Competition between model protocells driven by an encapsulated catalyst

Article abstract of DOI:10.1038/nchem.1650

The advent of Darwinian evolution required the emergence of molecular mechanisms for the heritable variation of fitness. One model for such a system involves competing protocell populations, each consisting of a replicating genetic polymer within a replic

Full text of DOI:10.1038/nchem.1650

ARTICLES
PUBLISHED ONLINE: 19 MAY 2013 | DOI: 10.1038/NCHEM.1650
Competition between model protocells driven by
an encapsulated catalyst
1
,2
1
Katarzyna Adamala and Jack W. Szostak
*
The advent of Darwinian evolution required the emergence of molecular mechanisms for the heritable variation of fitness.
One model for such a system involves competing protocell populations, each consisting of a replicating genetic polymer
within a replicating vesicle. In this model, each genetic polymer imparts a selective advantage to its protocell by, for
example, coding for a catalyst that generates a useful metabolite. Here, we report a partial model of such nascent
evolutionary traits in a system that consists of fatty-acid vesicles containing a dipeptide catalyst, which catalyses the
formation of a second dipeptide. The newly formed dipeptide binds to vesicle membranes, which imparts enhanced affinity
for fatty acids and thus promotes vesicle growth. The catalysed dipeptide synthesis proceeds with higher efficiency in
vesicles than in free solution, which further enhances fitness. Our observations suggest that, in a replicating protocell with
an RNA genome, ribozyme-catalysed peptide synthesis might have been sufficient to initiate Darwinian evolution.
odern cells are thought to have evolved from much earlier More recently, we observed that low levels of phospholipids can
protocells, simple replicating chemical systems composed drive the competitive growth of fatty-acid vesicles in a manner
M
of a cell membrane and an encapsulated genetic polymer, that circumvents this problem by causing growth into filamentous
1
18
that were the first cellular systems capable of Darwinian evolution . structures that divide readily in response to mild shear stresses .
Evolvability may have emerged in such systems via competition This model implies that a catalyst for phospholipid synthesis, such
2,3
between protocells for a limiting resource . As protocells lacked as an acyltransferase ribozyme, imparts a large selective advantage
the complex biochemical machinery of modern cells, such compe- to its host protocell because faster growth, coupled with division,
tition was necessarily based on simple chemical or physical results in a shorter cell cycle. A model system that illustrates the
4
processes . To gain further insight into such competitive processes, potential of an encapsulated catalyst to drive vesicle growth in a
various model protocell systems were developed. Fatty-acid vesicles similar manner would, therefore, be a significant step towards rea-
5–11
were widely employed as models of early protocellular systems
because fatty acids can be generated in plausible prebiotic scenarios,
lizing a complete model of the origin of Darwinian evolution.
Here we show that the simple dipeptide catalyst seryl-histidine
and because membranes based on fatty acids have physical (Ser-His) can drive vesicle growth through the catalytic synthesis
properties that are well suited to primitive forms of life1 . The pres- of a hydrophobic dipeptide, N-acetyl-L-phenylalanine leucinamide
2,13
ence of small molecules can induce size changes and fusion of lipid (AcPheLeuNH ), which localizes to the membrane of model proto-
2
vesicles1 , and it was shown previously that a membrane-based cells and drives competitive vesicle growth in a manner similar to
4,15
1
6
reaction can drive size changes of self-assembled lipid vesicles .
that demonstrated previously for phospholipids. Ser-His catalyses
The nature of the primordial genetic material remains uncertain; the formation of peptide bonds between amino acids and between
1
9,20
competing schools of thought support either RNA or some alterna- peptide nucleic acid monomers . Although Ser-His is a very inef-
tive nucleic acid as a progenitor of RNA. Irrespective of the nature of ficient and nonspecific catalyst, we found that it generates higher
the original genetic material, an important question in considering yields of peptide product in the presence of fatty-acid vesicles. As
the origins of cellular competition is how that genetic material could a result, vesicles that contain the catalyst generate sufficient reaction
impart a selective advantage to a primitive protocell. An early product to exhibit enhanced fitness, as measured by competitive
model17 for such a scenario postulated an autocatalytic self-replicat- growth, relative to those that lack the catalyst (Fig. 1). Therefore,
ing genetic material, such as an RNA replicase, that would accumu- we can observe how a simple catalyst causes changes in the compo-
late within vesicles at a rate corresponding to its catalytic efficiency. sition of the membrane of protocell vesicles and enables the origin
Mutations that led to greater replicase activity would result in a of selection and competition between protocells.
more rapid increase in internal RNA concentration and thus
internal osmotic pressure, which would lead to faster vesicle swel- Results
ling, which in turn would drive competitive vesicle growth. This Vesicles enhance Ser-His synthesis of AcPheLeuNH . The
2
model was supported by experimental observations that osmotically catalytic efficiency of the dipeptide Ser-His is extremely low, and
swollen vesicles could grow by absorbing fatty-acid molecules from quantitation is difficult because of precipitation of the
17
19
the membranes of surrounding relaxed vesicles . Although the hydrophobic product . In addition, Ser-His appears to be a
simple physical link between mutations that lead to faster replication more-effective catalyst of the hydrolysis of the AcPheOEt ethyl
of the genetic material and the consequent osmotic swelling and ester than of peptide-bond formation, which further limits
vesicle growth is attractive, this model suffers from the lack of a product yield. We decided to characterize the Ser-His-catalysed
plausible mechanism for the division of osmotically swollen vesicles. synthesis of AcPheLeuNH2 in the presence of fatty-acid
1
Howard Hughes Medical Institute, Department of Molecular Biology, and Center for Computational and Integrative Biology, Massachusetts
2
General Hospital, Boston, Massachusetts 02114, USA, Dipartimento di Biologia, Universit a` degli Studi di Roma Tre, Rome, Italy.
*
e-mail: szostak@molbio.mgh.harvard.edu
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
2013 Macmillan Publishers Limited. All rights reserved.
1
©

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1650
ARTICLES
b
a
c
NH
O
N
H2N
HO
O
N
H
O
H
O
OH
N
H
O
N
NH2
N
NH2
H2N
O
H
O
O
ii
i
O
Figure 1 | Schematic representation of adaptive changes and competition between protocell vesicles. a, Synthesis of AcPheLeuNH by catalyst
2
encapsulated in fatty-acid vesicles. The dipeptide Ser-His catalyses the reaction between substrates LeuNH and AcPheOEt (i), which generates the product
2
of the reaction, AcPheLeuNH . The product dipeptide AcPheLeuNH localizes to the bilayer membrane (ii). b, Vesicles with AcPheLeuNH in the membrane
2
2
2
(
red) grow when mixed with vesicles without dipeptide (grey), which shrink. c, After micelle addition vesicles with AcPheLeuNH in the membrane grow
2
more than vesicles without the dipeptide.
a
50
b
No OA
4
3
2
1
0
0
0
0
0
5 mM OA
60
1
2
3
5
5 mM OA
5 mM OA
7.5 mM OA
0 mM OA
45
30
15
0
No OA, no catalyst
50 mM OA, no catalyst
0
20
40
60
80
100
0
20
40
60
80
100
Time (h)
Time (h)
c
45
d
65
45
25
5
3
2
5
5
0
10
20
30
40
50
0
10
20
30
40
50
Oleic acid vesicles (mM)
Oleic acid vesicles (mM)
Figure 2 | Ser-His catalysis in the presence of fatty-acid vesicles results in increased synthesis of the hydrophobic dipeptide product and decreased
substrate hydrolysis. a, Ser-His-catalysed synthesis of AcPheLeuNH is faster and results in a higher yield in the presence of increasing concentrations
2
of oleate vesicles (OA). b, Ser-His-catalysed hydrolysis of the substrate AcPheOEt is progressively slower in the presence of increasing concentrations of
oleate vesicles. c, Yield of dipeptide AcPheLeuNH versus concentration of oleate vesicles. d, Yield of hydrolysed substrate AcPheOH versus concentration
2
of oleate vesicles in the same reactions. Data points are the arithmetic average of two independent experiments and error bars show extreme values.
þ
For all experiments, 10 mM of each substrate, 5 mM Ser-His catalyst, 0.2 M Na -bicine, pH 8.5, 37 8C. In these experiments, the Ser-His catalyst and the
AcPheOEt and LeuNH substrates were present both inside and outside the oleate vesicles.
2
2
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
2013 Macmillan Publishers Limited. All rights reserved.
©

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1650
ARTICLES
a
2.5
b
2.0
2
equiv. micelles added
2
1
1
0
0
.0
.5
.0
.5
.0
1
.5
1
3
.0
0
5
10
10
10
15
15
15
.8 4 equiv. micelles added
0
5
10
15
Dipeptide AcPheLeuNH2 (mol%)
2
.8
2
equiv. empty vesicles + dye-labelled
vesicles with dipeptide
1
5
4
.8
0
5
1
equiv. empty vesicles + dye-labelled
vesicles with dipeptide
.5
.5
6
equiv. micelles added
1
equiv. vesicles with dipeptide +
dye-labelled vesicles with dipeptide
1
+
equiv. dye-labelled empty vesicles
vesicles with dipeptide
3.5
1
equiv. dye-labelled empty vesicles +
vesicles with dipeptide
2
.5
0
5
Dipeptide AcPheLeuNH2 (mol%) in the
observed vesicle population
Figure 3 | Competition between vesicles with and without the hydrophobic dipeptide AcPheLeuNH . a, Oleate vesicles prepared without added buffer or
2
salt and containing the indicated amount of AcPheLeuNH were mixed with 1 or 2 equiv. empty vesicles. After 15 minutes the membrane surface area was
2
measured using a FRET-based assay. FRET dyes on vesicles with AcPheLeuNH show growth (filled symbols) and FRET dyes on vesicles without
2
þ
AcPheLeuNH show shrinkage (open symbols). b, Competition for added oleate micelles in a pH-buffered high-salt environment (0.2 M Na -bicine, pH
2
8
.5). Equal amounts of vesicles with and without AcPheLeuNH were mixed, then oleate micelles were added and the surface area measured after 15
2
minutes. FRET dyes on vesicles with AcPheLeuNH show growth. FRET dyes on vesicles without AcPheLeuNH (open symbols) show less growth than that
2
2
for the corresponding peptide-containing sample (filled symbols). Error bars indicate the standard deviation. This experiment could only be carried out in the
presence of added buffer, as without buffer the alkaline oleate micelles (1 equiv. NaOH relative to fatty acid versus 0.5 equiv. per fatty acid in a vesicle)
quickly and excessively changed the pH of the mixture, which destabilized the preformed vesicles.
membranes, which we suspected would dissolve the product and the selection of adaptations that are beneficial to the protocells in
prevent precipitation. We also suspected that colocalizing the which those adaptations originated. Thus, the adaptive property
hydrophobic substrates for the synthesis of the AcPheLeuNH2 must not be shared with other protocells in a population. It was
within or on fatty-acid membranes might improve the yield and therefore important to establish, in our simplified model system,
possibly minimize substrate hydrolysis. We found that the Ser-His that the Ser-His catalyst and the dipeptide product AcPheLeuNH2
catalyst produced progressively more AcPheLeuNH2 in the remained localized within the vesicles that originally contained
presence of increasing concentrations of oleate vesicles, with the the catalyst.
conversion of substrate into product increasing from 25% in free
We first asked whether the hydrophobic peptide AcPheLeuNH2
solution to 44% in the presence of 50 mM oleate vesicles (Fig. 2a). remained localized within an initial set of vesicles or was
The presence of vesicles also diminished substrate hydrolysis from exchanged between vesicles. Our results indicate that dipeptide
65% in free solution to 10% in the presence of 50 mM oleate AcPheLeuNH2 exchanges only slowly between vesicles when
vesicles (Fig. 2b). One possible explanation for these results is that the vesicles are prepared from oleic acid plus 0.5 equiv. NaOH,
the hydrophobic substrates partition to the membrane, which with .80% remaining in the original vesicles over a period of eight
allows the reaction to occur at the solvent–lipid bilayer interface, hours (Supplementary Fig. S1a). In contrast, the presence of additional
þ
or even within the bilayer, and thereby minimize ester hydrolysis buffer (0.2 M Na -bicine, pH 8.5) led to an accelerated exchange of
and enhance product formation.
peptide between vesicles. Therefore, all subsequent experiments
Also, we studied the catalysis of AcPheLeuNH synthesis by the were carried out over timescales that were short relative to peptide
2
tripeptide Ser-His-Gly, a less-effective catalyst than Ser-His, which exchange. We performed similar experiments to show that encapsu-
produced about half as much product dipeptide in the presence of lated Ser-His and Ser-His-Gly were also retained within vesicles and
oleate vesicles (with 50 mM vesicles, the yield of AcPheLeuNH syn- did not exchange between vesicles (Supplementary Fig. S1b,c).
2
thesis from 10 mM of each substrate and with 5 mM Ser-His-Gly
þ
was 27% after 96 hours in 0.2 M Na -bicine, pH 8.5 at 37 8C, Interaction of AcPheLeuNH with membranes. As a means of
2
compared to 44% for Ser-His).
examining the interaction of AcPheLeuNH2 with the vesicle
membrane, we measured the fluorescence anisotropy of the
Slow exchange of Ser-His and AcPheLeuNH peptides between molecular probe 1,6-diphenyl-1,3,5-hexatriene as a probe of
2
vesicles. Competition between protocells is expected to result in membrane order. We found that the fluidity of oleate membranes
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
2013 Macmillan Publishers Limited. All rights reserved.
3
©

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1650
ARTICLES
a Competitive growth and corresponding shrinking
c
Competition between vesicles with different catalysts
Final/initial
Reaction
scheme
Peptide in
vesicle A vesicle B
Final/initial
surface area
Reaction scheme
surface area
S-H
S-H
None
S-H
1.24 (0.009)
1.00 (0.013)
1.14 (0.009)
1.00 (0.014)
0.74 (0.007)
1.01 (0.011)
0.85 (0.008)
1.00 (0.006)
S-H S-H-G
1.09 (0.01)
A
A
B
B
A
B
S-H-G
S-H-G
None
S-H-G
None
S-H
S-H S-H-G
0
.92 (0.005)
.16 (0.011)
S-H
S-H
A
B
S-H-G
S-H-G
None
S-H-G
1
S-H S-H-G
S-H S-H-G
S-H S-H-G
b
Competitive micelle uptake
S-H
S-H
None
S-H
4.18 (0.025)
3.41 (0.024)
4.01 (0.036)
3.43 (0.021)
1.09 (0.01)
1
. React
A
B
A
B
S-H-G
S-H-G
None
S-H-G
2
.
0.69 (0.006)
S-H
S-H
None
S-H
2.78 (0.022)
3.44 (0.017)
2.96 (0.027)
3.41 (0.026)
1
. React
A
B
A
B
2.
S-H-G
S-H-G
None
S-H-G
Figure 4 | Competition between populations of protocell vesicles. a, Vesicle size changes following 1:1 mixing of the indicated vesicle populations. b, Vesicle
size changes following competitive oleate uptake after the addition of 6 equiv. oleate micelles. Populations of vesicles contained either Ser-His (S-H), Ser-His-
Gly (S-H-G) or no catalyst, as indicated. All vesicle populations were incubated separately with amino-acid substrates for 48 hours to allow for synthesis of
the hydrophobic dipeptide product prior to mixing. c, Each sample contained two or three populations of vesicles, as indicated, in a 1:1 or 1:1:1 ratio. In each
case, the FRET dye pair was placed in one of the populations to measure the size change after the reaction. Vesicles that contained Ser-His outcompeted
vesicles that contained Ser-His-Gly. The italic values are standard deviations (s), N ¼ 8.
decreased in the presence of the dipeptide AcPheLeuNH2
We monitored vesicle growth using an assay based on fluor-
(
Supplementary Fig. S2), similar to the previously observed effects escence resonance energy transfer (FRET) for the real-time
18
21
of phospholipids on oleate membrane fluidity . The presence of measurement of membrane surface area , in which the dilution
the hydrophobic dipeptide also decreased the off-rate of fatty of membrane localized fluorescent donor and acceptor dyes
acids from fatty-acid membranes (Supplementary Fig. S9). We causes decreased FRET (see the Supplementary Information for
found that the desorption rate of oleate from fatty-acid vesicles details of the materials and methods). We found that vesicles con-
decreased with increasing concentrations of dipeptide taining AcPheLeuNH grew at the expense of those that lacked
2
AcPheLeuNH2 in the membrane, consistent with previously this dipeptide when the two were incubated together (Fig. 3a and
reported observations of the effects of phospholipid on fatty-acid Fig. 4a). Similarly, when ‘fed’ with added fatty-acid micelles, vesicles
18
vesicles . The similar effects of AcPheLeuNH and phospholipid that contained AcPheLeuNH2 grew preferentially, taking up
2
on fatty-acid desorption suggests that the dipeptide might, as for more micelles than vesicles without dipeptide (Fig. 3b and
phospholipid, also drive vesicle growth at the expense of Fig. 4b). The time course of competitive vesicle growth, and the
surrounding pure fatty-acid vesicles.
We also examined the effects of the peptide on membrane per- dipeptide, match phospholipid-driven growth reported previously,
meability; neither AcPheLeuNH2 nor reaction substrates of the which suggests similar fatty-acid exchange mechanism
Ser-His catalyst affected vesicle permeability to the small molecule (Supplementary Fig. S3).
corresponding time course of the shrinking of vesicles without the
a
calcein or to oligonucleotides (Supplementary Fig. S4).
Surprisingly, we only observed competitive growth between ves-
icles with and without peptide in self-buffered vesicles (vesicles
Competitive growth of vesicles that contain AcPheLeuNH . Given with only the carboxyl groups of the fatty acids as buffering
2
that AcPheLeuNH2 decreases membrane fluidity and fatty-acid agents). After the addition of 0.5 molar equiv. NaCl to self-buffered
dissociation in a manner similar to that observed previously for vesicles (relative to oleate), less than half as much growth was
phospholipids, we investigated whether AcPheLeuNH2 also observed, and when 1 equiv. NaCl was added to self-buffered ves-
affected
membrane-growth
dynamics.
The
dipeptide icles, no significant competitive growth was observed (Fig. 5). The
AcPheLeuNH is practically insoluble in water, so the dilution of addition of 1 equiv. tetramethylammonium chloride (TMAC)
2
the insoluble peptide fraction present in fatty-acid membrane is affected the observed surface-area change to a lesser extent (ꢀ20%
favoured entropically. That, in addition to the decreased fatty-acid inhibition) (Fig. 5), which suggests that surface ionic interactions
off-rate (monomer efflux from the membrane), could lead to the strongly affect fatty-acid exchange processes. In contrast to competi-
accumulation of fatty acids in the dipeptide-containing tive vesicle–vesicle growth, we were only able to measure competitive
þ
membrane, when fatty-acid vesicles without the dipeptide are micelle-induced growth in high-salt Na -bicine-buffered vesicle
present to provide the oleate monomer.
samples (Fig. 3b).
4
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
©
2013 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1650
ARTICLES
can also result in the formation of filamentous vesicles and
subsequent division. We found that initially spherical vesicles with
10 mol% AcPheLeuNH2 in their membranes developed into
thread-like filamentous vesicles after mixing with 100 equiv.
empty oleate vesicles (Fig. 6a,b). Gentle agitation of the
filamentous vesicles resulted in vesicle division into multiple
smaller daughter vesicles (Fig. 6c). We showed previously that the
division of filamentous vesicles occurs without significant loss of
2
1
1
0
0
.0
.5
.0
.5
.0
encapsulated content, and therefore is a plausible mechanism for
Buffer*
–
–
–
–
–
+
–
+
+
spontaneous protocell division17,23.
Salt†
A
B
C
B
C
Competitive vesicle growth induced by Ser-His-catalysed
synthesis of AcPheLeuNH . The results described above show that
2
Figure 5 | Inhibitory effect of salt and buffer on competitive growth of
vesicles. Equal amounts of oleate vesicles with and without 10 mol%
Ser-His can catalyse the formation of AcPheLeuNH in vesicles and
2
that AcPheLeuNH2 can cause or enhance vesicle growth. We
sought to combine these phenomena to effect Ser-His-driven
vesicle growth. We prepared oleate vesicles that contained
encapsulated Ser-His dipeptide and FRET dyes in their membranes.
Unencapsulated catalyst was removed either by dialysis against a
vesicle solution of the same amphiphile concentration that lacked
the catalyst or, alternatively, by purification on a Sepharose 4B size-
exclusion column. Purified vesicles that contained 5 mM Ser-His
were then mixed with 10 mM of the amino-acid substrates
AcPheLeuNH were mixed and growth (or shrinkage) was monitored by a
2
FRET-based surface-area assay. Vesicles that contained AcPheLeuNH were
2
labelled with FRET dyes and the growth monitored (black bars). Vesicles
without AcPheLeuNH were labelled with FRET dyes and shrinkage
2
monitored (white bars). The presence of either salt or buffer strongly
inhibited both growth and shrinkage after the vesicles were mixed. Error bars
þ
indicate s.e.m. (N ¼ 5). *þ , vesicles in a high-salt buffer (0.2M Na -bicine,
†
pH 8.5); 2, self-buffered vesicles (50 mol% NaOH). 2, no salt added; A,
5
0 mol% NaCl (relative to oleate); B, 100 mol% NaCl; C, 100 mol% TMAC.
AcPheOEt and LeuNH , and incubated at 37 8C for 24–48 hours
2
to allow for synthesis of the dipeptide AcPheLeuNH . High-
2
In a preliminary effort to correlate structure and activity, we performance liquid chromatography analysis of vesicles after 48
examined the effect of several different small peptides on vesicle hours of incubation showed that in vesicles with Ser-His the
growth. Although other small hydrophobic dipeptides (for dipeptide product AcPheLeuNH was synthesized with 28% yield;
2
example, Phe-Phe) led to some competitive growth, none was as parallel experiments with vesicles that contained Ser-His-Gly
efficient as AcPheLeuNH2 (Supplementary Fig. S5). N-terminal showed that the product was synthesized in 16% yield
peptide acetylation was quite important, as Phe-LeuNH was less (Supplementary Fig. S8). After incubation, samples with catalyst
2
effective than AcPheLeuNH . Log P
(calculated logarithm of were mixed with 1 equiv. oleate vesicles that had been incubated
2
calc
octanol–water partition coefficient) values suggest that the more with amino-acid substrates, but without the Ser-His catalyst. After
lipophilic peptides induce a greater competitive growth effect.
the mixing, vesicle-size changes were measured using a FRET assay
for the surface area, as described above. We found that vesicles
Competitive growth can facilitate division of protocell vesicles. It containing the peptide catalyst Ser-His already incubated with
was shown previously that the growth of multilamellar oleate substrates (to allow for the accumulation of AcPheLeuNH2)
vesicles after micelle addition22 results in the development of increased in surface area by 24% when mixed with 1 equiv. empty
fragile, thread-like structures that can fragment easily to produce oleate vesicles. When the added empty oleate vesicles were labelled
daughter vesicles. Similar filamentous growth of phospholipid- with FRET dyes, we saw a decrease in their surface area, as
containing vesicles was observed after mixing with excess pure expected. In parallel experiments, vesicles that lacked Ser-His or
17
fatty-acid vesicles . We therefore asked whether competitive were incubated without substrates did not grow after the addition
growth caused by the presence of the hydrophobic AcPheLeuNH2 of oleate vesicles (Fig. 4a). We also examined vesicles that
a
c
b
Figure 6 | Vesicle growth and division. a, Large multilamellar vesicles with 0.2 mol% Rh-DHPE (rhodamine B 1,2-dihexadecanoyl-sn-glycero-3-
phosphoethanolamine) dye and 10 mol% dipeptide AcPheLeuNH in the membrane were initially spherical. b, Ten minutes after mixing with 100 equiv.
2
unlabelled empty oleic acid vesicles without the dipeptide, thread-like filamentous structures developed. The development of filamentous vesicles from
initially spherical vesicles is caused by the more rapid increase of surface area relative to the volume increase, which is osmotically controlled by solute
permeability. To recreate this effect in the absence of additional buffer, we used sucrose, a non-ionic osmolyte, to provide an osmotic constraint on vesicle
volume. c, After gentle agitation, the filamentous vesicles fragment into small daughter vesicles. Scale bar, 10
mm.
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
5
©
2013 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1650
ARTICLES
þ
contained the less-active catalyst Ser-His-Gly; these grew, but to a vesicles, after growth induced by the addition of oleate-arg micelles
lesser extent than the vesicles that contained Ser-His (Fig. 4a).
the vesicles that contained peptide developed a transmembrane
Next we examined the ability of Ser-His (and Ser-His-Gly) to pH gradient twice as large as that seen in the vesicles without
enhance vesicle growth after AcPheLeuNH2 synthesis and then peptide (Fig. 7).
oleate micelle addition. As expected from previous experiments
in which vesicles were prepared with AcPheLeuNH directly, ves- Competition between vesicles that contain different catalysts. To
2
icles in which AcPheLeuNH2 was synthesized internally also examine competition between vesicles that harbour peptides of
showed enhanced oleate uptake from added micelles, and they varying catalytic efficiencies, we prepared self-buffered Ser-His-
therefore grew more than empty vesicles (Fig. 4b). Vesicles containing vesicles and Ser-His-Gly-containing vesicles and
prepared with Ser-His-Gly showed a similar, but smaller, incubated them separately in the presence of substrates to allow
growth enhancement.
for the synthesis of the hydrophobic dipeptide AcPheLeuNH2.
These vesicles were mixed and, after 60 minutes of incubation, the
Vesicles containing AcPheLeuNH exhibited enhanced chemical change in the surface area of vesicles labelled with dye was
2
potential generation during competitive micelle uptake. As determined from the change in the FRET signal. The vesicles that
fatty-acid vesicles grow, protonated fatty-acid molecules flip across contained Ser-His increased in surface area, but those that
the membrane to maintain equilibrium between the inner and contained Ser-His-Gly shrank. This directly demonstrates
outer leaflets. Subsequent ionization of a fraction of the competition between protocells that contain two catalysts of
protonated fatty acids that flip to the inside of the vesicle acidifies varying efficiency (Fig. 4c).
the vesicle interior, which generates a pH gradient across the
In a separate experiment, we mixed vesicles with both catalysts and
membrane. This pH gradient normally decays rapidly because of added 1 equiv. empty oleate vesicles to serve as a ‘feedstock’ for the
þ
þ
H /Na exchange, but we showed previously that by employing vesicles with the dipeptide AcPheLeuNH . In this case, both vesicles
2
a membrane-impermeable counterion, such as arginine, the pH with Ser-His and those with Ser-His-Gly increased in size, but vesicles
2
1
gradient can be maintained for many hours (t1/2 ≈ 16 hours) . with Ser-His grew almost twice as much as those with Ser-His-Gly.
þ
We therefore used arginine as a counterion (arg ) to study the This demonstrates that where ‘feedstock’ is readily available (from
effect of AcPheLeuNH on the generation of a transmembrane empty vesicles), both populations of protocell grow, but the more-
2
electrochemical potential during micelle-mediated vesicle growth.
To monitor the effect of the peptide on pH-gradient formation
efficient catalyst allows for more growth (Fig. 4c).
induced by vesicle growth, we prepared two populations of vesicles Discussion
þ
in arg -bicine buffer, one with and one without the hydrophobic The ubiquitous role of proteins as the catalysts of metabolic reac-
dipeptide AcPheLeuNH in the membrane. Consistent with the tions raises the question of the origins of protein enzymes. That
2
enhanced surface-area growth of AcPheLeuNH -containing all modern proteins are synthesized through the catalytic activity
2
2
4
of the RNA component of the large ribosomal subunit suggests
that primitive enzymes might have been peptides synthesized by
ribozymes. By extension, the earliest enzyme progenitors might
have been simple peptides synthesized in a non-coding fashion by
one or more ribozymes acting sequentially. If short, prebiotically
available peptides played useful roles in the growth or division of
early protocells, there would have been a strong selective advantage
conferred by ribozymes able to synthesize more such useful pep-
tides. We found that a short hydrophobic peptide, whether supplied
exogenously or synthesized internally, can confer a growth advan-
tage to fatty-acid vesicles. Thus, a ribozyme capable of synthesizing
short hydrophobic peptides could accelerate protocell growth, and
thereby confer a strong selective advantage. The short peptide
Ser-His confers similar effects through its catalytic synthesis of the
hydrophobic peptide product, which supports the idea that a ribo-
zyme with similar catalytic activity would confer a selective advan-
tage, but also raises the possibility that a ribozyme that makes a
catalytic peptide product could amplify its own efficacy through
the indirect synthesis of the functional end product. Similarly, a
ribozyme that synthesizes a peptide with nascent phospholipid
synthase activity would confer a strong selective advantage
8
8
7
7
.2
.0
.8
.6
With AcPheLeuNH2
Without AcPheLeuNH2
0
1
2
3
4
5
Time (s)
Figure 7 | Transmembrane pH gradient generated by growth of vesicles
during competitive micelle uptake. Equal amounts of vesicles with and
þ
18
without 10 mol% AcPheLeuNH were mixed in high-salt buffer (arg -bicine,
through phospholipid-driven growth .
2
pH ¼ 8.1), and 1 equiv. arginine-oleate micelles was added to the mixture
to initiate growth. In both experiments, the peptide-containing vesicles
that carried the pH-sensitive water-soluble dye HPTS (trisodium
The polarity of nucleic acids means that ribozymes might have
difficulty catalysing reactions between membrane-localized sub-
strates; thus, the synthesis of intermediate catalytic peptides could
8-hydroxypyrene-1,3,6-trisulfonate) encapsulated in the vesicle interior (open be an effective strategy for the synthesis of membrane-modifying
circles) and the peptide-free vesicles that contained the HPTS (filled circles)
products. The hydrophobic environment found in the interior of
vesicle bilayers could provide a favourable reaction milieu for the
were mixed together and incubated for 30 minutes. No competitive growth
þ
occurred at this stage because vesicles were in a high-salt arg -bicine buffer. catalysis of many chemical reactions, similar to that afforded (in a
þ
We then added 1 equiv. oleate-arg micelles to the mixed vesicle sample,
which triggered growth of both sets of vesicles. We measured the change in
pH inside the vesicles versus time by monitoring the change in the
fluorescence emission of the HPTS dye (see the Supplementary
much more sophisticated fashion) by the interior of folded proteins
in contemporary biochemistry. This speculation is supported by our
observation that the presence of vesicles increases the yield of
AcPheLeuNH from amino-acid substrates with Ser-His as a cata-
2
Information). Vesicles that contained AcPheLeuNH developed a larger
lyst. This increased yield most probably results from the greatly
decreased extent of AcPheOEt substrate hydrolysis when it is
2
transmembrane pH gradient as a result of greater growth.
6
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
©
2013 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1650
ARTICLES
localized to the hydrophobic membrane interior and is thereby pro- 3. Szostak, J. W., Bartel, D. P. & Luisi, P. L. Synthesizing life. Nature 409,
3
87–390 (2001).
tected from attack by water (Fig. 2), much as labile intermediates are
protected from water within the active sites of enzymes. Membrane
localization of the leucyl-carboxamide substrate may also decrease
4.
5.
6.
7.
Budin, I. & Szostak, J. W. Expanding roles for diverse physical phenomena
during the origin of life. Ann. Rev. Biophys. 39, 245–263 (2010).
Gebicki, J. M. & Hicks, M. Ufasomes are stable particles surrounded by
unsaturated fatty acid membranes. Nature 243, 232–234 (1973).
Hargreaves, W. R. & Deamer, D. W. Liposomes from ionic, single-chain
amphiphiles. Biochemistry 17, 3759–3768 (1978).
Oberholzer, T., Wick, R., Luisi, P. L. & Biebricher, C. K. Enzymatic RNA
replication in self-reproducing vesicles: an approach to a minimal cell.
Biochem. Biophys. Res. Commun. 207, 250–257 (1995).
the pK of the N-terminal amino group, and so enhance its reactivity
a
by increasing the fraction of nucleophilic deprotonated amine. The
phenomena of enhancing yields of chemical reactions by colocaliz-
ing substrates, altering pK values and limiting side reactions will
a
probably be observed for many other membrane-associated sub-
strates, which would allow protocells to start membrane-localized 8. Apel, C. L., Deamer, D. W. & Mautner, M. N. Self-assembled vesicles of
monocarboxylic acids and alcohols: conditions for stability and for the
encapsulation of biopolymers. Biochim. Biophys. Acta 1559, 1–9 (2002).
Noireaux, V. & Libchaber, A. A vesicle bioreactor as a step toward an
artificial cell assembly. Proc. Natl Acad. Sci. USA 101, 17669–17674 (2004).
metabolism with assistance from very simple peptide catalysts.
We describe here a laboratory model system designed to illustrate
the principles of competition between protocells. As the catalyst we
9.
used is not heritable, our system cannot yet evolve. Nevertheless, we 10. Stano, P. & Luisi, P. L. Achievements and open questions in the self-
have demonstrated adaptive changes that arise from encapsulated
primitive catalysts acting at the protocell level, as well as competition
between populations of protocells that contain two catalysts differ-
ing in primary structure (Ser-His and Ser-His-Gly). Direct
reproduction of vesicles and synthetic minimal cells. Chem. Commun.
46, 3639–3653 (2010).
1
1. Kurihara, K. et al. Self-reproduction of supramolecular giant vesicles
combined with the amplification of encapsulated DNA. Nature Chem. 3,
775–781 (2011).
vesicle–vesicle competitive growth is analogous to a predator–prey 12. Deamer, D. W. & Dworkin, J. P. Chemistry and physics of primitive membranes.
interaction, in which one population acquires nutrients from
another population so that the ‘predatory’ population grows and
the ‘prey’ population shrinks. The competitive micelle uptake is ana-
logous to a ‘competition for feedstock’ among two populations. It is
interesting that, in the model system we describe, these two mech-
Top. Curr. Chem. 259, 1–27 (2005).
3. Deamer, D. W. & Barchfeld, G. L. Encapsulation of macromolecules by
lipid vesicles under simulated prebiotic conditions. J. Molecul. Evol. 18,
1
203–206 (1982).
1
4. Chiruvolu, S. et al. A phase of liposomes with entangled tubular vesicles. Science
266, 1222–1225 (1994).
anisms operate under different environmental conditions: direct 15. Bard, M., Albrecht, M. R., Gupta, N., Guynn, C. J. & Stillwell, W. Geraniol
interferes with membrane functions in strains of Candida and Saccharomyces.
competitive growth occurs only under low-salt conditions, but com-
petitive micelle uptake occurs only under high-salt/buffer con-
ditions. In each case, adaptive changes occur as a result of an
encapsulated reaction, the rate of which is enhanced by an encapsu-
Lipids 23, 534–538 (1988).
1
6. Suzuki, K., Toyota, T., Takakura, K. & Sugawara, T. Sparkling morphological
changes and spontaneous movements of self-assemblies in water induced by
chemical reactions. Chem. Lett. 38, 1010–1015 (2009).
lated catalyst. Thus, this system links intraprotocell chemistry (and 17. Chen, I. A., Roberts, R. W. & Szostak, J. W. The emergence of competition
between model protocells. Science 305, 1474–1476 (2004).
catalysis) to the ability of a model protocell to adapt to selective
1
8. Budin, I. & Szostak, J. W. Physical effects underlying the transition from
primitive to modern cell membranes. Proc. Natl Acad. Sci. USA 108,
pressure. As AcPheLeuNH exchanges only slowly between vesicles,
2
the presence of this hydrophobic dipeptide is truly adaptive, in that
5249–5254 (2011).
the product produced by the encapsulated catalyst remains with its 19. Gorlero, M. et al. Ser-His catalyses the formation of peptides and PNAs.
FEBS Lett. 583, 153–156 (2009).
0. Wieczorek, R., Dorr, M., Chotera, A., Luisi, P. L. & Monnard, P. A. Formation
of RNA phosphodiester bond by histidine-containing dipeptides. ChemBioChem
original vesicle, at least for a few hours, which affords it an advan-
2
tage (here, an enhanced affinity for membrane components,
which allows growth). Competitive protocell vesicle growth can
also result, under certain circumstances, in the development of a
higher transmembrane pH gradient (Fig. 7), which potentially
could be linked to the development of useful energy sources for pro-
14, 217–223 (2013).
2
1. Chen, I. A. & Szostak, J. W. Membrane growth can generate a
transmembrane pH gradient in fatty acid vesicles. Proc. Natl Acad. Sci. USA 101,
7965–7970 (2004).
21
22. Zhu, T. F. & Szostak, J. W. Coupled growth and division of model
tocell metabolism . Furthermore, rapid competitive vesicle growth
leads to the development of thread-like filamentous vesicles, which
can subsequently divide into small daughter vesicles as a result of
gentle agitation (Fig. 6). This process was observed previously for
protocell membranes. J. Am. Chem. Soc. 131, 5705–5713 (2009).
2
3. Zhu, T. F., Adamala, K., Zhang, N. & Szostak, J. W. Photochemically driven
redox chemistry induces protocell membrane pearling and division. Proc.
Natl Acad. Sci. USA 109, 9828–9832 (2012).
22
24. Moore, P. B. & Steitz, T. A. The roles of RNA in the synthesis of protein.
oleate vesicles grown by either micelle addition or by competitive
4
Cold Spring Harb. Perspect. Biol. 3, a003780 (2011).
lipid uptake driven by membrane composition , and thus appears to
be a general route to a cycle of growth and division. Here we show
that growth can be induced by the activity of a catalyst encapsulated
Acknowledgements
J.W.S. is an Investigator of the Howard Hughes Medical Institute. This work was supported
within the protocell, which thus brings the system one step closer to in part by National Aeronautics and Space Administration Exobiology grant
NNX07AJ09G. We thank A. Engelhart, C. Hentrich, I. Budin and R. Wieczorek for
discussions and help with manuscript preparation.
an internally controlled cell cycle. If such a system exhibited her-
edity, for example, via the activity of a self-replicating ribozyme
that forms peptide bonds, it would amount to a fully functioning
protocell capable of Darwinian evolution.
Author contributions
Both authors contributed to the design of the experiments and to writing the paper.
Experiments were conducted by K.A.
Received 30 November 2012; accepted 4 April 2013;
published online 19 May 2013
Additional information
Supplementary information is available in the online version of the paper. Reprints and
permissions information is available online at www.nature.com/reprints. Correspondence and
requests for materials should be addressed to J.W.S.
References
1
.
Monnard, P. A. & Deamer, D. W. Membrane self-assembly processes: steps
toward the first cellular life. Anat. Record 268, 196–207 (2002).
Szathmary, E. & Demeter, L. Group selection of early replicators and the
origin of life. J. Theor. Biol. 128, 463–486 (1987).
2
.
Competing financial interests
The authors declare no competing financial interests.
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
2013 Macmillan Publishers Limited. All rights reserved.
7
©