Hi, I'm not a membrane proteins specialist, but I used to perform SDS-PAGE and WB. According to me, you have a technical issue with your extraction and solubilization protocols. I would recommand not to perform your western blot with these prot extracts. The litterature is certainly full of protocols for this kind of proteins.

I've run many SDS-PAGE experiments. I get these sort of lines when I've not properly denatured the samples before loading - your samples have not run properly. I normally heat my membrane protein samples to 70 deg for a few minutes before running on SDS-PAGE - this doesn't cause me any aggregation problems (so far!). If I were you, I would also use less sample. 

Basically I think you need to optimise your SDS_PAGE setup before running WB. 

The vertical blue lines are protein. This may have occurred because the viscosity of the sample was too high owing to the high concentrations of urea and detergent. One thing you can try is to perform a TCA precipitation to get the protein away from all that stuff, then resuspend the pellets in 1X sample buffer.

I have experienced exactly such streaks due to DNA/RNA contamination of membrane fractions. The "outlining" of the lanes is particularly telling, and in highly contaminated samples you will also see "hourglassing" of the lanes. Any method of extraction that disrupts nuclei even to a small degree will cause chromatin contamination of membrane fractions. Chromatin is sticky and will adhere to membranes and copurify, even through sucrose gradient fractionation. Often the easiest solution is to include DNase I/RNase A  (obviously protease free MB grade) in your lysis buffer, or treat the samples with the enzymes for 30 min on ice after purification (less expensive).  As for the mechanism by which DNA/RNA contamination causes streaking (I doubt that the streaks are actually protein), I leave that to the theoreticians. From empirical observation, however, I can say that unusual viscosity of SDS-PAGE samples will almost always be due to DNA contamination. I have routinely run membrane protein samples containing 8 M (yes, 8M!) urea and high detergent, and there is no noticeable increase in viscosity... and it invariably improves the running of membrane protein samples (especially on gels also containing 8M (yes, 8M!) urea). Just as an added note, DNA doesn't necessarily come from just your samples. If you have bugs growing in your top electrode buffer, or in stock solutions you used to make your sample buffer, you will see the same thing (not to suggest that this is what has happened in your case; I'm just saying). Also, in my own experience, protein precipitation prior to SDS-PAGE, especially hydrophobic proteins, opens a whole new can of worms!

I have worked with membrane proteins for long time. if you are using isolated membrane from tissue, then your samples will be hydrophobic due to the membrane contents (lipid). This can cause membrane proteins to aggregate in the gel. to overcome this problem, i suggest to use this loading buffer that i used in my experiment:   loading buffer (3X) containing 0.3% bromophenol blue, 6% sodium dodecyl sulfate (SDS), 30 % glycerol, 5% β-marcaptoethanol, and 150 mM Tris-HCl (pH 6.8). Also you prepare 9 M urea stock (~maximum soulability) and add 1 microlitter to each sample you are loading. this is followed by placing the samples in hot path 100 °C for 5 min ( more or less than this time will form aggregation)

Good Luck

@Norbert Kartner    

What concentration of RNAse/DNAse do you suggest I use? Would it be better to treat at 37°C for higher enzyme activity?

Thanks a lot!