Concentration of a particular molecule (detergent or salt or substrate) in the vicinity of the protein in the proteoliposomes might affect its oreintation. Or may be the second method creates different sizes of proteoliposomes which does not allow the oreintation u need in the lipid bilayer. 

Also if the protein is not pure (contaminated with any other cellular protein) than unwanted protein might change the oreintation. 

Hello Hai,

I'm not entirely certain as to how well you can control and fine-tune the orientation of your transmembrane protein in reconstituted proteoliposomes.  However, it is definitely important for any kind of proteoliposome-based assay to be able to determine the overall proportion of molecules oriented one way or the other, as this can definitely affect the outcome of the experiment and how you interpret the results.

Am I correct in assuming that you have some kind of tag on your purified membrane protein?  If so, what is it?

I worked with a purified 12-TMS bacterial inner-membrane protein, reconstituted in proteoliposomes.  At the C-terminus of my construct, I had a GFP-His8 tag fused to the protein, separated by a TEV protease cleavage site.  In theory, the tag could be digested away after purification, and before you use the purified protein for any assays/structural work.

However, I left the tag completely intact, did my detergent solubilization, then reconstituted the protein in proteoliposomes (via detergent-removal spin columns).  To determine the approximate ratio of in-out vs out-in proteins, I digested my proteoliposomes with TEV protease +/- Triton X-100.  In the samples without TX100, only the outer surface of the proteoliposome was accessible, so only GFP-His8 tags exposed on the surface could be digested away.  In the TX100-disrupted samples, GFP-His8 tags both inside and outside the proteoliposome should have been accessible, thus representing the entire population of reconstituted protein molecules. 

After spinning down the membrane fractions in each sample (in the ultracentrifuge), all that was left was to measure the overall amount of GFP fluorescence in both supernatants.  In my case, I did this via running the samples on a SDS-PAGE gel and using a Typhoon fluorescence scanner to detect in-gel fluorescence for free GFP-His8.  Did the experiment in triplicate to get good statistics. Worked like a charm! 

We described that protocol that we developed (and to my knowledge which was not previously used) to determine protein orientation in a proteoliposome, in the following paper:

Islam et al. 2013. mBio 4:e00678-13. doi: 10.1128/mBio.00678-13.

Proton-dependent gating and proton uptake by Wzx support O-antigen-subunit antiport across the bacterial inner membrane.

So, depending on what kind of tag you have, different options are open to you.  What are you working with?

Cheers,

Salim

P.S. As an aside, there was a colleague of mine that was working with a purified membrane protein (containing 2 TMS).  He was able to purify and reconstitute versions that possessed a His6 tag, as well as one that lacked any kind of tag.  What he consistently found was that one of the constructs was always oriented in-out, while the other was always oriented the opposite way, i.e. out-in, in reconstituted proteoliposomes.  In his case, he was trying to generate 2D crystals in the membrane so orientation wasn't a problem for him.  However, the point of the story is that the type of tag (or lack thereof) can influence your overall protein orientation in the proteoliposome membrane.  He was using BioBeads.

Dear Hai,

In my own experience, you could look for the direction/orientation of proteoliposomes by some biochemical assays that are directionally/orientation specific. For example, if the substrate is only accessible through 1 side of the protein, and you have an assay to characterize this binding event then you can estimate roughly the orientation of the proteoliposomes.

Controlling the proteoliposome direction is always an interesting problem. You can try to control the liposome size by different uses of detergents, removal of detergent methods, incubation time with Biobeads (detergent removal agents)... to look at the size of the liposomes. This size of liposome can control the curvature stress on the protein and hopefully can control the orientation of the protein so that it is most thermodynamically stable. This is very protein-specific solution and it works differently for different protein type I believe.

I am thinking that you can also use different affinity tag to control the orientation of the proteoliposomes. For example, if the protein contains his-tags or flag-tags, you can use an affinity column to control the orientation of the proteoliposomes. The ones that are bound to the column will probably share similar orientation. This is just a though and I have not been able to find any article that addresses this method.

Hope this help!

Hi, Salim and Tuan,

 Thank you very much for your suggestions. My protein has a C-terminal histag. I can definitely try to use the method you suggest, to cut the histag before trion solubilization and after triton solubilization and run a gel to see the percentage of each. 

However, I also agree with Tuan's idea that some 'little' factors such as the incubation time with biobeads are important for the liposome orientation.

Last week I use a batch of protein and made liposome with 4 times of changes of biobeads (2 h rt, 2 h rt, ON 4 degree, 2 h rt) and test and find that the orientation is in-side-out, but this week I use a new batch of same protein and made liposome with less times of change (2 h rt, ON 4 degree, 2 h rt) and test and find that it is right-side-out. this makes me very confused. is it because of different batch of protein (which i donot think but will test) or because of more 'slow' of detergent removal? I think to control the detergent removal or liposome size is important for the orientation. However, to have different result is not desirable.

Do you think so?

thanks.

Hello Hai,

One potential "problem" is that biobeads are a relatively imprecise method by which to remove detergent in a specifically reproducible manner, especially if you are trying to fine-tune issues of reproducibility in your protocol.  Do you count the exact number of beads that you're transferring and replacing each time?  More/less beads than a previous transfer can affect the overall amount and/or rate at which detergent is removed from your protein and lipids.  This is how nit-pickey you may have to be (should you decide to stick with beads) to try and obtain precisely reproducible results each and every time.

Have you thought of detergent-removal resin/columns?  Granted, they're a bit more expensive, but they work like a charm each time.  Add your protein+lipid mixture to the top of the column, centrifuge it, and in your flow-through, you've got membrane proteins reconstituted in liposomes.  Works like a charm.  I successfully did it for my bacterial 12-TMS flippase, and the lab with which I was working, they use it frequently for eukaryotic CFTR.

Here's the link if you're interested in the columns:

http://www.piercenet.com/product/pierce-detergent-removal-spin-columns-plates

How exactly are you determining your orientation in your different preparations?  Are you seeing ALL of the proteins oriented one way in the first experiment, then ALL of the proteins oriented the other way in the second experiment?  There is no ratio between in-out:out-in?