Table 1). HPLC analysis of the products
showed the presence of the product of hydrolysis 1 to-
gether with 12, in approximately a 7 to 3 ratio (Table 1,
Supplementary Fig. 4). In the experimental conditions, a
fraction of both 1P and 2 are found associated with the
liposomal membrane. Clearly, in the experimental condi-
tions, the lipid membrane acts as a simple catalyzer,
Figure 2. A. Changes in relative concentration of 1P over
time derived from the UV data in the absence (grey circles)
and presence (black circles) of 2. The grey and black lines
represent the best fit to a single exponential decay. The panel
to the right shows a cartoon representation of the products
of reaction in the presence of 2. B. Idem in the presence of
vesicles. C. Idem for samples with 1P confined in the cavity of
liposomes. The red circles and trace correspond to a sample
with twice the concentration of 2. In all experiments the con-
centration of 1P was 30 ꢁM, that of lipids 500 ꢁM and that of
2 either 300 ꢁM or 600 ꢁM. (see Supplementary Information
and Supplementary Table 1 for further sample details).
ACS Paragon Plus Environment

Journal of the American Chemical Society
Table 1. Kinetic parameters and outcome of 1P reactions.
Page 4 of 8
1
2
Without 2
With 2
3
4
5
6
7
8
No vesicles
1P outside 1P
inside No vesicles
1P outside 1P
inside 1P
inside
vesicles
0.065
11
vesicles
vesicles
0.087
8
vesicles
vesiclesc
k (h-1) a
0.084
8.2
0.0034
200
0.089
7.8
0.11
6.2
23
0.14
4.8
18
t1/2 (h) a
% of 1 b
% of 12 b
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
>99
nd
>99
>99
>99
nd
68
nd
nd
32
77
82
a The error in k and t1/2 is on the order of 20%. b The percentage of 1 and 12 were estimated at the end point of the
reaction (after 48 hours) by HPLC. The error is on the order of 5%. See Supplementary Information for the experi-
mental details. c Used twice the concentration of 2. See Supplementary Table 1 for sample details.
1P does not permeate through the EYPC vesicular
membrane in a timescale of days (see Supplementary In-
formation and Supplementary Fig. 2). Vesicles loaded
with 1P were produced by suspending the solid lipids in a
buffer containing dissolved 1P. Most non-trapped 1P was
removed by gel permeation. Fluorescence quenching ex-
periments and analysis of the UV spectra showed that
around 85% of unreacted 1P (75% of the total 1P used to
prepare the experiment) was found inside the vesicles at
the beginning of the experiment (see Supplementary In-
formation and Supplementary Fig. 2). The average con-
centration of 1P inside the liposomes was estimated from
the concentration of lipids, the average area per EYPC
molecule and the average size of liposomes (see Supple-
mentary Information for details). The concentration was
estimated to be around 14.5 mM, which is 4.8 times larger
than the initial concentration of 1P in the buffer used to
prepare the vesicles. This increase in concentration is
attributed to the amphiphilic nature of 1P and its inability
to cross the membrane, once the vesicle is formed. Thus,
upon formation of the vesicles, a fraction of 1P incorpo-
rates into the membrane. The fraction that ends up in the
inner leaflet remains trapped, leading to an increase in
the concentration of 1P in the cavity of the liposome in
relation to the original concentration of 1P in pure buffer.
For samples containing liposome-confined 1P (but no 2),
changes in the UV spectra did not fit well to a single ex-
ponential function (typical of a simple first order kinetic
reaction) but were instead fitted to a double exponential
of the form:
tails), it follows that kappc is the rate of hydrolysis of 1P
associated with the membrane, khl. On the other hand,
the apparent rate constant for the fraction found outside
the liposomes, kapp, is a function of the hydrolysis of the
membrane-bound and free 1P (see equation (2)). In the
fitting of the data we use the value of kapp obtained from
the experiment with 1P outside the liposome as a known
parameter and kappc (that is, khl) is obtained as a parameter
of the fitting. For this experiment, the value reported in
Table 1 is kappc
some-confined 1P is very small and leads to a reaction
half-time of 200 hours (Table 1).
.
The rate constant of hydrolysis of lipo-
For samples with liposome-confined 1P, the addition of
2 led to an acceleration of the rate of between 30- to 40-
fold (depending on the concentration of 2 added) with an
observed half-time of 5 to 6 hours approximately (Fig. 2C,
Table 1, Supplementary Table 2). In these experiments
there were also two separated populations of 1P, where
the apparent rate constant for confined 1P is
ꢏꢐꢑꢑꢛ ꢁ ꢏꢔꢕ ꢖ ꢏꢟꢠ
(5)
The UV data however fits reasonably well to a single
exponential. This fact is attributed to a similar rate of re-
action of both populations. Clearly, whatever reduction in
the rate of hydrolysis afforded by the membrane for lipo-
some-confined 1P was offset by reaction of 1P with 2 in-
side the liposome, making kappc and kapp too similar to be
able to differentiate them. Analysis by HPLC of the prod-
ucts confirmed that 12 was the main product of reaction
in these conditions, at around 80% of 12 against 20% of
product of hydrolysis 1 (Table 1 Supplementary Fig. 4).
Since 25% of total 1P used was found outside the lipo-
somes and the hydrolysis outside the vesicles is around 25
times faster than inside the vesicles, we attribute the
presence of the product of hydrolysis to the hydrolysis of
1P outside the cavity of the liposomes, which is the main
reaction when found outside the liposome. Conversely,
liposome-confined 1P reacted almost quantitatively with 2
to produce 12 in the experimental conditions. The near
quantitative formation of 12 inside the liposome is readily
attributed to the combination of (i) the protection against
hydrolysis of 1P afforded by the lipid membrane at the
ꢀ ꢁ ꢂꢃꢄꢅꢃꢆꢇ ꢖ ꢒꢅꢃꢛꢜꢈ∆ꢂꢄꢅꢃꢆꢇꢉꢊꢋ
ꢎꢞ ꢖ ꢗ1 ꢈ
ꢌꢍꢍꢝ
ꢒꢅꢃꢛꢘꢜꢈ∆ꢂꢄꢅꢃꢆꢇꢉꢊꢋ ꢎꢞ
(4)
ꢌꢍꢍ
The need of a double exponential to fit the data is con-
sistent with the presence of two separated populations of
1P reacting at different rates. The fraction confined in the
vesicles is x1Pc, (estimated to be around 85% of the total
unreacted 1P at the beginning of the experiment), with an
apparent rate of reaction of the confined fraction, kappc
Since in excess of 99% of confined 1P is associated with
the membrane (see supplementary Information for de-
.
ACS Paragon Plus Environment

Page 5 of 8
Journal of the American Chemical Society
high local concentration inside the vesicle and (ii) the
ability of the membrane to bring both reagents, 1P and 2,
together.
1
2
3
4
5
6
7
8
9
It is clear that, in the presence of liposomes, the yield of
dipeptide 12 depends, amongst other factors, on the ex-
tent to which 1P and 2 are bound to the membrane of the
liposomes. On the one hand, the association of 1P with
the membrane decreases the rate of hydrolysis. On the
other hand, the association of both 1P and 2 brings the
reagents together, favoring the formation of 12. In the
presence of an excess of 2, the yield of 12, expressed as the
mole fraction x12, can be written as
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢒꢅꢚ ꢁ ꢒꢅꢃ∙ꢓꢒꢅꢅꢃꢚ∙ꢓ
(6)
where x1P•L is the fraction of 1P associated with the mem-
brane and x121P•L the fraction of 1P associated to the mem-
brane that undergoes reaction with 2. x1P•L depends,
among other factors, on the concentration of lipids. For
molecules confined in the cavity of the liposome the con-
centration of lipids depends on the radius r of the lipo-
some. x1P•L can therefore be written as a function of the
radius of the liposome as follows:
ꢛ
ꢒꢅꢃ∙ꢓ ꢁ ꢡꢢ
(7)
ꢤꢛ
ꢅꢃ∙ꢣ
where c is a confinement factor that depends on the area
per lipid molecule (see Supplementary Information for
details). On the other hand, x12
can be written as
1P•L
ꢋ ꢦ
ꢥ
ꢚ∙ꢣ
ꢅꢚ
ꢒ
ꢁ
(8)
ꢅꢃ∙ꢓ
ꢋ
ꢤꢋ ꢦ
ꢥ ꢚ∙ꢣ
ꢧꢨ
Figure 3. A. Graphical representation of the typical sizes of
choice of membrane-bound compartments. GUV, LUV and
SUV stand for giant, large and small unilamellar vesicle re-
spectively. Drug carriers and nanoreactors are typically in the
size range of LUV and SUV. B. Yield of the product of cou-
pling as a function of the size of the compartment and the
dissociation constant from the membrane of the component
confined (1P). The simulation has been generated using
equation (9), with values of ks = 1500 M-1s-1, [L] = 0.5 mM, [2]
= 0.3 mM and K2•L = 1.3 mM.
where ks is the intrinsic second order rate constant for the
coupling reaction, while the in-membrane concentration
expressed as mole fraction of membrane-associated 2,
x2•L, can be calculated from the values of total concentra-
tion of 2 and lipid, and the affinity constant K2•L
.
Combining equations (6)-(8) we have:
ꢛ
ꢒꢅꢚ ꢁ ꢡꢢ ꢤꢛ ꢩ ꢋ
ꢅꢃ∙ꢣ
ꢋ ꢦ
ꢥ ꢚ∙ꢣ
(9)
ꢤꢋ ꢦ
ꢥ ꢚ∙ꢣ
ꢧꢨ
These experiments show that the confinement of a re-
active molecule inside a liposome has a strong impact on
the reactions it undergoes. Reactions of hydrolysis may be
drastically slowed down, to a much larger extent than
when found in the presence of lipid vesicles, but not con-
fined within the aqueous cavity.33 This effect is readily
attributed to the larger concentration of lipids that a
membrane-impermeable molecule experiences when con-
fined in a small liposome cavity. For example, using non
encapsulated reagents, it can be seen that the suppression
of hydrolysis is enhanced with increasing lipid concentra-
tion (Supplementary Fig. 5 and Supplementary Table 3).
Since the concentration of lipids in the cavity is inversely
proportional to the radius of the lipid vesicle, it is clear
that the stabilization against hydrolysis reactions increas-
es as the compartment shrinks. A reactive molecule in the
cavity remains therefore available to react with a mem-
brane-permeable molecule to yield dimers. The overall
effect can be approximately emulated in non-
which allows one to estimate the yield of dipeptide 12 as a
function of all the relevant parameters – in particular, as a
function of the size of the compartment and the affinity
of the confined building block for the membrane (Fig. 3).
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 8
encapsulated systems, using large concentrations of lipid. the dissociation of the monomer sitting in the outer leaf-
1
2
3
4
5
6
7
8
(Supplementary Fig. 6). In general, equation (9) allows
for the quantification of the ability of a lipid-bound com-
partment to inhibit hydrolysis and promote oligomeriza-
tion reactions. While enzymatic reactions may not be
directly affected by the compartmentalization factor de-
scribed in equation (9), they may be affected indirectly.
For example, depending on the size of the compartment,
an enzyme will interact to a larger extent with the mem-
brane, which may modify its catalytic ability. The same
reasoning applies to the availability of the substrates. It is
therefore conceivable that biomolecular processes are
regulated, to some extent, by the size of the compartment
in which they take place. It is however in the absence of
let of the vesicles leading to its hydrolysis. By contrast,
the fraction located in the inner leaflet would be protect-
ed against hydrolysis. The other monomers present in the
medium (for example, membrane-permeable mono-
phosphate nucleotides) would associate with the mem-
brane and readily react with the trapped starter, especial-
ly if they feature activated versions of the phosphate
group (Fig. 4).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
enzymes where the presence of lipids has a clear and
33
measurable effect on the outcome of the reactions.
.
Thus equation (9) can assist in the design of more effi-
cient nanoreactors with applications as drug carriers, for
example, were the correct choice of reagents may lead to
a better control of the release of a confined drug.
It is in abiogenesis where the implications of these find-
ings are the clearest. Abiogenesis is a phenomenon that,
among other chemical changes,34-38 requires the oligomer-
ization of simple building blocks into more complex,
functional macromolecules. Any process that assists oli-
gomerization and inhibits hydrolysis is therefore an im-
portant candidate to have played a role in abiogene-
sis.25,26,33,39 The candidacy becomes even clearer when the
promotion of oligomerization and inhibition of hydrolysis
takes place in such a way that yields readily made com-
partments, or proto-cells, enriched with the oligomeric
molecules trapped inside. In the context of abiogenesis,
the phenomenon described here will lead to the accumu-
lation of oligomers in the cavity in the presence of (i) a
starter monomer that is an amphiphile as well as mem-
brane-impermeable and (ii) the presence of additional
membrane-permeable, amphiphilic building blocks. In
our experiment, the starter monomer is 1P, which has
been chosen for convenience: the pyranine moiety pre-
vents membrane permeation and is a convenient molecu-
lar probe; tryptophan is suitably hydrophobic. In prebi-
otic settings, this role could be fulfilled by a nucleotide
derivative bearing an oligophosphate chain as a mem-
brane-impermeable moiety. The rest of the components
may include any suitable sugar (including ribose) and a
nitrogen base. Interestingly, it has been shown that deca-
noic acid, a candidate for prebiotic membrane lipids, in-
teracts with the bases and sugar found in RNA.40 A plau-
sible starting point towards a compartment enriched in
polymers would begin with the accumulation of simple
lipids (such as fatty acids and fatty alcohols) together
with the starter polyphosphate nucleotide. There are a
number of natural processes that could bring these com-
ponents together, including the formation of eutectic ice
phases upon water freezing, or simply by evaporation.
Upon re-dilution (by, for example, melting of the ice),
vesicles would spontaneously form, incorporating the
amphiphilic membrane-impermeable starter monomer.
Further dilution (as the ice fully melts) would result in
Figure 4. A plausible scenario for the initial stages of proto-
cell evolution. A mixture of lipids and an amphiphilic nucleo-
tide polyphosphate is hydrated (I). Following formation of
the vesicles, the nucleotide-polyphosphate associated with
the outer membrane is diluted leading to dissociation from
the membrane (II). In the presence of membrane permeable
nucleotides, the formation of dinucleotides is catalyzed by
the inner leaflet of the membrane (III). Some dinucleotide
may be lost by permeation (IV). Slow permeation of larger
molecules may however enable the extension of the nucleo-
tide in the inner compartment leading to increasingly less
permeable oligomers that would remain trapped (V).
The reaction may result in the loss of the membrane-
impermeable oligophosphate moiety. Nevertheless, larger
oligomers are likely to be less membrane-permeable and,
as they extend, the likelihood of remaining in the cavity
for an extended period increases, allowing for further
growth of the polymer located in the cavity. The scenario
depicted is a plausible sequence of events that would lead
to the formation of primordial oligonucleotides from very
simple building blocks. A similar mechanism can be de-
picted for the formation of other functional oligomeric
molecules, such as oligopeptides, a possibility that is con-
sistent with a scenario where peptides and oligonucleo-
tides evolve independently in the first instance, being
only integrated in the same metabolic network at a later
stage.
ACS Paragon Plus Environment

Page 7 of 8
Journal of the American Chemical Society
(8) Baumler, H.; Georgieva, R. Biomacromolecules 2010, 11,
1480.
(9) Palivan, C. G.; Fischer-Onaca, O.; Delcea, M.; Itel, F.; Mei-
er, W. Chem. Soc. Rev. 2012, 41, 2800.
(10) Tanner, P.; Balasubramanian, V.; Palivan, C. G. Nano Lett.
2013, 13, 2875.
Concluding remarks
1
2
3
4
5
6
7
8
9
In summary, the phenomenon described here is an
emerging property of chemical systems41,42 featuring small
membrane-bound compartments. In addition to the clear
implications that it has for abiogenesis (or rather, because
of these implications) this phenomenon can also inform
the development of synthetic biology approaches that
relate to the bottom-up design of artificial, cell inspired,
devices. These devices may have applications ranging
from chemical synthesis (as an alternative methodology
for the synthesis of peptides or oligonucleotides) to bio-
sensing and as drug delivery vehicles. The plausibility of
these approaches, together with further implications of
these results in prebiotic research, is currently being ex-
plored in our laboratories.
(11) Szostak, J. W.; Bartel, D. P.; Luisi, P. L. Nature 2001, 409,
387.
(12) Simpson, M. L. Mol. Syst. Biol. 2006, 2, 69.
(13) Forster, A. C.; Church, G. M. Genome Res. 2007, 17, 1.
(14) Doktycz, M. J.; Simpson, M. L. Mol. Syst. Biol. 2007, 3, 125.
(15) Zhang, Y.; Ruder, W. C.; LeDuc P. R. Trends Biotechnol.
2008, 26, 14.
(16) Ichihashi, N.; Matsuura, T.; Kita, H.; Sunami, T.; Suzuki,
H.; Yomo, T. Cold Spring Harb. Perspect. Biol. 2010, 2, a004945.
(17) Forlin, M.; Lentini, R.; Mansy, S. S. Curr. Opin. Chem. Biol.
2012, 16, 586.
(18) Stano, P.; Luisi, P. L. Curr. Opin. Biotechnol. 2013, 24, 633.
(19) Schrum, J. P.; Zhu, T. F.; Szostak, J. W. Cold Spring Harb.
Perspect. Biol. 2010, 2, a002212.
(20) Walde, P. Bioessays 2010, 4, 296.
(21) Ruiz-Mirazo, K.; Briones, C.; de la Escosura, A. Chem. Rev.
2014, 114, 285.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information. Experimental methods ( includ-
ing synthesis and characterization of 1P, 2 and 12, sample
preparation, titration experiments, kinetic experiments and
HPLC characterization of the products of reaction), and de-
tailed derivation of the equations. This material is available
free of charge via the Internet at http://pubs.acs.org.
(22) Mann, S. Angew. Chem. Int. Ed. 2013, 52, 155.
(23) Segre, D.; Ben-Eli, D.; Deamer, D. W.; Lancet, D. Origins
Life Evol. B. 2001, 31, 119.
(24) Hanczyc, M. M.; Fujiyama, S. M; Szostak, J. W. Science
2003, 302, 618.
AUTHOR INFORMATION
(25) Hazen, R. M.; Sverjensky, D. A. Cold Spring Harb. Per-
spect. Biol. 2010, 2, a002162.
Corresponding Author
(26) Ferris, J. P.; Ertem, G. Science 1992, 257, 1387.
(27) Adamala, K.; Szostak, J. W. Science 2013, 342, 1098.
(28) Zhu, T. F.; Szostak, J. W. J. Am. Chem. Soc. 2009, 131, 5705.
(29) Mansy, S. S.; Schrum, J. P.; Krishnamurthy, M.; Tobe, S.;
Treco, D. A.; Szostak, J. W. Nature 2008, 454, 122.
(30) Hanczyc, M. M.; Szostak, J. W. Curr. Op. Chem. Bi-
ol.2004, 8, 660.
(31) Chen, I. A.; Roberts, R. W.; Szostak, J. W. Science 2004,
305, 1474.
(32) Attwater, J.; Wochner, A.; Holliger, P. Nature Chem. 2013,
5, 1011.
(33) Adamala, K.; Szostak, J.W. Nature Chem. 2013, 5, 495.
(34) Powner, M. W.; Gerland, B.; Sutherland, J. D. Nature
2009, 459, 239.
(35) Sutherland, J. D. Cold Spring Harb. Perspect. Biol. 2010, 2,
a005439.
(36) Ritson, D.; Sutherland, J. D. Nature Chem. 2012, 4, 895.
(37) Hein, J. E.; Blackmond, D. G. Acc. Chem. Res. 2012, 45,
2045.
(38) Danger, G.; Plasson, R.; Pascal, R. Chem. Soc. Rev. 2012, 41,
5416.
* Dr. Salvador Tomas, s.tomas@bbk.ac.uk
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version
of the manuscript.)
Funding Sources
This work was funded by the Faculty of Sciences and the
Department of Biological Sciences, Birkbeck, University of
London and by CIRCE Crystal Engineering.
ACKNOWLEDGMENT
The authors would like to thank H. Saibil and N. H. Williams
for their critical reading of the manuscript.
ABBREVIATIONS
EYPC, egg yolk phosphatidylcholine.
(39) Attwater, J.; Wochner, A.; Pinheiro, V. B.; Coulson, A.;
Hollinger, P. Nat. Commun. 2010, 1, 76.
(40) Black, R. A.; Blosser, M. C.; Stottrup, B. L.; Tavakley, R.;
Deamer, D. W.; Keller, S. L. Proc. Nat. Acad. Sci. U. S. A. 2013,
110, 13272.
REFERENCES
(1) Rasmussen, S.; Chen, L. H.; Nilsson, M.; Abe, S. Artif. Life
2003, 9, 269.
(2) Blocher, M.; Liu, D.; Walde, P.; Luisi, P. L. Macromolecules
1999, 32, 7332.
(3) Bansho, Y.; Ichihashi, N.; Kazuta, Y.; Matsuura, T.; Suzuki,
H.; Yomo, T. Chem. Biol. 2012, 19, 478.
(4) Ichihashi, N.; Yomo, T. Curr. Op. Chem. Biol. 2014, 22, 12.
(5) Matsuura, T.; Hosoda, K.; Kazuta, Y.; Ichihashi, N.; Suzuki,
H.; Yomo, T. ACS Synth. Biol. 2012, 1, 431.
(6) Nardin, C.; Thoeni, S.; Widmer,a J.; Winterhalterb, M.;
Meier, W. Chem. Commun. 2000, 1433.
(7) Broz, P.; Driamov, S.; Ziegler, J.; Ben-Haim, N.; Marsch, S.;
Meier, W.; Hunziker, P. Nano Lett. 2006, 6, 2349.
(41) Ludlow, R. F.; Otto, S. Chem. Soc. Rev. 2008, 37, 101.
(42) Lehn, J. M. Angew Chem. Int. Ed. 2013, 52, 2836.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment