The issue might be further downstream than in the tissue preparation. As far as I can tell the buffer should be fine for membrane protein isolation. (If you are unsure just try total lysis in 1x loading buffer).

Which type of membranes do you use? I usually found that PVDF works a lot better for membrane proteins than nitrocellulose. Also there might be a difference in transfer technique (semidry vs. tank blotting).

Maybe, you need to use more aggressive extraction agent? (Triton X - 100, for example, in case membrane-binded proteins)

Thank you Sascha, I was using a nitrocellulose so perhaps I can try PVDF. I am using wet transfer and will also try with Triton-x100.

In my hands (large mammalian membrane proteins) PVDF works a lot better than NC for two reasons:

• PVDF doesn't fragment during incubation, NC all too often does :-(
• PVDF has a higher affinity and capacity for proteins than NC. I once used spot-blotting a 125I-labelled protein to assess this, under identical conditions PVDF bound 95% of the radioactivity, NC 60%. Considering that separation of iodinated protein from free iodate was by desalting only, the protein sample may have contained a small amount of unbound iodine, so recovery with PVDF was essentially quantitative.

Blotting at least of large proteins (> 100 kD) works a lot better with tank than with semi-dry blotters (possibly because gels heat up much more with the latter).

From my experience, I would also recommend Dunn's transfer buffer (10 mM NaHCO3, 3 mM Na2CO3, see doi:10.1016/0003-2697(86)90207-1) over Towbin's. It not only works better, it is also cheaper. For large, hydrophobic proteins add 100 µM SDS.

Last, but not least, membrane proteins are often heavily glycosylated and hence produce fuzzy bands in SDS-PAGE. CTAB-PAGE (doi:10.1016/S0003-2697(02)00639-5 ) produces much crisper bands and is compatible with 2D-electrophoresis, blotting and the common staining protocols.

CTAB is also very effective for initial solubilisation of membrane proteins (actually more so than SDS), at the same time it is mild enough that there is a good chance to maintain enzymatic activity of your protein during solubilisation (which SDS only very rarely does).

You mean you extract plants expressing GFP only with the same protocol and you get signal on WB?

With transmembrane proteins you may have problem with the denaturation. Instead of 5-10 min at 95°C, you may incubate your proteins with the loading dye at 80°C or even 65°C for up to half hour.

Otherwise I'd agree with Sascha Röth.