Between those two assays, SRB has a higher sensitivity and better linearity. there's a few papers that compare them.

The two assays measure two different things: MTT assays for metabolic activity of cytoplasmic enzymes while SRB measures total protein amounts. Both assay can be correct yet give different results.

I´m agree Mario, but I need to find the most exact GI50, in order to estimate the LD50 from it for toxicological studies, and values can be really different, since curves have different shapes.

Marian, I do not believe that there is such a thing as a 'true GI50'. You can make a case for either one. Or one based on ATP. Or cell number. Or apoptosis. The assays can be all accurate and all give different results.

Marian, you would like to use GI50 for invivo experiments (on mice,...) from invitro results? If yes, i think it can not, you have to set LD50 on invivo models which you want

Marian, Very few in vitro cytotoxicity assays can give GI 50 values which you can extrapolate to in vivo model. But, if you are using tumor cell line, many studies show that clonogenic assay on soft agar gives data which correlates well with in vivo models for growth inhibition of tumors. For example, if you have 3 drugs with 3 different GI 50 values in clonogenic assay.... high, medium, and low GI 50 value...then this same trend is also seen when you use the three drugs in animal model. So, absolute GI 50 values maybe different in clonogenic assay and animal model...BUT, the drug which is most potent in clonogenic assay will usually be most potent in animal model. Likewise, Drug which is least potent in clonogenic assay will usually be least potent in animal model. The clonogenic assay is doable and not too expensive-and gives you a good idea of what your drug can do in animal model. Hope this helps-Good Luck!

Most of the replies to your question give really good advice, but the real question is what do you want to study in your experiments? If you want to compare toxicities of a series of drug as a screen then I have found SRB to be better because all SRB is doing is using protein staining as a surrogate for cell counting, so all you are really doing is determining how many cells have replication potential after drug treatment. Having said that I would add the caveat that you must start with a low number of cells in your wells which do not become confluent by the end point of your assay, which is usually about 1 week after drug treatment. MTT can be quite useful for short term exposure and measuring viability. Venil makes a really good point about clonogenic assays, which are the best assay you can possibly do, but are difficult if you a running a screening programme.

Finally Mario makes the really good point in that all these assays measure different things, and so will give different answers. Consequently their usefulness really depends on the biological question that you are going to ask and answer.

good luck

You can also refer the following question. Their is a lot of relevant literature shared in this discussion also. I think you will find the merits and demerits of both of them here and then you can correlate with your query, if applicable:

https://www.researchgate.net/post/Why_can_sulphorhodamine-B_assay_SRB_assay_for_preliminary_anticancer_screening_not_be_done_with_HL-60_Cells

Regards

Thank you all for your really good answers and viewpoints. I also agree Dr Sumantran, he got my point totally, because I´m trying to estimate my LD50 in order to use less animals for my in vivo experiments.

Thank you, Dr Nahata for the link

Just a doubt on SRB assay. Measuring total protein can be misleading because one would also measure protein within dying, autophagic, or dead cells-right?? Is this a dumb question?? A medical doctor once asked me how PCR can be used a s dignostic for pathogenic bacteria...he said PCR only measures bacteria that already died! I did not give him a good reply!