1. The glucose uptake assay should be done for longer durations, say 24hrs or time points for every 6 hours (see my paper below).

2. 30uM is high..reviewers might question you on dose.

3. 50,000 vs 5000 cells can be brought to same density (50,000/ml vs 5000/100ul) so that you could nullify density issue.

https://www.researchgate.net/publication/292615481_Mitochondrial_oligomers_boost_glycolysis_in_cancer_stem_cells_to_facilitate_blebbishield-mediated_transformation_after_apoptosis

Article Mitochondrial oligomers boost glycolysis in cancer stem cell...

Thanks! But you think that there is a difference because of density and treatment time. I mean 15 minute and 24 hour is a big gap but I just want to know if there could be some secondary differences as well.

You should have tested both assays at the same time for the comparison to be valid. You do have to remember that apoptosis/necrosis caused the activity of you compound might not occur in that first 15 minutes but after 24 hours. 30uM is actually quite high unless you have your positive control drug and your therapeutic window/ selectivity index ratio is high when compared to normal cell. It would be better if your compound has IC50 in the nanoMolar range. For the difference in cell number, I am not really sure about how big is the effect. Need to wait for other expert to comment on this.

You may assess the MTT activity by replacing the Glucose with Galactose to see the inhibition. There is so called Crab effect. Galactose as carbon source would increase the mitochondrial toxicity.

There is a very interesting article, may help you.

Marroquin et al. (2007) Circumventing the Crabtree Effect: Replacing Media Glucose

with Galactose Increases Susceptibility of HepG2 Cells to Mitochondrial Toxicants. TOXICOLOGICAL SCIENCES 97(2), 539–547.

1. For MTT assay, 24 hour exposure is rather short. Many active compounds show inferior potency at this timepoint. I suggest MTT assay time 48 h or even 72 h at which tested compounds at 30 uM will likely show more than 50% inhibition, which was reached  at 24 h assay timepoint.

2. The number of cells per well in MTT assay should not be taken as fixed (e.g. 5000 cells/well for any cell line/condition), but it should be determined empirically for each cell line and each assay time to ensure strong assay signal and assay linearity. In addition, cells are usually more sensitive in exponential growth phase, so growth curves should be determined (for different numbers of loaded cells and specific  assay time) to ensure that cells do not reach stationary phase too early during drug treatment. 

3. Antiproliferative IC50 for glucose uptake inhibitors in mid-uM range is plausible to me. For instance, another GLUT inhibitor quercetin showed average IC50 value across 60 cancer cell lines ~48 uM at 48 hr timepoint. Likewise, WBZ117 showed 41% inhibition of H1299 lung cancer cell growth rate at 10 uM concentration when the inhibition of glucose uptake was  93%.

4. Inhibition of glucose uptake and inhibition of cell proliferation by MTT assay do not have to show simultaneously. Even if the inhibition of GLUT is apparent after 15 minutes, much more time is needed to show apparent inhibition of cell proliferation when cell numbers (MTT signals) are compared for untreated and drug-treated cells. The more important question is for how long the inhibition of GLUT lasts when the cells are constantly exposed to the inhibitor in cell culture medium.

You are measuring two different things. MTT measures how many cells are left after a treatment (cell viability), so if your glucose uptake inhibition cause cell death, this may be apparent at 24 , 36 or X hours. Thus you need to do an MTT time course to find out if the cells are affected in this manner and when.

Since the IC50 value looks credible for glucose uptake inhibitors,  it is conceptually not necessary to change the assay time, and the only reason to do it would be pleasing  reviewers with appealingly lower IC50. In fact, the IC50 of about 20 uM is not that bad, and not all anticancer drugs have IC50 values in nanomolar ranges. For instance, clinically used carboplatin has an average IC50 ~ 100 uM at 48 hrs across panel of 114 cancer cell lines.

Concerning assay times: Longer assay times usually increase apparent antiproliferative effect of drugs; however, this often merely reflects increasing number of cell divisions in undisturbed control cells, which optically enhances the difference between treated and control cells. Assay times longer than 72 hrs may be needed when using cells that proliferate too slowly (e.g. cancer stem cells), or when testing drugs with low solubility in the medium, but in these cases spent medium with drugs or solvent control needs to be replaced with fresh medium. This may add variability to the assay, and so 72+hr assay times would be counterproductive, unless used for these specific reasons.  

On the other hand, 24 hrs assay time is usually too short and one may miss active compounds, not necessarily because treated cells are not affected by drugs before 24 hrs (cells may be committed to cell cycle arrest or cell death), but because the control cells did not have enough time to proliferate and show meaningful MTT signal difference with treated cells. 

The NCI uses in its screening 48-hr time point, because they found it more likely than 24 hrs to detect active compounds, and at this time point, the assay is less likely to show non drug-related differences that only reflect proliferation rate of the used cell line. Thus, 48 hrs or 72 hrs are good assay times. If something happens to the treated cells earlier, e.g. within 36 hrs, this effect will be reflected in 48 or 72 hrs end points due to the cumulative nature of the assay. For these reasons, I would give lower priority to the time course study in this case. It is more important to make sure that the assay is in linear range (signal linearly dependent on the number of cells) and that the cell culture is in exponential stage of growth for most of the assay time, which needs to be assured during the assay optimization step.

@ Roman: Thank you for your answer. But, we have observed that A549 and H1299 cells nearly double during a course of 24 hours. We actually work like this:

Day 1 : cell plating in 96 well plate 100ul (5000 cells/well)

Day 2: remove old media and treatment addition (100ul)

Day 3: Mtt reagent addition/ 4 hour incubation/ Absorbance measurement

@Pratik, yes density and treatment time greatly influence the availability of glucose in media. Just like you had mentioned above in one of your replies, cells can double and consume most of the glucose by 24 hours. So it is advisable to have controls for every time point you test. this will give idea about the availability of glucose in media at every time point, so that you'll know up to which time point you can rely on your results.