Download Nutrient uptake by protocells: a liposome model system

NUTRIENT UPTAKE BY PROTOCELLS: A LIPOSOME MODEL
SYSTEM
PIERRE-ALAIN MONNARD∗ and DAVID W. DEAMER
Department of Chemistry and Biochemistry, University of California, Santa Cruz, 1156 High Street,
Santa Cruz, CA 96064, U.S.A.
(∗ Author for correspondence)
(Received 2 November, 1999)
Abstract. Over the past decade, several liposome-based models for protocells have been developed.
For example, liposome systems composed of polymerase enzymes encapsulated with their substrates
have demonstrated that complex compartmentalized reactions can be carried out under conditions
in which polymeric products are protected from degradation by hydrolytic enzymes present in the
external medium. However, such systems do not have nutrient uptake mechanisms, which would
be essential for primitive cells lacking the highly evolved nutrient transport processes present in all
contemporary cells. In this report, we explore passive diffusion of solutes across lipid bilayers as one
possible uptake mechanism. We have established conditions under which ionic substrates as large as
ATP can permeate bilayers at rates capable of supplying an encapsulated template-dependent RNA
polymerase. Furthermore, while allowing the permeation of monomer substrates such as ATP, bilayer
vesicles selectively retained polymerization products as small as dimers and as large as a transfer
RNA. These observations demonstrate that passive diffusion could be used by the earliest forms of
cellular life for transport of important nutrients such as amino acids, phosphate, and phosphorylated
organic solutes.
Keywords: liposome, membrane, nucleotide triphosphates, passive diffusion, protocell
1. Introduction
The pathway by which prebiotic molecular systems undergoing chemical evolution
assembled into the earliest forms of cellular life remains an unsolved problem.
The general assumption is that, some 3.8 billion years ago, simple self-replicating
systems of polymeric molecules appeared spontaneously on the early Earth. It is
also assumed that the polymerization processes by which growth and replication
occurred required a source of chemical energy, were catalyzed in some way, and
were directed by a primitive genetic system. At some point the polymerization
reactions began to take place in membrane-bounded compartments (Deamer and
Oro, 1980; Deamer, 1997).
Compartmentalization is a necessary prerequisite for maintaining the integrity
of such interdependent molecular systems and for permitting the variations required for speciation (Tawfik and Griffiths, 1998). The components of the boundary
structures defining the earliest cellular compartments are unknown. However, the
availability of self-assembling amphiphilic molecules on the early Earth has been
Origins of Life and Evolution of the Biosphere 31: 147–155, 2001.
© 2001 Kluwer Academic Publishers. Printed in the Netherlands.
148
P.-A. MONNARD AND D. W. DEAMER
established, produced either by abiotic synthesis, (Hargreaves et al., 1977) or by
delivery of extraterrestrial organic material during late accretion (Deamer, 1985;
Deamer and Pashley, 1989). Furthermore, the fact that all contemporary living
cells are defined by lipid-bilayer membranes suggests that protocellular structures
could be modeled in the laboratory by dispersing lipid molecules as bilayer vesicles in aqueous phases (Deamer and Oro, 1980; Lazcano, 1994a, b; Luisi et al.,
1999). However, the use of lipid bilayers for this purpose has a significant drawback, which is their relative impermeability to polar or ionic molecules (Deamer
et al., 1994; Mouritsen et al., 1995). To overcome this permeability barrier, complex protein transport systems have evolved in all contemporary cells (Aidley and
Stanfield, 1996) which were presumably absent in early cells. Therefore, the latter
necessarily relied on simple transport mechanisms such as passive diffusion across
bilayer membranes to take up nutrients and sources of chemical energy from the
environment. At the same time, early membranes must have provided a selective permeability barrier that permitted permeation of monomers while retaining
polymeric products of metabolism. Otherwise newly synthesized substances would
diffuse into the surrounding bulk phase and the potential for interactive systems and
speciation would be lost (Deamer et al., 1994).
The permeability properties of bilayers have been extensively investigated for
cations, anions (Kanehisa and Tsong, 1978; Rosenquist et al., 1981; Deamer and
Bramhall, 1986; Paula et al., 1996) and small molecules (Langner and Hui, 1993).
Permeability has been shown to depend on the gel-fluid phase transition (Mouritsen
et al., 1995), the length of the lipid hydrocarbon chains (Paula et al., 1996), and
on the nature of the solutes, especially size and charge density. Solute molecules
such as water, glycerol, urea and glucose can cross the membrane barrier either
by partitioning and diffusing in the bilayers (Brunner et al., 1980)) or by using
transient defects resulting from disturbances in the lipid packing order (Paula et
al., 1996) whose frequency increases at the lipid phase transition temperature Tm
(Kanehisa and Tsong, 1978; Mouritsen et al., 1995).
What solutes might have been important to early cellular life? Several obvious
examples include phosphate, simple carbohydrates like glyceraldehyde, and perhaps amino acids. Of these, phosphate and amino acids are ionic, and bilayer membranes have been shown to represent significant permeability barriers to their free
diffusion (Chakrabarti and Deamer, 1992). In the work reported here, we will show
that under certain conditions molecules as large as nucleoside triphosphates can
permeate across bilayers composed of an intermediate-chain length lipid (DMPC).
Furthermore, the selectivity of these membranes is very high: oligonucleotides as
small as dimers are unable to permeate under the same conditions.
NUTRIENT UPTAKE BY PROTOCELLS: A LIPOSOME MODEL SYSTEM
149
2. Materials and Methods
2.1. R EAGENTS
1,2-Dimyristoyl-sn-glycero-3-phosphocholine (DMPC), and 1,2-dipalmitoyl-sn-glycero3-phosphocholine (DPPC) were purchased from Avanti Polar Lipids. ATP, adenyl(30
→ 50 )uridine 30 -monophosphate (AU), adenyl(30 → 50 )-cytidyl (30 → 50 )-cytidine
(ACC), myristic acid (MA), and sodium cholate were purchased from Sigma Chemical Co (St. Louis, MO). tRNA (from wheat germ) was obtained from Fluka (Buchs,
Switzerland). BioGel 15 A (mesh 200–400) was purchased by Bio-Rad. Other
chemicals used were of the highest grade commercially available.
2.2. L IPOSOME
PREPARATIONS
.
Pure lipids or lipid mixtures were prepared by dissolving the lipids in CHCl3
or MeOH/CHCl3 (9:1) followed by evaporation to form thin films on the interior
surface of the glass vessel. Multilamellar liposomes were formed by swelling the
film with 40 mM Tris, pH 8.0 (or pH 8.75 when MA was present), and 1.75 mM
MgCl2 to obtain a total concentration of 160 mM lipid. In permeation experiments
monitoring efflux of encapsulated solutes, the permeant solute was also present
at this step. The suspensions were dried under N2 -flow, and carefully rehydrated,
first using a dampened fabric above the film at 37 ◦ C for 20 min, followed by
deionized water addition to obtain a final lipid concentration of 160 mM. These
suspensions were then extruded ten times through 200-nm pore size polycarbonate
filters (LiposoFastTM , Avestin, Inc.).
2.3. P ERMEABILITY
MEASUREMENTS :
U PTAKE
OF EXTERNALLY ADDED
SOLUTES
Permeability measurements were usually performed at 23.3 ◦ C with 33 mM pure
DMPC, or DMPC/MA mixed liposomes, or at 32 ◦ C for those composed of DMPC/DPPC. The solutes were added from concentrated stock solutions to produce the
following final concentrations: 7 or 40 mM ATP (from a 200-mM stock solution,
pH 8), 1 mg mL−1 AU or ACC, or 11.55 mg mL−1 tRNA. Samples were taken after
incubation times up to 16 hr at specified temperatures. External solutes that were
not encapsulated were removed on a spin column (Bio-gel A 15 m, mesh 200–
400). Over 99.5% of non-entrapped nucleotides or tRNA can be removed by this
procedure (Chonn et al., 1991; Monnard et al., 1997) at 12–16 ◦ C. Encapsulation
efficiency was determined by UV absorption at 260 nm after solubilization of the
liposomes with cholate (molar ratio of cholate to surfactant was always higher than
1.8).
150
P.-A. MONNARD AND D. W. DEAMER
Figure 1. Permeability of DMPC liposomes to ATP and tRNA at phase transition temperature
(23.3 ◦ C): Uptake and release (insert). #: 7 mM ATP; : 11.55 mg mL−1 tRNA.
2.4. P ERMEABILITY
MEASUREMENTS :
E FFLUX
OF ENCAPSULATED SOLUTES
Liposome preparations containing trapped solutes such as ATP were incubated at
23.3 or 32 ◦ C (DMPC/DPPC (1:1) mixed liposomes) for up to 16 hr. Released
solute molecules were separated on a spin column at 12–16 ◦ C, and solute that
remained entrapped was spectroscopically determined at 260 nm after vesicle solubilization with cholate as described above.
3. Results
The permeability measurements of nucleoside triphosphates were conducted using
only ATP at the transition temperature of each liposome suspension (as established
in earlier studies by anisotropy measurements with diphenylhexatriene – data not
shown). Because ionic groups on solutes are by far the major factor governing
permeability, we assume that the permeation rates of all nucleotides are similar.
In Figure 1, the release (insert) and uptake of ATP in 200-nm liposome preparations are shown. After 16 hr incubation at 23.3 ◦ C, the concentration of ATP
had clearly equilibrated across the membrane with a half-time of 33 min, and 3.5%
(about 245 nmol) of the initial externally added ATP eluted with the liposomes. In
control experiments, where the free ATP was separated directly after addition of the
solute (that is, without incubation), less than 0.5% of ATP eluted with the vesicles.
In other experiments, the partition coefficient (K) of ATP in octanol was measured. The value of 8.4 × 10−5 suggests that solute molecules rarely partition into
the DMPC bilayers. These results demonstrated that measurable amounts of ATP
NUTRIENT UPTAKE BY PROTOCELLS: A LIPOSOME MODEL SYSTEM
151
TABLE I
Permeability coefficients calculated from uptake data with 7 mM ATP as solute
Liposome
composition
DMPC
DMPC
DMPC:DPPC (50:50)
DMPC:MA (95:5)
DMPC:MA (89:11)c
DMPC:MA (80:20)c
Temperature
Half-time
k × 10−4
(◦ C)
(min)
(1 s−1 )a
Permeability
coefficient
(P) × 10−9
(cm s−1 )b
23.3
37.0
32.0
23.3
23.3
23.3
33.3
53.7
36.6
10.8
4.1
18.6
3.90
0.03
0.55
4.63
35.90
5.69
1.3
0.01
0.2
1.6
12.0
1.9
a k-values were obtained by linear regression on a plot ln(([ATP]
at equilibrium –
[ATP]t )/(([ATP]at equilibrium) versus t (s).
b P values were calculated assuming a monodisperse size distribution of unilamellar
200-nm extruded liposome according to P = (k r)/3.
c At pH 8.75.
diffuse across bilayers at the phase transition temperature (Tm) by the transient
defect model (Mouritsen et al., 1995; Paula et al., 1996).
The permeability coefficients (see Table I) calculated from these data were
independent of the solute gradient (7 and 40 mM ATP) across the bilayers. As
expected, the flux and the total amount of encapsulated ATP increased as the solute
gradient increased (data not shown). Furthermore, magnesium present in the buffer
markedly inhibited ATP permeation when the ratio of Mg2+ to ATP was 1:3.5 or
greater, which indicates that ATP molecules do not permeate if they have bound
a magnesium ion. This observation is not consistent with earlier studies carried
out with mixed lipid (lecithin based) planar membranes in the presence of large
magnesium concentrations (500 mM), or small SUVs of mixed lecithin/cholesterol
(Petkau and Chelack, 1972; Stillwell and Winter, 1974). In these studies, the increase of ATP permeability was found to be related to the presence of divalent
cations in the solutions. However, these bilayers contained lipids with charged
headgroups that could have directly interacted with metal ions, thereby causing
disruption of membrane, so that the results are not directly comparable.
The permeability of DMPC bilayers at the phase transition temperature could be
only useful if it is sufficiently selective to permit monomer permeation yet prevent
the release of larger oligomers produced in the liposomes. To check this critical
aspect, the permeation rates of AU (1 mg mL−1 ), ACC (1 mg mL−1 ), and tRNA
(11.55 mg mL−1 ) were investigated. Under the same conditions used to demonstrate ATP permeation, tRNA data showed no significant uptake (see Figure 1) or
release (see insert Figure 1). Furthermore, no uptake or release could be detected
152
P.-A. MONNARD AND D. W. DEAMER
using either the dinucleotide (AU), or the trinucleotide (AAC). The absence of
released tRNA after 16 hr incubation also demonstrated that DMPC bilayers were
stable even in the presence of large amounts of encapsulated polynucleotides. It
should be noted that no hydrolysis of RNA molecules could be detected during the
incubation period.
The temperature dependence of bilayer permeability to ATP was determined at
two critical temperatures: The bilayer phase transition temperature (Tm = 23.3 ◦ C)
and the optimal temperature for typical enzyme catalyzed reactions (37 ◦ C). The
liposomes exhibited a decrease of permeability to ATP permeability exceeding 2
orders of magnitude when the temperature was increased from Tm to 37 ◦ C (see
Table I) This result is consistent with the suggestion that the passive diffusion of
ATP occurs through transient defects in the bilayer that are most numerous at Tm.
3.1. P ERMEABILITY
OF MIXED LIPID SYSTEMS
It is probable that lipid-like molecules were available only in the form of mixtures in the prebiotic environment. Furthermore, mixed lipid systems would have
markedly different permeability properties than the pure DMPC used in the experiments described above. Therefore we also tested lipid compositions in which a
single chain lipid – myristic acid (MA) – or a second phosphatidylcholine derivative (1,2-dipalmitoyl-sn-glycero-3-phosphocholine, DPPC) was added. As shown
in Table I, the permeability coefficient of mixed DMPC/DPPC (1:1) systems was
significantly decreased, presumably because of the higher Tm of the DPPC (41 ◦ C).
In the presence of MA – at molar ratios where no changes in the morphology of the
liposomes were observed – the permeability coefficient first increased as the molar
ratio of MA to DMPC increased to approximately 11 mol %, then decreased to a
value similar to that of pure DMPC liposomes.
4. Discussion
The passive diffusion of NDPs across the bilayers of membranes composed of
intermediate-chain length lipids (DMPC) or fatty acids (oleic acid/oleate) has been
already established (Chakrabarti et al., 1994; Walde et al., 1994). For instance, in
earlier work from our laboratory, the half-time for ADP entry into DMPC liposomes was approximately 2 hr. (The temperature was not maintained at the Tm
of 23.3 ◦ C, so the half-time was longer than reported here for ATP.) In oleic
acid/oleate systems, available data show that about 4% of externally added ADP
permeated after 30 hr incubation at 25 ◦ C. Again, the longer half-life reflects the
fact that the lipid was 18 carbon chain length, rather than the 14 carbon chains
of DMPC used here. The present study extends these results to NTPs bearing an
additional negative charge. We found that ATP can diffuse across bilayers of pure
DMPC or DMPC/MA mixtures, and that the rate was greatest at Tm, suggesting
NUTRIENT UPTAKE BY PROTOCELLS: A LIPOSOME MODEL SYSTEM
153
that the mechanism was transient defects in the bilayer rather solubility-diffusion
(Paula et al., 1996). It follows that transcription of DNA templates by a templatedirected enzyme, such as T7 RNA polymerase, should be possible with the substrates being supplied in the external bulk medium. These are currently underway
in our laboratory and will be reported elsewhere.
In contrast to enzyme-catalyzed reactions with limited amounts of entrapped
substrates (Oberholzer et al., 1995a, b), the substrate permeation rates described
here would permit catalyzed reactions to continue as long as the enzymes retain
catalytic activity. The possibility that continuous polymerization, replication and
transcription reactions might be achieved in liposome model systems lends plausibility to the idea that passive diffusion across lipid bilayers would allow early cells
to have access to ionic nutrients in the environment.
It remains uncertain whether the observed permeation rates could sustain the
activity of an enzyme with a relatively high turnover rate, and we are now testing this possibility. The ATP permeability coefficients (calculated here under the
assumption of monodisperse and unilamellar 200-nm liposomes) mean that 5 ATP
molecules per s entered each pure DMPC liposome at Tm. With mixed DMPC/MA
(89:11) liposomes, approximately 60 molecules should cross the bilayer each second,
a value that approaches the turnover of NTPs by a single entrapped T7 RNA
polymerase molecule.
Permeation of ionized substrates through transient defects at the Tm of pure
lipids provides a useful laboratory model, but early cells must have relied on a
more general mechanism to ensure nutrient uptake. The simplest such permeation
process also incorporates the idea of transient defects, but uses shorter chains and
lipid mixtures to increase their frequency, rather than Tm. For instance, DMPC
was used here for experimental simplicity, and it is obviously implausible the early
cells would have membranes composed of a pure lipid with a phase transition
temperature. More likely is that their membranes were composed of mixtures of
relatively short chain lipid-like amphiphiles that happened to have a sufficiently
high passive permeability to be able to support internal metabolic activity with
external nutrients.
For instance, in the experiments reported here, mixed DMPC/MA liposomes
containing a significant amount of single-chain fatty acids (11 mol %) displayed
markedly enhanced permeability toward NTPs. Therefore, membrane-forming compounds such as fatty acids with simpler chemical structures than DMPC, could be
used to increase bilayer permeability. Single chain lipids also answer two major
concerns about a protocell membrane consisting of long-chain phosphatidylcholines.
First, the longest chains reported for meteoritic monocarboxylic acids so far are
octanoic and nonanoic acid (Lawless and Yuen, 1979), and these would need to
condense with a linker species (glycerol, phosphate) to form a phosphatidyl molecule. Second, a source of phosphate groups for phospholipid synthesis is not
at all clear (Deamer et al., 1994). We are therefore undertaking investigations of
mixed fatty acid/fatty alcohol vesicles with shorter chains (8–12 carbon single
154
P.-A. MONNARD AND D. W. DEAMER
chain amphiphiles), and initial results indicate that such vesicles readily encapsulate macromolecules like nucleic acids and enzymes (C. Apel, unpublished).
An important result of the present study is related to the selectivity of the passive
diffusion. Although ATP is permeable under the conditions described here, DMPC
bilayers exhibited a surprisingly high barrier to larger solutes. For example, even
the shortest oligomer, a dimer, remained encapsulated with no measurable loss over
periods of many hours. It follows that if longer nucleotide oligomers are produced
within the aqueous compartment of a liposome, they would remain entrapped. This
is an essential prerequisite for a protocell which would then be able to gradually
increase molecular complexity in the interior volume, while releasing metabolic
end products.
The indefinite encapsulation of tRNA even at high solute concentrations (466
µM assuming a tRNA length of 75 nt) demonstrates that a lipid bilayer could have
withstood the osmotic pressure related to captured macromolecules and associated
metabolism, thereby preventing the osmotic disruption of protocells suggested by
some authors (Hartman, 1995). Furthermore, experiments, such as the entrapped
PCR reaction by Oberholzer et al. (1995a), show that lipid bilayers alone can preserve the integrity of a complex enzymatic system, even though they are exposed
to wide temperature variations. Therefore, liposomes seem to be a suitable model
for a protocell during early evolution leading to cellular life.
Acknowledgement
This work was supported by NASA grant NAG5-4665.
References
Aidley, D. J. and Stanfield, P. R.: 1996, Book Ion Channels. Molecules in Action, Cambridge
University Press, New York.
Brunner, J., Graham, D. E., Hauser, H. and Semenza, G.: 1980, Ion and Sugar Permeabilities of
Lecithin Bilayers: Comparion of Curved and Planar Bilayers, J. Membrane Biol. 57, 133–141.
Chakrabarti, A. C., Breaker, R. R., Joyce, G. F. and Deamer, D. W.: 1994, Production of RNA by
Polymerase Protein Encapsulated Within Phospholipid Versicles, J. Mol. Evol. 39, 555–559.
Chakrabarti, A. C. and Deamer, D. W.: 1992, Permeability of Lipid Bilayer to Amino Acids and
Phosphate, Biochim. Biophys. Acta 1111, 171–177.
Chonn, A., Semple, S. C. and Cullis, P. R.: 1991, Separation of Large Unilamellar Liposomes from
Blood Components by a Spin Column Procedure: Toward Identifying Plasma Proteins which
Mediate Liposomes Clearance in vivo, Biochim. Biophys. Acta 1070, 215–222.
Deamer, D. W.: 1985, Boundary Structures are Formed by Organic Components of the Murchison
Carbonaceous Chondrites, Nature 317, 792–794.
Deamer, D. W.: 1997, The First Living Systems: A Bioenergetic Perspective, Microbiol. Mol. Biol.
Rev. 61, 230–261.
Deamer, D. W. and Bramhall, J.: 1986, Permeability of Lipid Bilayers to Water and Ionic Solutes,
Chem. Phys. Lipids 40, 167–188.
NUTRIENT UPTAKE BY PROTOCELLS: A LIPOSOME MODEL SYSTEM
155
Deamer, D. W., Mahon, E. H. and Bosco, G.: 1994, Self-Assembly and Function of Primitive Membrane Structures, in S. Bengtson (ed.), Early Life on Earth, Columbia University Press, New
York, Chichester, West Sussex, Nobel Symposium No. 84, pp. 107–123.
Deamer, D. W. and Oro, J.: 1980, Role of Lipids in Prebiotic Stuctures, Biosystems 12, 167–175.
Deamer, D. W. and Pasley, R. M.: 1989, Amphiphilic Components of the Murchison Carbonaceous
Chondrite: Surface Properties and Membrane Formation, Origins of Life and Evolution of the
Biosphere 19, 21–38.
Hargreaves, W. R., Mulvihill, S. J. and Deamer, D. W.: 1977, Synthesis of Phophosplipids and
Membranes in Prebiotic Condition, Nature 266, 78–80.
Hartman, H.: 1995, Speculations on the Origin of the Genetic Code, J. Mol. Evol. 40, 541–544.
Kanehisa, M. I. and Tsong, T. Y.: 1978, Cluster Model of Lipid Phase Transition with Application to
Passive Permeation of Molecules and Structure Relaxations in Lipid Bilayers, J. Am. Chem. Soc.
100, 424–432.
Langner, M. and Hui, S. W.: 1993, Dithionite Penetration through Phospholipid Bilayers as a
Measure of Defects in Lipid Molecular Packing, Chem. Phys. Lipids 65, 23–30.
Lawless, J. G. and Yuen, G. U.: 1979, Quantitation of Monocarboxylic Acids in the Murchison
Carbonaceous Meteorite, Nature 282, 431–454.
Lazcano, A.: 1994a, The Transition from Nonliving to Living, in S. Bengtson (ed.), Early Life
on Earth, Columbia University Press, New York, Chichester, West Sussex, Nobel Symposium
No. 84, pp. 60–69.
Lazcano, A.: 1994b, The RNA World, its Precessors, and its Descendants, in S. Bengtson (ed.),
Early Life on Earth, Columbia University Press, New York, Chichester, West Sussex, Nobel
Symposium No. 84, pp. 70–80.
Luisi, P. L., Walde, P. and Oberholzer, T.: 1999, Lipid Vesicles as Possible Intermediates in the Origin
of Life, Current Opinion in Colloid and Interface Science 4, 33.
Monnard, P.-A., Oberholzer, T. and Luisi, P. L.: 1997, Encapsulation of Polynucleotides in
Liposomes, Biochim. Biophys. Acta 1329, 39–50.
Mouritsen, O. G., Jorgenseb, K. and Honger, T.: 1995, Permeability of Lipid Bilayers Near the Phase
Transition, in E. A. Disalvo and S. A. Simon (eds.), Permeability and Stability of Lipid Bilayers,
CRC Press, Boca Raton, 1, pp. 137–160.
Oberholzer, T., Albrizio, M. and Luisi, P. L.: 1995a, Polymerase Chain Reaction in Liposomes,
Chemistry and Biology 2, 677–682.
Oberholzer, T., Wick, R., Luisi, P. L. and Biebricher, C. K.: 1995b, Enzymatic RNA Replication in
Self-Reproducing Vesicles: An Approach to a Minimal Cell, Biochim. Biophys. Res. Commun.
207, 250–257.
Paula, S., Volkov, A. G., Van Hoek, A. N., Haines, T. H. and Deamer, D. W.: 1996, Permeation of
Proton, Potassium Ions, and Small Polar Molecules through Phospholipid Bilayers as a Function
of Membrane Thickness, Biophys. J. 70, 339–348.
Petkau, A. and Chelack, W. S.: 1972, Can. J. Biochem. 50, 615.
Rosenquist, K., Gabran, T. and Rydhag, L.: 1981, Studies of Permeability Across Bilayers of Lecithin,
Lipidforum, Scandinavian forum for lipid research and technology, Bergen, Norway, 1, pp. 85–
89.
Stillwell, W. and Winter, H. C.: 1974, Biochem. Biophys. Res. Commun. 56, 617.
Tawfik, D. S. and Griffiths, A. D.: 1998, Man-Made Cell-like Compartments for Molecular
Evolution, Nature Biochtechnology 16, 652–656.
Torchilin, V. P., Klibanov, A. L., Huang, L., O’Donell, S., Nossif, N. D. and Khaw, B. A.: 1992,
FASEB J. 6, 2716.
Walde, P., Goto, A., Monnard, P.-A., Wessicken, M. and Luisi, P. L.: 1994, Oparin’s Reaction Revisited: Enzymatic Synthesis of Poly(adenyl acid) in Micelles and Self-Reproducing Vesicles, J.
Am. Chem. Soc. 116, 7541–7547.