Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Purification of proteins from human tissue, SDS_PAGE

  1. Aug 16, 2005 #1
    I plan running SDS-PAGE to get the protein profile of human tissues. Which kit is best for this purpose and how can i measure the concentration of a mixture of different proteins? I can measure the protein concentration when it is just a single kind of protein in the mixture, but how about when it is a mixture of different porteins?

    Thanks for any helps and ideas!
  2. jcsd
  3. Aug 16, 2005 #2


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    The assay to measure can be the same as the one used for a purified protein such as colorimetric assay. However, the bradfort assay might not be the best because bradfort reagent has a bias for certain amino acids and it is more sensitive to BSA. The assay by Gornal which used the biuret reagent might be better.

    Also, stay away from absorbance assay such as A280 readings.
  4. Aug 16, 2005 #3
    Depending upon how much money you have, the best setup you could get would be the BioRad with pre-cast gels. Most ppl, however, have to pour their own polyacrylamide.

    Become familiar with Western Blotting, since it sounds like that is what you really want to do. You will need antibodies specific against the protein(s) you wish to study.

    As far as protein assay goes - Bradford will work fine, but you may want to consider how you build your standard curve. IgG, for example, has a standard curve whose reponse is closer to 1:1 than BSA, yet most people use BSA still because it's readily available. It all depends upon your particular experiment and what you are trying to do.

    Typically, after cell harvesting and possible some cell fractionation, a protein assay will be done to determine the concentration of proteins from each cell culture. Then you will mix up the samples with solubilizing buffer, bromo blue dye and BME or DTT - at this point you will dilute them appropiately, using the assay data, so that they all have the same concentration. Then when you run them on a gel and probe with an antibody, changes in the appearance of the autoradiograph can actually be used to deduced cellular changes in a protein (or translocation).
  5. Aug 16, 2005 #4
    btw, you also know that SDS-PAGE is not a good purification method (except perhaps as a final step).

    however, SDS-PAGE can give you a good idea of purity while working out a purification protocol, particular when you silver stain the gels to see that you have narrowed down the bands.
  6. Aug 16, 2005 #5

    The Biorad stuff makes it so much easier. If you do go that route play around with reagent amounts sometimes you can get away with less than what they tell you.
  7. Aug 18, 2005 #6

    It sounds like you are familiar with this method. :biggrin: I also plan doing it after running SDs-PAGe. i am going to measure the concentration of each band and for doing it i will use B-actin as a normalizer. My question; how can i detect both the target band and B-actin band in the same lane? Should i cut the membrane and hybridize each of them in their specific antibody or can i hybridize a membrane filter with two antibodies at the same time? Any suggestions?

  8. Aug 18, 2005 #7
    Well, i will work with solid tissue not cells. :confused:
  9. Aug 18, 2005 #8


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    If you are doing quantification of the protein with your SDS-PAGE, you should not use silver staining, it offers poor quantification. Coomasie blue stainning is better for quantification but it is less sensitive than silver stainning. If you require both quantification and sensitivity, I recommend that you use fluorescence stainning but you will need special equipment for that.

    It is possible to detect two different proteins in a western blot but you have to choose your primary and if needed your secondary antibody carefully.
    Also, detecting multiple protein is called Multiplex Western Blotting.

    First, you can choose to use directly labeled primary antibody with different amine dye. This will give you different colours on your blot. The other optium is to use primary antibodies from a different animal (ex: goat and mouse). So you will need two different secondary antibodies that are conjugated with different compound. This will give you different colours.

    Look for the Multiplexed Western Blots: Detecting Multiple Protein Targets Simultaneously section. You should find some info that may help you.
    Last edited by a moderator: Apr 21, 2017
  10. Aug 18, 2005 #9


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Have you confirmed that in the tissue you're examining, beta-actin is stable? A lot of people use beta-actin to normalize results, but you should always double and triple check that there's no evidence in the literature that its expression can vary in your tissue or cell type of interest, because there are situations where it could differ. Just be sure you're not using it only because you've seen a lot of other people use it, and that it's really the appropriate loading control for the tissue you're studying.
  11. Aug 19, 2005 #10
    painfully so, after having done hundreds of them.

    first of all, yes, you can detect both bands if their molecular weight if far enough apart that they are distinguishable on the autoradiograph. you don't need to bother that the primary or secondary antibodies are from different species in this case.

    if their mw is such that they are too close together, then you can use different species antibodies - this can be tricky and will depend upon how strong your signal is since you will inevitably get some signal from the last antibody that didn't completely wash off. you western for one protein, strip the antibody off in glycine solution, then probe with the other. (if your protocol works well, this isn't a bad practice if you are looking at alot of proteins - it can save you time).

    normalizing against actin can work out ok, but i would personally recommend against it unless you are look for something that is very dilute (such as receptors). i would go ahead and do the assay, and make them all the same concentrations.

    some ppl - instead of making them different concentrations - will just load different amounts on the gel. i have found that this is poor practice. if you do this, the solubility may vary enough so that one band appears darker when in fact there is the same amount of protein there - not good. of course, run several controls.
    Last edited: Aug 19, 2005
  12. Aug 19, 2005 #11
    it doesn't matter - you will need to homogenize the proteins, only in your case i suspect that you're grinding up a tumor instead of sonicating cells in solution.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook