There are various possible reasons for your results. First, not all proteins will interact with your protein quantitation reagent the same way. When you have a mixture of proteins, as with a cell lysate, it is not that uncommon that loading based upon a protein standard curve will not results in consistant results for your protein. We have found that, where possible, counting the cells before lysing and loading lysate from the same number of cells generally gives us much more consistant results. Another option is to try different protein standards and see if you can find one which gives you more reproducible results. In addition, you may want to consider the possibility that your treatment affects levels of your housekeeping gene product and try different housekeepking gene products Another possibility might not be with the loading but with uneven transfer. Coomassie staining of the gel and Ponceau S staining of the membrane can help to determine whether this may be a factor. Finally, it may seem trivial, but sometimes the problem is in the preparation of the stock. Even small dilution variations can be compounded and throw off your quantitation. I like to include a few dilutions of my sample to make sure I am getting a similar linear relationship with dilutions as seen with the standard.

Truth is I haven't checked more housekeepings. As for the transfer I think it is so far quite succesfull since you can even see the marker very clearly on the membrane and my blots do not lack protein amount. I usually extract proteins from my cells with RIPA or Urea and I have noticed better results whenever urea is used. I do not know if it is related to my problem. I haven't considered using dilutions of the sample so far but I don't usually have much sample available. I ll try it and also see what I can do with cell counting before loading. Thank you very much Mrs. Hanson.