You can make phospholipid vesicles with different percentages of cholesterol.

Hi Jaime Guillén thanks for your help. Can you help link with a suitable protocol .

I also suggest you to add cholesterol to the membrane. Of course the ratio of lipid to cholesterol depends on what kind of lipid you are interested in and what temperature you are looking at. I can suggest you a reference: J Phys Chem B 2017, 121 (16) 4081-4090. Hopefully you will get some idea from here.

I assume from your question that you are working with actual cells and that making the cell membranes de novo is not an option. Thus your only option (excluding extensive genetic modification of the cells) is pretty much to try to load the outer leaflet of the membrane with cholesterol or saturated, relatively long chain sphingolipids or phospholipids. (Well, easiest would be to decrease temperature, but because temperature affects pretty much every chemical and physical process, it would not tell you much about the specific role of fluidity.) Usually people use beta-cyclodextrins for cholesterol loading and gamma-cyclodextrins for phospho-/sphingolipid loading. In principle lipid exchange proteins would probably be better but with far, far higher cost.

There are a few things to keep in mind.

First, you need a good control to ensure that it is the change in membrane composition that affects the vesicle formation. Preferably you should also verify the fluidity of the cell membranes. This brings us to our second point to keep in mind. One problem with the cholesterol loading of the membranes is that cholesterol tends to uncouple the fluidity from the lipid chain order, resulting at higher mole fractions with saturated acyl chain rich membrane in fluid but ordered liquid-ordered phase. Otherwise the fluidity (i.e. how easily the lipids diffuse in the membrane) and chain order (i.e. how many kinks the chains tend to have) are more or less coupled, as kinks tend to push the lipid molecules farther apart. So, you will need preferably a few different types of reporter systems on fluidity, e.g. FRAP of a fluorescent lipid, Laurdan GP value, and, ideally some method for monitoring chain order, but this is very hard to do in cells that have very strong background fluorescence at the wavelengths of good chain order reporters like DPH-PC.

The third problem is that because of the coupling between the chain order and area occupied by chains the preferred shape of the lipid molecules shifts with temperature from the initial (for unstressed bilayers) cylindrical towards a cone (with the polar headgroup at the peak). In addition, because the cholesterol has a very small headgroup, the increasing mole fraction of cholesterol causes similar effect. Occasionally this is referred to as HII propensity, because an easy way to quantify the effect is to look at the temperature where the system undergoes a phase transition from a (bilayer) L-alpha phase to cylindrical HII phase. If this happens on one leaflet only and if there is nonuniform concentration of the components, it tends to favour invagination and budding of the vesicles. This is something that is pretty much guaranteed to affect the vesicle budding process. Firstly, by reducing the headgroup area and inducing clustering by cleaving sphingomyelin with sphingomyelinase into ceramide it is possible to induce the invagination and vesiculation of the ceramide rich domains in model membranes in the absence of any other protein than sphingomyelinase (see the papers by Holopainen and Kinnunen, Vectorial Budding etc. in Biophys. J., and by Nurminen, Holopainen, and Kinnunen in J. Am. Chem. Soc.). In addition, it affects the energetics of the fusion/fission stalk, I think you can find papers on this by Kozlov in Biophys. J.

The fourth problem is that the changes in the composition tend to affect the bending rigidity of the membrane that of course affects the energetics of the membrane bending processes like invagination and vesiculation.

It is probably easy to see that successful, significant modification of the membrane composition affects the vesicle formation, but it is so complex to try to isolate the effect of fluidity of the vesicle formation that I doubt it can be done convincingly in live cells, at the very least in would need to involve numerous modifications while using sophisticated systems to record their effects on the properties of the membranes, including microdomain formation. Most likely it would require experiments using a full, reconstituted vesicle budding machinery with giant vesicles and accompanied by coarse-grained simulations to isolate the various effects of the modifications.

So, to summarize, quick and dirty experiments with cholesterol loading can be done, but they do not tell you much about what property it was that caused the effect observed.

Juha-Matti Alakoskela thanks for your suggestions.

If you work with actual cells perhaps another option is subject to cholesterol depletion by treatment with mβCD or enrich with cholesterol by exposure to cholesterol complexed with mβCD.

See: Biophys J. 2009 Aug 19; 97(4): 1003–1012.