The plasma membrane is the outermost layer of a cell and separates it from the extracellular environment. The membrane, as well as the proteins that are anchored in it, plays a crucial role in the uptake and efflux of molecules. Lipid bilayers are therefore an important object of research in many fields of science, such as biophysics, pharmacology and synthetic biology. In this thesis, we develop new microfluidic methods that allow us to study the laws that govern the transport processes through lipid bilayers.

We use the novel microfluidic Octanol-Assisted Liposome Assembly (OLA) technique to obtain lipid vesicles which serve as model membrane for our studies. We perform a literature review, where we discuss OLA and other techniques to obtain liposomes in detail, before we provide a biophysical analysis of liposomes generated with OLA and compare them to vesicles obtained via traditional techniques. The biophysical analysis represents the first systematic evaluation of the properties of GUVs produced with the OLA technique to date. Using a fluorescence intensity assay, we show that OLA allows for the production of GUVs with binary lipid mixtures of DOPC-DOPG and DOPC-DOPE in the 1:3, 2:2 and 3:1 lipid ratio. GUVs with binary lipid mixtures of DOPG-DOPE are only stable in the 1:3 and 2:2 mixing ratio, but not in the 3:1 ratio. We attribute this behaviour to the high charge density of this mixture and the lipid polymorphism which makes it energetically unfavourable for certain lipid compositions to form lamellar structures. We furthermore investigate the lateral lipid diffusivity of DOPC and POPC vesicles produced with OLA using fluorescence recovery after photobleaching (FRAP) and compare it to that of vesicles obtained via the established electroformation technique. We find the lateral diffusion coefficients to be quantitatively similar and in the range of 1 μm²/s for the different lipid systems and techniques tested. Finally, using a dithionite bleaching assay, we quantitatively show the unilamellarity of OLA vesicles, confirming previous results.

After examining OLA vesicles for their suitability for transport measurements, we expand the OLA technique and develop a platform that allows for the on-chip fabrication and controlled exposure of liposomes to a solute of interest. This novel microfluidic platform combines OLA with a flow through system that enables us to measure transport and other membrane-active processes that occur in time scales of tens of seconds on chip. In this platform, we either use fluorescent labels or exploit the solute molecule’s autofluorescence to visualise the transport across the vesicle membrane. Using this new method, we investigate the permeability of small antibiotic molecules of the fluoroquinolone family and quantify the transport of the drugs ciprofloxacin and norfloxacin through liposome membranes. For PGPC membranes, we measure median permeability coefficients of 3.57 × 10⁻⁶ cm/s for norfloxacin and 4.83 × 10⁻⁶ cm/s for ciprofloxacin in a PBS buffer at pH 7.4. These values correlate with the partition coefficients of the drugs, as well as previous studies on their permeation into lipid vesicles.

In a second series of experiments, we investigate whether or not DNA nanostructures can act as ion channels and increase the membrane’s permeability towards protons. For these studies, we use a microfluidic perfusion assay that was previously developed by the Keyser group. In this assay, OLA-generated vesicles are immobilised on chip via vesicle traps and are either incubated with DNA nanopores or a buffer control, before being perfused with a low pH solution. By means of the fluorescent pH indicator HPTS, encapsulated inside the GUVs, the intravesicular proton concentration is monitored throughout this exposure. Our results do not suggest substantial enhancement of proton flux as a result of the incubation with the DNA nanopore. We attribute this behaviour to the low insertion efficiency of the DNA nanopore and the temporal resolution of the assay, which might be too low when put in context with the timescale of passive proton permeation to show a potentially flux enhancing effect of the DNA nanostructure.

Finally, we widen the scope of our technique by taking the first steps towards a new visualisation method based on the deep UV absorbance of molecules rather than fluorescence. Our results indicate that this new visualisation method can in principle be incorporated into a microfluidic assay and used to measure membrane permeability, however this imaging mode features a substantially lower signal to noise ratio compared to the design based on fluorescence, limiting its capabilities. The issue of low signal to noise ratio must be overcome before the absorbance assay can provide a true alternative to fluorescence-based microfluidic assays as a means to study membrane permeation.

All in all, our studies show that GUVs obtained with the novel OLA technique are viable alternative to vesicles generated with established methods such as electroformation. We successfully use OLA-generated vesicles to study a series of biophysical membrane parameters and conduct measurements on membrane permeability. We also demonstrate the versatility of the OLA technique by successfully incorporating it into complex microfluidic assays, exemplifying its potential for the investigation of membrane properties and its use in transport studies.