Hello Mauricio,

the time of incubation depends on the compound and on the cells. In HeLa, 24 h is usually quite enough; in primary cells, 48 h might be insufficient. It is probably best to take a look at the literature and see what incubation times others use for this compound and/or this cell line. Changing medium every 24 h sounds good.

Also, you did not mention which plate format you are using - I guess, 96-well? 2000-3000 cells might be a bit too few, if those are proliferating slowly - every small pipetting error gets augmented. I would also suggest using a well-known compound that should definitely affect proliferation of cells, so that you can be sure that your assay works fine.

The poor reproducibility can also be related to the assay protocol - which assay do you use? Does it require taking several steps (e.g., washing, resolubilization, etc.) before you can get your results? Generally, the more steps there are, the larger are errors.

Furthermore, poor reproducibility within one measurement plate might indicate that cells seeded into different wells do not in fact reside in equal conditions. For instance, it is sometimes a good idea to leave the outer wells of the plate empty (occassionally, I just put PBS or cell culture medium there, which prevents the neighbouring wells from drying out). It will also protect you from some optical artefacts arising from taking readings close to the edges of the plate.

I would suggest to try different incubation durations: 24 h, 36 h, 72 h and 96 h. Some compounds depending in their mechanism of action, possess the effect on cell viability only after specific time point (eg. some antimetabolites are not effective after 24 h, but they show a strong effect after 72-96 hours). Also, we always make the same dilution of compound in triplicate (technical repetition), as you could make some technical errors during plating the cells into wells or making compound dilutions, etc. We do not change the medium, as you never know how much compound already penetrated into cells, how much was left outside. If you add the new solution of your compound, you may slightly change the conditions. Cell number depends a lot on the cell line and the duration of experiment, as some cells divide very fast, some are very slow. I suppose, 2000-3000 could be OK, but to be sure, you have to check how your negative control (cells without treatment) look like on the last day of experiment, just before measurement cell viability. They should not be overgrown. Also, it is very important to maintain the same concentration of your solvent in all wells, especially if solvent of your tested compound is DMSO, as it also affects cell viability.

@Darja Lavogina

Yes, 96 well plates. The assay is cyquant direct cell proliferation. I just need to prepare the reagent, add 100 ul to each well and incubate at 37°C for one hour. That´s pretty much it. No washes or any other additional steps. Considering this i guess that my most important source of variation should be the seeding step, although i always try to avoid those wells at the edges of the plate and i do fill them with PBS to avoid as far as possibe the effect of evaporation of the media in wells that have cells.

@ Vilma Petrikaite

I´m currently incubating the cells with the treatment and desing the experiment in such way that i could be able to read at least at 48, 72 and 96 hours. The triplicates i always include them.

In regard to the solvent, i always include a control for the solvent and all the dilutions have the same concentration of it (0.3 %). I always check that the DMSO cotrol behaves as the cells with no treatment.

@Andrea Magionni

Thanks for the information. Actually i´searching for a cheaper alternative to the kit that i´m currently using (cyquant direct cell proliferation), since i will be carrying out this type of assays very frequently. I´m intrested in kits or techniques that would allow me to determine the growth inhibition by DNA content without killing the cells.

Hi Mauricio,

Do you synchronize your cells by removing FBS for 24h before adding it back in with the treatment media?

I've been doing antiproliferation assays using breast cancer cells and noticed that I got good results when I used 5000 cells/well but not with 2500 or 10k cells/well. Using lymphoma cells I found just the opposite and settled on 20k/well.

I use TNFa as the positive control which seems to inhibit proliferation very consistently by about 40-50%.

Finally, I have to agree with you that the cell proliferation methods in the literature are generally not detailed enough to be able to repeat the experiments. I'm hoping that this will change now that Richard Harris' book 'Rogor Mortis' is shining a light on the fact that methods are written poorly and that much of the published research remains irreproducible.

@ Kerstien Padgett

To be honest I haven´t read anything related to synchronization of cells prior to treatment in dose/response assays. I´ll check it further to see how this could help me getting better results.

Isn´t 10k cells/well (96 well plate) too much for growth inhibition assays? ...considering that one of the cell lines that i am using is MCF7 and they got confluent very fast. I´ll definitely make some tests seeding 5k cells/well, since maybe having few cells/ml will affect the homogenization of the sample and thus the variation of cells/well.

Exactly. That´s why i´m struggling with the design of these assays, mostly due to the lack of detailed information on how the experiments are carried out.

Thanks for your help!

Any suggestions on RPMI 8 cell line seeding?