Not sure if possible with your project, but you could recombinantly express these membrane protein complexes for isolation purposes to get better results. Additionally, here is a review that may be of some use

http://onlinelibrary.wiley.com/doi/10.1002/0471140864.ps0408s53/pdf

Aside from the regular list of detergents, nanodiscs and isotropic bicelles work to preserve the activities of membrane complexes. You will first have to isolate the protein in detergent though (limited to DHPC for bicelles, nanodiscs are more flexible) and their size and composition interferes with some assays. Non-detergent sulfobetaines (NDSBs) and amphiphols have reported to work for some people. Our experience in my former lab was positive with bicelles and negative for NDSBs and amphiphols for the particular systems we were working with.

http://onlinelibrary.wiley.com/doi/10.1002/cbic.200300830/pdf

Another option might be SMALP's although the membrane complex would have to be relatively small as they have a ~10nM disc diameter. They have the advantage of being able to remove your protein(s) of choice directly from the membrane and encapsulating it in the lipids which directly surround it. This can aid in stability and activity. There is a nice review of nanodiscs, detergents and SMALPs here;

http://www.ncbi.nlm.nih.gov/pubmed/21599653

Surfactant-free purification of membrane proteins with intact native membrane environment. Jamshad M, et al., Biochemical Society Transactions (2011) 39, (813–818).

Hope it helps.

I used DDM 0.01%,it works with E.coli complex I and complex II

Be sure to isolate the membranes first and the make the exchange.

1% DDM (dodecyl maltoside) for lysis

it is a very soft detergent. I prefer 0,03% for purification because the cmc is very closed at 0,01% and you could loss your protein easyly during steps

If you have membrane targeted expression, it is better to use mild detergents like DDM of OG. In most cases these detergents will save protein structure and allow to perform purification and fuctional tests.

There is no absolute answer to that question :( Choice of a detergent is really protein dependent. It could be difficlut to find a detergent that solubilized cell membrane and simultaneously preserve complex. Mild detergent with low CMC are often used in the membrane protein field. For example, the Photosystem II complex was isolated with dodecylmaltoside from membrane of thermophile cyanobacteria. Nervertheless some other large membrane protein complex were isolated without any loss of acitvity with high CMC detergent like octylglucoside. It was the case of the Light harvesting complex from the membrane of Rhodospirullum Rubrum. The best will be to try to find a detergent that can be used for solubilization and activity measurement. Dodecylmaltoside and TX100 could be good candidates. However, if you find a detergent that make possible solubilization and complex isolation possible but without any activity of your complex, you can try to switch by adding a Size Exclusion Chromatography step to exchange detergent 1 for a detergent 2. In that case, think about the possible role of lipids too that can have a critical role on maintaining structure/function. SEC is the only techniques that allow efficient exchange of low CMC detergent.

Some microbes contain very unusual lipids, e.g. those in Thermoplasma Volcanium so that overpurification removing all of these could be harmful. It could be hamful in any case and adding lipids back/reconstitution of complexes often does not work.

Glycosylation and other modifications of the membranes/proteins - could that rather be the cause of the problems?

I agree with M Yegres and as also seen in literature, DDM seems to be compatible with a majority of transmembrane proteins. Switching detergents is very common, some membranes are really tough to disrupt at all so that one has to store the protein in a milder detergent.

In some cases the GraFix method of slight crosslinking is useful. http://www.nature.com/nmeth/journal/v5/n1/full/nmeth1139.html

Good luck

If you are interested in detergents as opposed to more complex partners like SMALPs and amphipols, then there are almost too many to suggest and to try out!

You have probably tried many of these, but the standard classes that are least destabilising are non-ionic detergents, e.g. of the maltoside, glucoside and Cymal families. A more recent development has been neopentyl glycol based amphiphiles, as described initially in this paper http://www.ncbi.nlm.nih.gov/pubmed/21037590 (available to purchase from Anatrace). There have been a handful of reports that these are more stabilising than many other detergents, e.g. for stabilisation of labile GPCRs.

Whether you need to solubilise in one detergent and then exchange to another really depends on the endpoint that you intend for the protein complex. For example, in terms of crystallisation, solubilisation is often performed in a longer chain detergent, often DDM, and then switched over to a shorter chain detergent, e.g. DM, to provide smaller micelles and hence promote crystal contact formation. Other pitfalls may be if a particular detergent interferes with an assay, etc.

One detergent worth mentioning is digitonin. This has the downside that it has been impossible to crystallise a single membrane protein in it, and it is significantly more expensive than most other detergents. However, it is very mild and often preserves membrane protein complexes where other detergents have failed.

all excellent suggestions, but the old adage "horses for courses" is certainly worth bearing in mind, which almost forces you into a great deal of trial and error I'm afraid to say,

Dodecyl maltoside remains the most popular detergent for isolation of native membrane proteins. I would try a range of detergents on an isolated envelope preparation and look at the % solubilization in small test sample, spun at 100K. Other detergents to try are DDAO and C12E8, with SDS as the 100% denaturing control- all dodecyl chains, only head groups differ. Octylglucoside or C8E5 have higher cmcs and may be more potent. Lots of more exotic detergents to test. We usually do solubilization test with about 12 different detergents at 1-2% concentration, 1 h on ice.

To solubilize membrane protein COMPLEXES or to solubilize INDIVIDUAL integral membrane proteins mean two different goals and two different strategies. They need in general two different detergents. So if you have solved the problem of solubilization of membrane protein complexes in functional stage you very possibly have to change detergent for solubilization the already solubilized protein complexes. However, I do not think that there is a more definite answer to your question as this. In my experience, each integral membrane protein I have ever solubilized needed another detergent than the previous one. Moreover, even the very same TYPE of protein needed different detergent for solubilization from plant or from animal plazma membranes - which points also to the differences in all other membrane constituents in the membrane than your protein in question. The best strategy is always to test 6-10 different detergents before choosing the one for large-scale solubilization.

I highly recommend reading a comment written above by Cédric Montigny. There is nothing to add may be except one thing: to exchange the detergents can be tricky, but there are the ways to do that.

You can try Triton ,or digitonin ,I have tried them & they were excellent.

Empirical ‘‘stability screens’’ can be helpful in choosing which detergents, and which detergent concentrations, may

be optimal for a given system.

It is worth remembering which detergents worked best for what and when. Some day some sense will dawn of much of this empirical hit and miss approaches. Digitonin came to fame with rhodopsin and paid huge dividends in controlled solubilization/ disruption to achieve sub-mitochondrial particles. The conventional wisdom to keep things in an ice bucket while solubilizing took a severe jolt when mitochondrial ATPase died when kept in cold. Do also heating a protein to do SDS-PAGE. Both rhodopsin and coupling factor B in mitochondria (D.R.Sanadi told me of the latter) tend to aggregate badly when heated, almost impossible to reverse. It is always good to run an unheated SDS gel at the outset. DDM paid good dividends with cytochrome oxidase if I am not mistaken and Octyl glucoside because of its dialysability became a signature candidate among detergents. I wonder if any one remembers Ammonyx LO, which made debut as a gentle solubilizing agent for adenyl cyclase in late 70's. A word of caution on what activity means. If proton pumping is an activity for cytochrome oxidase, neutral detergents retain stimulation by uncouplers AFTER complete solubilization, while ionic detergents differentially affect this.

I worked with the isolation of membrane proteins, and found that most detergents (SDS, Triton, Digitonin, etc.) work well in maintaining activity of membrane bound enzymes as long as one reduces the % of the detergent while testing for enzymatic activity. My suggestion is that it is better to dissolve the detergent in a buffer suitable for testing some specific enzymatic activity, until one maximizes the activity. I would recommend SDS (Sodium Dodecyl Sulfate) as a detergent for these purposes because SDS is better at dissociating the phospholipid-membrane protein bonds than most other detergents. Should you detect some enzymatic activity but at a lower rate than desired, it is most likely due to an essential or specific phospholipid interaction that the enzyme needs which might be lost with the detergent action. Should you do a lipid analysis of the sample and are able to pretty much identify the phospholipid involved, (the one showing up loose in the lipid analysis in greater concentration), you can always gradually add that phospholipid as a reagent until you see the enzymatic activity increase. Most likely the added phospholipid as a reagent will result in phospholipid micelles forming and incorporating the enzymes, and if the micelles are large enough you should be able to see them upon microscopic inspection of a drop of the sample. You might have to experiment reducing the % detergent in the solution gradually at 0.5% increments, and always making sure to maintain the temperature of the solution suitable for whatever enzymatic activity you are targeting.

I recommend to use digitonin

Hello John,

As others I would recommend your to make a screen, each protein is unique, and one cannot tell in advance what is the best detergent. There are a couple things I would like to add. First, not only detergent concentration, but also the ratio protein:detergent could be important. We are working with DDM, and we have found that there is rather sharp optimum of this parameter: if it is too high the solubilization is not complete, if it is too low than the protein of interest is destroyed. Second, you need to keep in mind that some detergents have particular features, e.g. Triton tends to remove a bound ubiquinone, and it is a specific inhibitor for the respiratory Complex I, LDAO can be reduced, what could make a problem if the protein of interest has redox activity.

Dodecyl maltoside and octyl glucoside (both avaiable from Sigma) have been used by myself and many investigators for your attended use. The surfactants may be seperated from the proteins after isolation by gel filtration or ion exchange chromatography.

Hello John,

Unfortunately, I also recommend reading a comment written above by Cédric Montigny.

I know example that, in the case of detergents such as OG or DM, the difference in carbon chain length (only one or two carbon) of the alkyl group of detergent, or number or type of sugar of hydrophilic part of detergent, causes different results in structure and activity of the membrane receptor (Tsukihara, T. et al.: Science, 269, 1069-1074 (1995)). So that, I recommend screening of detergent for your target protein.

I think that, because non-ionic surfactants are mild for the protein activities in many cases, they are good to start the screening.

Depending upon your purpose you may need to choose detergent or a combination of detergents. I found DHPC works better in retaining native protein structure and function when proteoliposome is prepared with DOPC after Ca-ATPass purification. C12E8 and Triton X-100 are also widely used.

You can read my BBA article 2004 for further reading. Na deoxycholate and SDS are also used for this purpose in other systems.

Sorry, I refer my BBA 2006 article not 2004.

So is there a possible top 5 or 10 list of detergents that would be most appropriate for many membrane proteins or does each protein require more comprehensive separate testing? 

My personal opinion is that optimization of detergent solubilization of any integral membrane protein needs a thorough study. No general rules can be established; or, more precisely, there are a few points that must be considered. It is because in the solubilization process we try to replace the hydrophobic environment of the integral membrane protein in question; and each integral membrane protein might have a special micro-environment (e.g. lipids, sterols, interacting proteins). So your idea about a list of 8-10 different detergents for trying is exact. Good luck for the experiments.

Here are two additionnal comments.

Detergent exchange after purification should be a good opportunity to recover activity. Some detergent could be very good candidates from the solubilization/purification point of view but the purified protein should be inactive in this detergent. However, it doesn't necessarily mean that our protein is denaturated. As you probably know, interaction of a membrane protein with the hydrophobic phase (lipids or detergent) should be critical for its activity. A particular detergent that appears to be good for extraction should interfere with rotein activity without denaturing it e.g. for the Ca2+-ATPAse SERCA1a, it has been shown that DDM is good for its extraction from native membrane and its long-term stabilization but C12E8 was revealed to be clearly better to measure activities.

One of the contributors suggested the use of amphipols. I encourage you to use it with caution. Amphipols (amphiphile polymers) were shown to be good candidates for membrane proteins stabilization. Nevertheless, for some membrane transporters (that may explore large conformational changes), it has been demonstrated that amphipols block catalysis. Authors from the amphipols field naed it the Gulliver's effect, suggesting that the polymer surrounds the membranous part of the transproter like a rope, mechanically preventing the transporter to "breath" i.e. to change conformation.

One additional option which we have been using in the lab is styrene maleic acid co-polymers (SMALP). The advantage of these is that they remove membrane proteins directly from the membrane keeping them trapped in a layer of natural lipids. For AcrB we found an increase in activity compared to DDM, a similar thing has been seen for PgP. Detergents are not used at any point its different to most other techniques. We are not sure on how much flexibility it allows for movement, this is on our list of things to do.  If interested the paper is here. 

http://www.sciencedirect.com/science/article/pii/S0005273614003538

 would strongly recommend to a screen followed by  protein activity assay. This is very important as detergents many differ in solubilisation efficiencies and functional isolation of proteins. We have worked with DDM and LMNG and almost always tested several others before doing any real experiments. SMA is a nice emerging option but not necessarily a magic. Also test CHS. And last point you could also test mixed micelles that may actually stabilise your protein much more effectively than detergent alone.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4192586/

John Tainer ... Good question… Please check references below.. hope it will be helpful for you…

Regards…

http://wolfson.huji.ac.il/purification/Course92632_2014/Talks%202015/1M%20MembraneProteinPurification2015.pdf

https://www.anatrace.com/Technical-Documentation/Catalogs/Anatrace-Detergents-Booklet-FINAL

One aspect of protein isolation from membranes using detergents has not been emphasized in literature (at all? we have not seen that) and it caused us some heartburn in the early days. The question relates to heterogeneity in detergent:membrane/protein interaction. Much depends on the' inital hit concentrations', readily seen with ionic detergents since an already negatively charged (due to the first few molecules of detergent) complex tends to take less of further detergents molecules due to repulsion. This is even true for non-ionic detergent since the surface is being modified. We normally add the the detergent to an actively stirred suspencion in one go to ensure a constant hit on all particles/molecules. Reproducibility of data also improves dramatically.