[Protein-crystallography] Re: help! membrane protein solubilization

Dr Engelbert Buxbaum engelbert_buxbaum at hotmail.com
Mon Jun 13 13:53:59 EST 2005


aumium wrote:

> Hi, friends.
> 
> I have a membrane protein which consists of 12 TMs. I expressed it in
> E. coli and isolated the membrane fraction. Then I tried to extract it
> from the membrane with different detergent, but just failed even with
> 1%SDS.

This could be a question of lipid/protein/detergent ratio. A detergent
micell can solubilise 10 molecules of lipid (you can assume an average
MW of 750 Da) or 1 molecule of protein (assume an average of 100 kDa).
If you have not determined the lipid/protein ratio in your membranes,
you can in first instance assume 1 mg of lipid for each mg of protein.
Using these data and the cmc and aggregation number of your detergent
you can calculate roughly the detergent concentration required for
solubilisation. Do a quick trial of 1/4, 1/2, 1, 2 and 4 times that
calculated concentration, spin 100,000 rpm 15 min (desktop
ultracentrifuge with 0.2 ml vials is ideal) and determine protein in
pellet and supernatant. Plot solubilisation vs [detergent]. Measure the
activity of your protein in an aliquot of the solubilisation mixture
(before spinning) and plot that vs [Detergent] too.

This method should be repeated with as many detergents as you can lay
your hand on, expect about 1 out of 10-20 detergents to solubilise your
protein in a functional state. Unfortunately it is impossible to predict
which one will work with your protein. In my experience some non-ionic
detergents (Triton X100 and X114, C12E8) and some positively charged
ones (CTAB) are good bets. I had less success with alkylated sugars
(dodecyl(thio)maltoside, octyl(thio)glucoside, although these are often
recommended in the literature), amphoteric detergents (CHAPS), and
negatively charged ones like SDS.

Using this method you should get a good idea of the protein/detergent
ratio required for solubilisation. However, the resulting
protein/lipid/detergent mixed micells often contain more than 1 molecule
of protein, so for purification you need to increase the detergent
concentration further. To protect the protein from striping of essential
("annular") lipids, extraneous lipid needs also to be added.

If you get this right, spinning your sample on an isopycnic sucrose- or
glycerol gradient (the Martin & Ames procedure) should result in protein
separation by molecular weight, which you can demonstrate by SDS-PAGE of
the various fractions. Once that has been achived, you can purify your
protein like a soluble one, but include the detergent at the cmc in all
buffers to prevent protein denaturation and aggregation. Note that the
cmc depends on both temperature and the concentration of salts and
osmolytes (glycerol, sucrose...) and needs to be determined for each
buffer.




More information about the Xtal-log mailing list