Biology Reference
In-Depth Information
the dimension of vesicles was regulated by extrusion, a mechanical procedure that converts
large vesicles into small ones with a desired size. After incubation, the synthesis of eGFP was
revealed fluorimetrically, demonstrating that the process of protein synthesis can occur
inside conventional lipid vesicles with diameters of 200 nm. It is noteworthy to say that the
eGFP yield was estimated as 0.01
0.05 molecules/vesicle (indicating that not all vesicles
were able to produce eGFP, this observation will be important for the discussion below).
This apparently trivial result, however, at close inspection, reveals a conundrum. The fact
that eGFP was successfully produced inside small vesicles implies that at least one copy of
the 83 different macromolecules (DNA, RNA polymerase, ribosome, etc. see
Table 14.1
)
contained in the PURE system were simultaneously encapsulated within the same lipid
vesicle in the moment of vesicle formation. It should be remarked that the concentration of
each PURE system component is between 0.1 and 1
M. It is interesting to calculate the
probability of co-entrapping 82 different molecules on the basis of the expected number of
solutes/vesicles and the Poisson distribution. For each solute, this probability is dependent
on the vesicle size (as the average number of entrapped molecules is given by
CV
, where
C
is the solute concentration in bulk and
V
the vesicle volume), and the overall probability is
calculated by considering the independent entrapment of 82 species. The final curve is
shown in
Figure 14.2A
. As intuitively expected, the overall probability of co-entrapping at
least one copy of the 82 macromolecular species (each at concentrations below between 0.1
and 1
μ
M) decreases when the vesicle volume decreases, reaching extremely low values (ca.
10
2
26
!) for vesicles with diameters of 200 nm. In order to have physically significant
probabilities, the only reasonable hypothesis is that the
local
concentration of the entrapped
solutes was not the same as the bulk values, but at least 10
μ
20 times higher. This
hypothesis, together with the observation made above (not all vesicles were capable of
synthesizing the protein) brought us to the conclusion that the encapsulation of solutes was
not following the Poisson distribution. Rather, few
vesicles were formed by
following a mechanism that is not related to the expected capture of solutes (proportional
to the vesicle volume). In these vesicles, the entrapment of solutes should follow a
mechanism that facilitates solute capture during vesicle formation, so that the
corresponding probability of entrapment results were underestimated by the Poisson
distribution. Our next question was on how to test this hypothesis.
special
'
'
269
The current available knowledge related to entrapment of solutes inside lipid vesicles has
been generally obtained via averaging techniques (batch absorbance or fluorescence, etc.),
and little attention has been devoted to studying individual vesicle encapsulation. But only
the direct measurement of the content of each vesicle can determine the true solute
occupancy distribution (i.e. the distribution of the number of entrapped solute molecules).
We therefore based our investigations on the direct visualization of the entrapped solutes
(inside liposomes) by means of cryogenic transmission electron microscopy (cryo-
TEM).
29,30
As solutes we employed: (1) ferritin (from 4 to 32
M), a protein that contains
in its core about 4 500 iron atoms, and it is therefore clearly visible in cryo-TEM images as a
black round spot; (2) ribosomes (from 0.4 to 8
μ
M), that are also electron dense enough to
be distinguishable in cryo-TEM pictures; (3)
E. coli
extracts and the PURE system, which also
contain ribosomes (1.2
μ
M in the PURE system). The main idea was to study the statistics
of entrapment of these macromolecules in liposomes by varying the solute concentration,
the vesicle preparation method, and the kind of lipids. We focused on unilamellar vesicles
with diameters between 50 to 300 nm, formed by thin-film hydration (with or without
extrusion), and by the injection method, starting from POPC, POPC/cholesterol 8/2, or
POPC/oleate 8/2 or 6/4.
μ
If the solute occupancy distribution was a Poisson one, we would expect to have a bell-
shaped curve centered around the average number
N
0
of entrapped solute, calculated as
stated above from the product of solute concentration
C
and the vesicle volume
V
