The problem is the color of dox that interfers with the MTT color.

You must define the wavelength of dox (it is published in the literature) and check the wavelengths you are using for your MTT measurements.

You can encounter various technical problems with the MTT assay as detailed in the attached articles.

And last but not least, the MTT assay does not measure "cytotoxicity" but "cell growth inhibition".

If you are interested by this last point, I can develop it .... as I am doing it on a regular basis ...

Best regards

Robert

PS: thus, why not trying LDH assay for dox-induced "growth inhibitory effects" ....

Thank you, Robert, especially for the papers.

MTT is an established method in our lab, and we use the corresponding corrections on color interference, etc. As far as I can see in the published papers on doxorubicin, the description of MTT is standard, there are no comments on any pitfalls. As for using the LDH assay, it would take several months just to buy a necessary kit in Russia. So, alas, currently we're stuck with MTT.

Are you using the same growth medium as was used for the measurement in the literature? I'm thinking that a difference in fetal bovine serum concentration could have an effect on DOX toxicity if it binds to serum proteins.

I understand Ksenia.

While I know that dox-related IC50 are published in the literature on a regular basis with the MTT assay, we encountered several times (but not always!) the same problem as the one you describe.

You should simply look by means of phase-contrast microscopy what happens on the bottom of your flasks with and without dox. You can perform a "quick and dirty" evaluation of dox IC50 by counting the cells on the bottom of the glass (with a grid in the occular of your microscope).

And yes, I fully agree with Adam's remark: the serum can markedly modify the data (see the attached article).

And why not using a compound other than dox?

Best regards

Robert

Dear Ksenia,

The problem may not be with the reproduction of protocol for IC50 assessment. Problem also arises with from where you have procured the cell lines from. Maturity period of cell lines, state of cancer when the cells were isolated, all together play role in IC50 determination. Duplicating the result with same line will not yield satisfactory result. Try MTT assay again with another batch of same cell line. That may work.

Secondly, if Doxorubicin is commercially procured, then there is 50% chance that problem is with Doxorubicin. Because, efficacy concentration varies with each batch of commercial packing.

Regards

Azeez

Robert: as we both know, MTT determines number of viable (and metabolically active) cells. If we have a treatment T, which decreased MTT assay signal (absorbance by formazan) relative to untreated control C, we can conclude that we have less cells in T relative to C. The question is what is the underlying cause behind it: Is it due to inhibition of cell proliferation, or due to cell death, or both? If our treatment was short  enough, and control cells did not have time to proliferate considerably during this time, then MTT assay would be detecting lower signal due to cell death in T (or loss of metabolic activity thereof). If our treatment lasted long enough, and control cells could have proliferated, a lower MTT signal in T signifies inhibition of cell proliferation relative to C, cell death in T, or both. Now, even inhibition of proliferation is sometimes a result of a sub-lethal cytotoxic insult, which did not kill cells, but induced their arrest and DNA repair etc...,so MTT assay can indirectly detect cytotoxicity as well, but it requires careful interpretation depending on how the assay was performed. I agree that MTT, XTT etc. are classified as proliferation assays, while LDH, dead-cell protease, PI staining etc. are classified as cytotoxicity assays, but proliferation assays performed under specific designs do allow to infer about cytotoxicity.

Ksenia: reproducibility or even repeatibility  of IC50 values is generally not very high, but 200 uM for doxorubicin is definitely inflated. To get comparable IC50s in reproduced experiments, one needs to use the same exposure time, the same cell loading, the same growth medium, check whether phenol red is present (it has mild estrogenic activity and can stimulate proliferation of some cells), drug solvent and solvent control in untreated control cells...to name just a few. Importantly, cell lines should be validated, which is quite increasingly required by journals to accept a manuscript, because cell line contamination by other cell lines (HeLa and others) is far too frequent. Talking about contamination: mycoplasma infection can also affect cell metabolic parameters and response to drugs, so it needs to be checked  too. In addition, out of my experience, certain cell lines with higher cell passages may respond differently then cell lines with lower passage numbers. Were cells loaded for the assay in the same status as in compared experiment (cells from an exponentially growing "healthy" culture not exposed to acidified spent medium? There is just many variables potentially involved, which need to be taken care of.

Roman,

Here below is my very own interpretation about "MTT versus cytotoxicity" ...

"Cytotoxicity" cannot be determined by means of the colorimetric MTT assay. You must use other tests than the MTT one if you want to measure cytotoxicity. The MTT test enables you to evaluate "cell viability".

Cytotoxic means direct "cell killing effects" induced by the drug of interest.

Cytostatic means that the compound of interested lowers the growth rate of a given cell population without direct cell killing effects. Cell death will occur as a consequence of a too long cytostatic effect.

A colorimetric assay can only bring "relative global growth inhibition information" because it is a relative test in which you compare the ODs of a treated cell population to the ODs of a control condition (untreated cells) arbitrarily scaled at 100%.

Thus, the IC50 / GI50-related values obtained by means of a colorimetric assay do actually not translate “cytotoxic” effects.

If one wants to determine actual cytotoxic effects for a compound of interest, the MTT colorimetric assay can be completed with the lactate dehydrogenase (LDH) test.

LDH is a soluble cytosolic enzyme that is released into surrounding culture medium upon cell damage or lysis during apoptosis or necrosis for example. The quantitative determination of LDH in the cell culture medium can be used as a marker for cytotoxicity.

Coming back to a colorimetric assay such as the MTT one, when one obtains a concentration (for a given compound) decreasing by 50% the global growth (after x days (usually 2 or 3)), i.e. the GI50 concentration (or the IC50 as commonly used in the literature) you do not know whether your compound of interest killed 50% of the cells (cytotoxic effects), whether it inhibited 50% of the cell proliferation (cytostatic effects), whether it detached 50% of the cells (anti-adhesive, i.e. "in vitro antimetastatic" effects), etc..., etc...

Once you have determined the GI50 / IC50 concentration for a given compound on a given cell line, you must use complementary biochemical and/or morphological techniques to determine whether your compound is cytotoxic, cytostatic, anti-adhesive, etc..., etc...

The two attached articles by Galluzzi and colleagues (2012, 2015; Appendix-1 and Appendix-2) are of great help in this domain.

The attached article by Kornienko et al. (2013; Appendix-3) reviewed various chemicals that are able to induce non-apoptotic cell deaths in cancer cells.

Coming back to the IC50 / GI50 values obtained by means of a colorimetric assay (as for example the MTT one):

in the Mathieu et al. (2009 (Appendix-4) and 2015 (Appendix-5)) articles, the MTT test-related GI50 concentrations relate to actual cytotoxic effects.

In the Lefranc – Nunzo et al. (2013; Appendix-6) article, the MTT test-related GI50 concentrations relate to cytotoxic effects that in turn do not relate to apoptosis …

This means that each cytotoxic effect does not “universally” translate into pro-apoptotic ones.

In the Van Goietsenoven et al. (2010; Appendix-7) article, the MTT test-related GI50 concentrations relate to cytostatic effects, neither to cytotoxic nor to pro-apoptotic ones.

Be aware that you cannot always translate the MTT test-related growth inhibition of a given compound into a precise GI50 value. Some compounds reach a “plateau” of inhibition (see Lefranc – Nunzo et al., 2013; Appendix-6).

Lastly, you can also have "false" data generated with colorimetric assays (see the attached article by Chan et al. (2013; Appendix-8) and the first NCI-60-cell line-related article (Shoemaker, 2006 (Appendix-9)).

The US NCI set up a fantastic tool to characterize the effects of a given drug in terms of growth inhibition in a panel of 60 cancer cell lines belonging to >10 histopathological types (Shoemaker, 2006; Appendix-9).

The US NCI clearly defined by means of the combination of the GI50 (growth inhibition), the LD50 (lethal dose by 50%) and the TGI (total growth inhibition) how to make the difference between a cytotoxic and a cytostatic compound:

The US NCI-related GI50 value corresponds to a global growth decrease by 50% induced by a compound on a given cell line “x” days after having cultured the cells with the drug and in comparison to an untreated control condition (= 100%) grown during the same time;

The US NCI-related LD50 value corresponds to the a global growth decrease by 50% induced by a compound on a given cell line “x” days after having cultured the cells with the drug and in comparison to the initial number of cells in the untreated control condition;

The TGI is the US NCI-related parameter to determine the concentration needed to kill 100% of the treated cells.

It is by comparing the GI50 to the LD50 value that one can determine whether a compound is cytotoxic or cytostatic, and not at all with the sole GI50 value.

We are using morphological approaches in the research unit to which I belong for determining whether a compound is cytotoxic or cytostatic (see Lefranc-Nunzo et al., 2013 (Appendix-6); Mathieu et al. 2009 (Appendix-4), 2015 (Appendix-5); Van Goeitsenoven et al., 2010 (Appendix-7)).

The US NCI is not so far for having tested about 800,000 anticancer drugs, whose data are publicly available on the NCI website https://dtp.cancer.gov/databases_tools/data_search.htm

I actually benefited several times from the amazing help of the NCI in identifying the mechanism of action of an innovative anticancer compound (see for example Frederick et al. JMC 2011 (Appendix-10)).

Hoping that this long explanation would not be too boring,

Best regards

Robert

Dear Robert and Roman,

Both of you discussed good example,

Dear Robert,

you find any differences in the text you have writen...

The US NCI-related GI50 value...

The US NCI-related LD50 value ...

I didn't find any difference except the name.

Lethal compounds includes both inhibition of proliferation  and induction of  death. No any compounds induce cell death...Cell death is consequence of the inhibition of metabolic process and other functional process either by damaging, breaking, binding etc.compounds interacting with DNA may be directly cytotoxic because of our traditional model of defining cytotoxicity.. e.g. campt, 

It is very difficult to say which compound is cytotoxic or cytostatic or apoptotic or genomic instability inducer.

Dear Ksenia...

One of the most important point, which none of the above reviewer mentioned and seems to be importnat in your case is oxidative stress mediated mitochondrial proliferation and increase in mito mass. Cells adopt to enhance mitochondrial number to cope the stress, this leads to enhanced formazan crystal formation and reduced observed toxicity. Therefore, selection of cell lines is important.

Second, option is cell lysis using DMSO vs SDS...Many cells can't be lysed simply applying DMSO and inability to dissolve formazan completely leads to abrupt observed results.

Third, is concentration of MTT and solubility of tested compounds, interaction with serum.

 Best regards

Krishna

Dear Krishna,

Please read carefully the review of Schoemaker in Nat Rev Cancer.

The GI50 represents the concentration that decreases by 50% the growth of the treated cell population as compared to the control one AT THE END of the experiment (48-96 h).

The LD50 represents the concentration that decreases by 50% the growth of the treated cell population as compared to the INITIAL (at t = 0) CONTROL cell population.

It is by comparing the GI50 and LD50 values of a given compound that you can define whether this compound is cytotoxic or cytostatic.

Best regards

Robert

PS: once again, please read carefully the above-mentioned article.

Dear Krishna,

While your response above seems to be issued by an actual expert in the field of anticancer compounds, and more precisely about the definition of cytostatic versus cytotoxic effects, I did not find any references from your CV listed on RG indicating me your actual expertise in these domains.

Would you be so kind in sending me your article(s) that I missed.

Thank you very much,

Robert

Dear Robert...

I worked during my PhD on this topic...its part of my thesis..many work remains unpublished due to publication shop, rented reviewers.and last but not least project nature etc.etc.my thesis is some one else properties and i can't forward it to you. Hope i am able to convey my message.

Best regards

Krishna

Dear Krishna,

I am asking for high-level peer-reviewed journals in which you have published your affirmations, not a thesis.

Best regards

Robert

Thanks to everyone for this very helpful discussion! I hope we'll solve our issue soon.

Dear Ksenia,

maybe you can take in account the sulforhodamine B method as a solution of your problem. We use it routinely with HT-29 cells in the screening of natural compounds with anticancer activity. Here I leave a reference that describes the method:

Vichai V, Kirtikara K (2006) Sulforhodamine B colorimetric assay for cytotoxicity screening. Nat Protoc 1:1112-1116

Best regards,

Jose Manuel

Thank you, Jose Manuel!

Regarding the possible interference of doxorubicin (spectral properties) in the assay, there are also concerns with the PI staining method.

https://www.researchgate.net/post/Can_I_use_Doxorubicin_treated_cells_for_PI_staining_and_FACS