The Challenges of Measuring Membrane Protein Function in Giant Unilamellar Vesicles

Dolder, Nicolas (2022). The Challenges of Measuring Membrane Protein Function in Giant Unilamellar Vesicles (Unpublished). (Dissertation, Universität Bern, Philosophisch-naturwissenschaftliche Fakultät)

[img]
Preview
Text (Dissertation)
Dolder_Nicolas_PhDThesis.pdf - Other
Available under License BORIS Standard License.

Download (17MB) | Preview
[img]
Preview
Text
Dolder_Nicolas_PhDThesis_Revised.pdf - Other
Available under License BORIS Standard License.

Download (17MB) | Preview

Giant unilamellar vesicles (GUVs) are a desired membrane‐mimetic system for the study of many
membrane‐related phenomena, the function of MPs and the creation of synthetic cells. These
micrometer sized vesicles are similar in size to bacteria and eukaryotic cells, and thus mimic these
organisms more closely in terms of surface and volume. The size allows the integration of complex
systems and entire metabolic processes such as transcription and translation, as well as the investigation
of individual vesicles using light microscopy techniques, potentially cutting the costs of purified MPs
needed to perform experiments by a factor of 100 compared to bulk methods. This makes them a very
attractive system to investigate MPs, for example to test and develop novel drugs, and to create a
bottom‐up synthetic cell. However, lipids do not spontaneously assemble into cell‐sized vesicles which
has prompted the development of several different techniques for GUV formation. For the same reason,
these vesicles are more fragile towards the use of detergents, which complicates MP reconstitution. To
harness the power of light microscopy measurements, GUVs have to be immobilized to enable real‐time
observation over several minutes to hours. Lastly, the success of measuring MP function in a GUV also
depends on the choice of detection system.
In previous work in our lab, GUV electroformation on indium‐tin‐oxide (ITO) coated glass slides and
reconstitution of MPs using charge‐mediated fusion of oppositely charged vesicles was established.
GUVs were immobilized using a streptavidin‐biotin system to enable measurement of MP function. One
of the disadvantages of GUV electroformation on ITO coated glass slides is the poor compatibility with
high ionic solution, which could result in low protein activities due to formation at non‐physiological
conditions. One of the aims of this project was to establish GUV formation under physiologically relevant
conditions to allow formation in buffer compositions optimal for MP function. We thus compared
previously established electroformation on ITO coated glass slides with electroformation on platinum
(Pt) wires and the more recently developed polymer assisted swelling using PVA. We observed that both
Pt wire and PVA formation produced GUVs using various buffer compositions and that polymer assisted
swelling produced a high yield of GUVs without much optimization, showing the potential and versatility
of this method. Interestingly, we discovered that the immobilization was affected by the buffer
composition, and that strong adhesion can lead to leakage and loss of encapsulated cargo, especially in
PVA GUVs. This is an important finding as MPs are frequently followed using encapsulated fluorescent
dyes, showing that immobilization conditions have to be tuned according to the buffer composition to
provide sufficient immobilization while preventing too much cargo loss. Protons play an important role
in many cellular processes, they are involved in many transmembrane transport reactions as well as in
the production of ATP by the ATP synthase. Thus, GUVs should be able to maintain a proton gradient.
Our measurements suggest that this is indeed the case also in immobilized vesicles that have not leaked
encapsulated fluorophores. We further presented a simple strategy that could be used to estimate the
protein concentration in GUVs after charge‐mediated reconstitution by fusion of GUVs with small
vesicles containing labeled lipids and labeled MP. We could show that GUVs with more lipid‐coupled dye
signal also showed more MP‐coupled dye signal, which could simplify the quantification of MPs in GUVs
after fusion by following lipid‐coupled dye incorporation without the need for MP labeling. However,
this strategy is not able to distinguish between simple adhesion or hemifusion of small vesicles and full
fusion, which would be required for functional reconstitution of a MP. However, the same is true for
simply following labeled MP signal, meaning that potentially other methods such as content mixing
assays would be needed to get a better idea on the amount of functionally reconstituted MP.
Finally, knowledge gained from the characterization of GUVs was applied for the reconstitution and
measurement of cytochrome c oxidase from Rhodobacter sphaeroides using carboxyfluorescein and
pyranine (HPTS). The former produced only weak and unclear signals and was prone to fast bleaching.
Using ratiometric dyes such as HPTS, pH calibration can be performed, where the observed ratios should
be independent of dye concentration and bleaching. This proved to be challenging in GUVs, as the
vesicles size seemed to have an effect on the observed HPTS ratio. Despite that, cytochrome c oxidase
measurements using HPTS yielded better results and by characterization of vesicle shape and
measurement of lipid‐coupled dye signal introduced via fusion, a correlation between the proton
translocation and the relative amount of MP per vesicle could be observed, showing that thorough
characterization of the GUVs can help to relate vesicle activities. Nonetheless, the different sizes of the
vesicles were a considerable challenge for data analysis. We thus plan to use monodisperse GUVs
produced by microfluidic techniques in collaboration with members of the deMello group form the ETH
Zürich and we could show that these GUVs are fusogenic and can be potentially used to measure proton
translocation.
In a second project, we used bifunctional DNA duplexes to establish a new tool for the measurement of
MP function. By linking pH‐sensitive dyes via DNA oligomers to cholesterol moieties, fluorescent probes
can be anchored to the lipid membrane, allowing more efficient encapsulation compared to soluble dyes
which could safe costs using precious probes. Addition of DNase I allows fast and simple removal of
probes facing the liposome exterior, which are exposed to the buffered solution and thus do not
contribute to the measurement of proton translocation in or out of the vesicle, a common problem with
lipid‐coupled probes. Incorporation of bifunctional DNA into GUVs was slightly more challenging.
Depending on the GUV formation method, as well as lipid and buffer composition, differences in the
degree of incorporation into the membrane were observed, ranging from no membrane localization to
complete incorporation. Formation further seemed to be negatively affected by the probe. Thus, further
optimization of the probe might be needed to enable measurement of MP function in GUVs.

Item Type:

Thesis (Dissertation)

Division/Institute:

08 Faculty of Science > Department of Chemistry, Biochemistry and Pharmaceutical Sciences (DCBP)

Graduate School:

Graduate School for Cellular and Biomedical Sciences (GCB)

UniBE Contributor:

Dolder, Nicolas

Subjects:

500 Science > 570 Life sciences; biology
500 Science > 540 Chemistry

Language:

English

Submitter:

Christoph von Ballmoos

Date Deposited:

27 Apr 2022 14:25

Last Modified:

05 Dec 2022 16:19

BORIS DOI:

10.48350/169460

URI:

https://boris.unibe.ch/id/eprint/169460

Actions (login required)

Edit item Edit item
Provide Feedback