Your cell lines maybe forming resistance to the anti cancer drugs that you are treating them with. When given at low concentrations for long periods of time cancers cells have the ability to mutate and/or start up regulating the expressions of drug resistance proteins such as pgp. You could test your cell lines before and after treatment for these drug resistant markers.

Hello,

from what you describe, that sounds like a hormetic mechanism to me. It is quite common that at low doses, various compounds have effects opposed to what you observe at higher concentrations (a.k.a. biphasic responses), for more details, c.f. Wattson, M., Calabrese, E., 2010. Hormesis: a revolution in toxicology, risk assessment and medicine.

However, I assume the graph you posted is the dataset of the compounds dose-response curve you refer to. Usually you should find quite continuous dose-response-relationships without significant "hops", which is the case with concentrations of E-06 and E-03. How do your individual replicate treatments look? Is the hop caused by strong single outliers? That can happen if your cell monolayer is detached and lost during removal of the treatment medium. I assume that was the case in this test, so I recommend repeating the test taking care to not lose the monolayer.

Thanks Rasika, that was what came to my mind at first. Coincidentally I found about hormesis just like what Patrick said. However the only way to confirm this is to do some assays to determine which pathway was activated, eventually will be carried out once I am done with monotherapy treatments. Patrick, I did not have any single strong outlier so far and the same pattern appeared in 4 cell lines and 4 anticancer drugs. The graph was just an example, the actual lowest concentration was 2.56x10-11 M. From 10microMolar with 5-fold dilution. At concentration from 10microM to 6.4x10-10 M was a nice curve. Thanks for the help guys. I need to prepare for my PhD progress presentation, that is why I have to figure out possible explanation for this.

Hi Reagen, this is a tricky one. The answer is either going to be a technical problem or a real effect.

With regards to the technical do you also add your drug carrier to your negative control (i.e. if your drug is in DMSO and you add 1 ul/ml do you also add (just) 1 ul of DMSO per ml to your negative control)? I sure you are but if not this is essential for any drug treatment of cells (to separate out real response to response to the carrier).

As you are looking at absorbance then the presence of the carrier may have an effect (especially if you are adding at more than 1 ul per ml).

Personally I would look at another platform to examine the drug effect. If, for example, you have a FACS machine you could look at either proliferation differences (using true-count beads or similar) or use assays for signs of early apoptosis (such as Annexin V staining).  If you get the same result then one can eliminate the technical and focus on the biological. It may be that low doses act as  "irritants" which somehow stimulates proliferation (I am thinking of the recent discovery that low doses of alcohol actually stimulate neuron growth in the hippocampus rather than killing the neurons).

Sorry I can't be more help,

Gary

Thanks Gary. I did vehicle control at 0.1% DMSO and without DMSO. There was no statistical significant. The low doses treated cancer even had higher absorbance value than the controls (vehicle and no DMSO added)

Often, doses lower than those that would be cytolytic or anti-mitotic are inducers of the ER stress and/or heat shock protein-like factor response.  These doses are anti-apoptotic because of induction of protective protein expression, while higher doses of the same drugs overwhelm the inductive apparatus and are pro-apoptotic or anti-proliferative. 

Thanks Nina. Yes, I have read some information on various mechanisms that potentially give cancer protection/resistant against drugs such as heat shock protein and others. It was just that there are so little publications that I can accessed from my University, showing u/ inverse u shaped graph curve for me to make the connection and/ or relate to my findings in the beginning. The publications that I could find will only show best fit lines or just some tables with IC50. However, I will be doing assay via flow cytometry at the end of single drug treatments. Another 4 cell lines and 12 drugs to go.

It could simply be a plate edge effect if your negative controls are consistently placed in the outer wells. In my experience, those effects can be quite strong and introduce a systematic error into your calculations. Those effects are particularly pronounced in a long experiment like 72hrs incubation with lesser volumes of the medium. The solution is quite simple - do not use the outside perimeter wells of the plate but still fill them with the medium to reduce evaporation from the inner-positioned wells. Alternatively, vary the position of control replicates such as to have DMSO controls on the inside as well, not just at the edge of the plate. Hope it helps.

Hi Vadim, I agree with you about the outer wells. That is why I do not use the outer wells and they were filled with media to avoid any evaporation. They were carried out in triplicates (seeded with the same cell line from 3 different flasks per plate to ensure reproducible result and triplicate wells in a plate). I am using MTS and PrestoBlue that do not require any washing. I am still waiting for SRB to arrive from vendor and I will compare the IC50 of the drugs to see if there is any significant changes/shifts. I managed to determine the IC50 using MTS and PrestoBlue, it is just that I am curious as to why this phenomena occurs. It could be a question that will be asked by lecturers during my progress report presentation. However, I am grateful that everyone is helping me to figure out possible explanation because at this stage it is actually hard to pinpoint the exact cause until necessary assays are carried out to confirm the mechanisms involved. I will take all of the comments given and read further details on them.

Hi,

another possibility may be the presence of different effects that can not be differentiated by the MTS assay. High concentration => reduced proliferation and increased apoptosis is the main effect and reduced dye turnover is measured. Low concentration => no effect on proliferation but stimulation of (mitochondrial) metabolism (increased turnover of the dye)? You should try to validate the results with a "real" proliferation assay (3H, BrdU).

Good luck

Norman

It could be reaction from MTS but I repeated the experiments with PrestoBlue and the same pattern observed. However I will carry out the proliferation assays to confirm these findings. Thanks Norman

...but even PrestoBlue is modified by the cellular environment and possibly the turnover can be influenced by the cellular state (or the substances). But I have not worked with this assay so it is an assumption only.

That could be the culprit, I hope.. I am still waiting for SRB to arrive, I will compare the results and update in this forum.

Are you sure the compounds you are using are not doing some funny business with the MTS or Presto blue by themselves? Try just medium plus the titration of the compound without cells present in the wells and go through the assay as a negative control. Sometimes compounds can react with ATP sensor (admittedly rare event, but then it's an easy experiment). I do agree with previous comment that BrdU assay could be the best choice in this case. Roche used to have a good kit

Thanks Vadim, I will be carrying the experiment tomorrow and will include controls with drugs alone to make sure that there is no significant difference with control media alone. I will update the findings here at the end of the 72 hr incubation.

Good luck! You just might need some with this one

We have seen this effects many times in the past. Drugs at very low concentrations (far below IC50)  stimulate cell growth.

The drugs stimulate various kinases (stress reaction).

There is another question that I would like to know regarding the IC50 since at very low dose there was higher reading so it will affect the IC50 calculations. Do we take into account the low doses or we omit as the literatures' IC50 was only at the range of nanoMolar to microMolar, mine was 10^-12 nM and maximum at 10 microMolar? 

From the graph you attached, I found the top concentration was 1M, was it true? You also mentioned the range was 10^(-12) to 10 uM. What type of anticancer drugs you were testing? I didn’t see the error bars, was it done in duplicate or triplicate? What’s the Z’ of your assay? Not sure whether you have the results from MTS and drug interaction. If the drug can chemically reduce MTS, the higher concentration should give higher reading (if the inhibitory effect on cell growth is not so strong). May I know what cell lines you are using, the cell number per well, and the plate layout? If there are too many cells in a well, they may die after reaching the plateau. What’s the media color like of your control wells? Wells treated with low concentration of drug may slow down the cell growth a bit but the cells are still healthy.

Regarding the IC50 modelling: If you encounter significant hormesis, a standard model (e.g. 4 parametric logistic) may not be appropriate, as you cannot fit the function with a satisfactoy coefficent of determination. In such cases you should stick to functions dedicated for such hormetic dose-response-relationships. For example, the Brain and Coussens, cf. http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0033432. This function works great for hormesis (see figures), although calculation is quite elaborate and may not be possible in your particular statistics package.

If you want to include the apparently increased viability into your discussion as effect, you should stick to parameters other than IC50, such as the "no observed effect level" (NOEL) and / or "no observed adverse effect level" (NOAEL) concentrations.

Thanks Patrick. Hi Haishan, I have mentioned in my previous replies that the attachment was just example. I did triplicate per plate and triplicate plates where cells of the same line but from different flask per plate to ensure the result is reproducible. I cannot really disclose the anticancer drugs and cell lines that I am using but there are about 16-20 drugs and 8-10 cancer and normal cell lines. I have plated 2000 to 2500 cells/100 ul per well and incubate for 72 hours taking into account the 0 day too. Right now I cannot proceed with testing the drugs interaction with MTS alone as I am out of MTS. I am still waiting for the delivery from vendor.

With so many different combinations of drugs and lines you are bound to get some unusual responses - stress pathways, mitochondrial activators, stimulation of some proliferation pathways while targeting others, if it's B- or T-cells, cell activation, etc., etc., At the end you'll have plenty of less complicated and more straightforward responses you should focus on and just let some other ones go - not everything can be always explained in this world, after all...

Thanks Vadim. I am doing single drug and multiple target drugs for synergistic effect. I am using various inhibitors and also probably compounds from plants in collaboration with other university. I have a lot on my plate as this is just the first stage of my PhD study to screen the effectiveness of the drugs alone and combined (2, 3 or even 4 drugs combinations targeting various proteins and cell cycles), the later stage will involve flow cytometry, pcr, wester blot and few other assays may also be added in.

The MTT assay, like the others of the family, measures mitochondrial quinone redox levels. In many cancer cells I have examined I find changes in the balance of glycoloic/aeroboic respiration. Essentially the cells can either up or down regulate their mitochondrial levels in response to a stimulus. I do not trust either MTT or the LDA assays for cell numbers. I now only use direct cell counts. 

Yes, low doses of anticancer drugs might cause higher cell proliferation. Choosing very low doses (depending on the type of the cell) can cause opposite results ( increase in cell proliferation). You can read: The Comparison of The Effects of Silybin and Silybin-Phosphatidylcholine on Viability and ESR Expression in Human Breast Cancer T47D Cell Line (see figure 1A and 2A, dose 50 µM).

Yhis is an important question for a number of reasons but one is that vascular growth is abnormal in solid tumours and the concentration of systemically administered drugs will be very low in some regions where cells survive for long periods anaerobically.

It is probably either a change in the metabolism (induction of enzymes); or your compounds really have a proliferative effect on cancer cells in lower doses. I think you should simply count the cells with Trypan blue after treatment, just to see whether the number of cells is actually higher then in the control or not. If it has a proliferative effect, no matter which cytotoxicity assay you do, the lower doses will always have more then 100 % viability. That is why I think the counting of the cells should be the next step. 

Yes. I will be conducting flow cytometry to confirm the effect. It's just that journals always get a perfect curve even at low doses, using the same MTS assay. This effect not only happened to me but also my lab colleague doing different kind of cell lines. Then a friend of mine suggested me to use wst-1 that would give better curve and would not show any sudden bump at low dose as she also encountered the same issue until she changed to wst-1.

 Please find the attached file.

I seeded at 5000 cells/well/200ul complete rpmi with 10% serum, incubation for 72 hours before MTS. Technical triplicate and biological triplicate.

I have another comment which is not related to your recent question.

You mentioned your volume is 200 µL in each 96 well. As you know the total volume for a flat bottom well is 360 µL and the recommended working volume is 75 to 200 µL; and you are using the maximum recommended working volume of a well in a 96 well plate. I think I know why; you are not renewing your drug and medium at all. Am I right? Please check the half-life (the duration of action of a drug is known as its half-life) of each drug that you are using. For example some drugs should be renewed each 24 hours (or less and more) for cell treatment.

I just want you to be sure you are doing the right thing for each drug. So check it.

Thanks Narges for the input

Hai Paige, I have also tested an Akt inhibitor but the lowest 2 doses had slight higher absorbance reading on a few NPCs. I will be using flowcytometer to validate my screened targeted therapies. Using MTS/SRB is just to narrow down the most active.

HI 

I have a question too, not sure might be  related.

I am treating cancer cell with different concentration of drug.

Drug start to reduce  viability as concentration increases but suddenly in highest concentration  there is a peak of high viability 

Bahram, what viability assay did you use? MTT? When I did my flowcytometer with 7AAD highest concentration of drug has lower dead cell than half the concentration. This pattern might be the same as yours but since it is just 7AAD so it is still inconclusive untill I run with other stain.

It could be.

it is reported that anticancer drug WNT-974, a porcupine antagonist, could induce cardiac and renal cells proliferation, although it has an anti-proliferative effect on hair follicles, bone and cartilage cells.

 http://www.pnas.org/content/early/2017/01/30/1621346114/F9.expansion.html

We can see this effects in some plant extract also. at very low concentrations below CC50  stimulate cell growth,and at the concentration above CC50 reduced proliferation of the cell