Big differences between BCA and Bradford? - Two assays give very different results (Feb/10/2010 )
Hi!
I've been trying to achieve equal loading with my protein samples and for this I have been using a Bradford Assay. My samples are crude total soluble protein extracts from tobacco leaves and the buffer I'm using contains HEPES, KAc, MgAc, EDTA, DTT and PMSF. I've always used the bradford before with this buffer and never had any problems like this before.
Anyway, no matter how hard I try, by using the values from the bradford I never get anything like equal loading, even though my replicates are all almost identical and my curve is always near perfect (Rsq=0.998). So, on the advice of a professor in my dept (not my advisor, because she has no protein knowledge), I tried the BCA assay to see if I would have any luck with it. And this is where I hit another problem:
Based on my bradford, my samples were mostly around 1ug/ul so I set up my BCA curve from 2-10ug/ul and added 4ul and 8ul in triplicate of all my samples to the plate, hoping either one of those would be in the range of my curve. What I ended up getting was a lovely standard curve but my samples were way beyond the range of it, above even the highest concentration on my curve. This would imply that I actually have a lot more protein in my samples than I thought, OR that the BCA is way off. Since the values are so different, which one should I trust? Or is there another option I can take?
Thanks for your help!
vojera on Feb 10 2010, 01:38 PM said:
I've been trying to achieve equal loading with my protein samples and for this I have been using a Bradford Assay. My samples are crude total soluble protein extracts from tobacco leaves and the buffer I'm using contains HEPES, KAc, MgAc, EDTA, DTT and PMSF. I've always used the bradford before with this buffer and never had any problems like this before.
Anyway, no matter how hard I try, by using the values from the bradford I never get anything like equal loading, even though my replicates are all almost identical and my curve is always near perfect (Rsq=0.998). So, on the advice of a professor in my dept (not my advisor, because she has no protein knowledge), I tried the BCA assay to see if I would have any luck with it. And this is where I hit another problem:
Based on my bradford, my samples were mostly around 1ug/ul so I set up my BCA curve from 2-10ug/ul and added 4ul and 8ul in triplicate of all my samples to the plate, hoping either one of those would be in the range of my curve. What I ended up getting was a lovely standard curve but my samples were way beyond the range of it, above even the highest concentration on my curve. This would imply that I actually have a lot more protein in my samples than I thought, OR that the BCA is way off. Since the values are so different, which one should I trust? Or is there another option I can take?
Thanks for your help!
I may be no help at all, but I would try a THIRD protein quantification method, and go from there... GE´s 2D quant-Kit is awesome, I was having lots of trouble quantifying my protein samples, tried lowry and biuret method, and all gave very strange results... GE kit worked wonders and I have been using it ever since!!! =)
what did you use for the standard curve? with bsa the bradford gives a response 2x that of igg. there is a lot of protein-protein variation.
as rachelhauser said, you should try a third method to validate one of the other methods. we used a fluorescamine assay and found that the bradford with igg as standard was valid for our protein (it also matched the lowry and bca methods). you can also use kjeldahl (or microkjeldahl) nitrogen determination to determine which method is valid for your protein (you should only need to do this once, then you will know which method you can rely on.
It could also be that the tobacco has some component that absorbs strongly at or near the wavelength you are using to read your plates for the BCA, or an inhibitor of the Bradford assay (though what that would be I don't know, I'd go with the former idea first).
Thirding the try another assay/protein determination method.
Further to Bob’s comment.
What concentration of DTT do you use? Any reducing agents in the buffer will generate the Cu1+ ions used for BCA detection.
Sometimes the presence of NP-40 could affect the result. Just try to add some NP-40 buffer to your bradford and you will see a really nice blue color. (I dont know if you use Np-40 but if so, maybe that could affect)
Hi everyone, thanks for your responses! I'll have an ask around the dept to see if anyone uses those other methods - we're on a buying freeze unless I can show that it'll definitely work for me.
I'm using 1mM DTT in my buffer, is this too much for the BCA?
Also, I always suspect a big part of my problem is that these extracts are colored (i.e. they still have chlorophyll in them are are green or brown rather than clear). Is there any way to get around this?
1 mM DTT should be okay for BCA. See: http://www.piercenet.com/Objects/View.cfm?...CA-F0C564454B72
vojera on Feb 12 2010, 07:30 AM said:
I'm using 1mM DTT in my buffer, is this too much for the BCA?
Also, I always suspect a big part of my problem is that these extracts are colored (i.e. they still have chlorophyll in them are are green or brown rather than clear). Is there any way to get around this?
you can blank for your buffer components, most will have a small enough effect. edta may have a more serious effect on the bca (and lowry and biuret). how much do you use in your buffer and how much goes into your protein assay?
color in the extract, whatever the source, will interfere with your readings. you may want to try a fluorescent protein assay like these (available from invitrogen). but you should also ensure that the chlorophyll doesn't interfere with the assay (but you should be able to blank for it).