OD 630nm is considered as noisy signal. In my opinion, OD 570nm is enough

This is what I came up with:

http://www.moleculardevices.com/Documents/general-documents/mkt-appnotes/microplate-appnotes/Maxline_AppNote_05_MTT.pdf

It looks like the 630 reference wavelength is to negate the effect of precipitated proteins or cellular debris that may interfere with the reading. If there is particulate matter in the path of the light, it will affect absorbance in all channels by scattering. You normally subtract the reference wavelength from the test wavelength to cancel out this effect.

As suggested by Jason above, a reference wave length is always used in colorimetric measurements to account for the chromogenic reaction-independent well to well variations. Its not MTT specific. Small differences in plastic, medium components, precipitation, bubbles etc etc could be normalized

We at MatTek recommend to use reference filter at 650, since 630 still cuts away significant part of the MTT spectrum. One should be however very careful in using reference filters, when the test article tested is coloured and has its absorbance spectrum in the spectrum of MTT and ref. filter.

In my experience it was only necessary to read at 2 wavelengths if you leave the cell culture medium in the well during the experiment and add the MTT solution to this - then the medium (phenol red?) interferes with the absorbance of formazan so you need to subtract the background.

Whether or not you leave the culture medium in of course depends on how long your incubation with test compound is & your experimental design- could you take out the medium and replace it with warm (37o) Krebs solution on a hot plate for a short test compound incubation time e.g. 1hr, for example?

Or could you do the test compound incubation in sterile serum free culture medium, e.g. for 24 hrs, then remove this, rinse gently in Krebs/ PBS and add MTT dissolved in Krebs solution or PBS (still at 37 degrees) for 1-2 hr. This would remove the need to read absorbance at 630nm as well as 570m.

Look up the absorbance for phenol red (in medium) and MTT and this should explain it for your method write-up...or other articles that have done a similar thing. This was my experience during my PhD, I hope it helps you out. Good luck!

P.S. Make sure you have controls:

A) without MTT , with cells, & in solubiliser (DMSO or propanolol) to subtract the background (cellular debris) absorbance;

B) without cells, with solubiliser and if necessary MTT solution - If you haven't removed the remaining MTT (from all wells prior to DMSO addiition) before reading, this will still be yellow;

C) 100% (no damage) control (with cells, with MTT, with solubiliser, without test compound)

if you think the plastic has an effect, you could read this independently but with control (A) I didn't find that it was necessary.

Sorry if this is a bit obvious/repetitive! Good luck again!

Abi

We usually use two wavelength in flourimetry method extinction and emission

OD 630nm is reference to see the base line. When you have background signals from solution, some debris, unclear solution, scratches on/in your wells, subtraction (A570-A630) would help to get more exact results.

You can also have reference on your plate with extra wells having all solution without cells. So you can measure the base line at 570nm and subtract it from readings of your samples. It doesn't include individual condition, but it would work in most cases. So see how your results vary using extra wells without cells (usually it works well because variation of each well is not that large). If you get large variation, you might have to consider someway to eliminate backgrounds of individual wells.

This is not fluorometry, it is absorbance spectrophotometry. The reference wavelength is used to correct for optical variation and is unrelated to the actual chromophore. It allows individual corrections to made for each experimental well, due to say a scratch or a thumbprint in the optical path that would change the apparent background absorbance value of that well relative to the unmarred wells next to it. Establishing background values in separate control wells does not make sense unless you are doing an overall statistical error analysis. Many workers simply ignore reference wavelengths and go with the assumption that the wells are uniform, which is OK if you do enough replicates in your experiments (and the wells are not obviously different). Optical errors are then simply subsumed into the experimental variance. If you use a reference wavelength two concerns are relevant: the spectrophotometer must have equal sensitivity at the two wavelengths (usually true in good quality specs) and (as was pointed out above) you have to be outside the absorption curve for the chromophore (it was suggested that 650 nm is required to achieve this). The curve should be available somewhere online and you can judge for yourself. All this applies to cuvette methods also.

The reference value is related to the different adsorbance of plastics from each wells. you can perform the test using only the "measurement" value but if you have a low number of cells (low seeding density or high citotoxic effects) the influence of plastics can be significant.

I can suggest you the MTS assay (for example by promega)... It's a tetrazolium salt like MTT but it is soluble in medium... so you don't have to dissolute formazan with DMSO. this test is quickier (2 hours of incubation can be enough), requires less handling and it is more accurate. Moreover it is not a lytic assay so you can use cells for other techniques (such as proteomic or nucleic acid extractions). Regarding money i think MTS costs is comparable with the MTT

MTT is a multi-well measurement and, even, is repeated 3-5 times. Therefore, in my opinion, is not needed to subtract the background because 'background' is present in all of the measurements including controls. However the amount of solubilised formazan must be so adjusted to be in a good range (0.2-0.7 OD) of measurement by a plate reader. Too high or too low absorbance is not good even after subtracting 'background'.

I need to mention that using two wavelengths in colorimetry, not required always, is not equal to that in fluorometry where emission fluorescence can be measured only after exciting that fluorophore. In colorimetry, therefore, 'background' subtraction can take place some times but in fluorometry there is no chance of such background subtraction.

Thank you so much for all these valuable info, which I've never think about or read somewhere. I'm so glad, thank you everyone.

You don't have to read your samples at two different wavelengths. These are my papers you can check . I read my samples at 570nm.

N. T. Laçin et al. / Journal of Biomaterials Science 23 (2012) 577–592.

TIMP-2 gene transfer by positively charged PEG-lated monosized polycationic carrier to smooth muscle cellsJ Nanopart Res (2012) 14:694.DOI 10.1007/s11051-011-0694-3

I agree with Mohd Javed Akhtar. i think You don't have to read your samples at two different wavelengths. just read the samples at 570nm is ok. it is not needed to subtract the background, because 'background' is present in all of the measurements including controls. but the most important thing is that the amount of solubilised formazan must be adjusted to be in a good range (0.2-0.7 OD) of measurement by a plate reader, because too high or too low absorbance is not good even after subtracting 'background'.

630 nm measurement for the blank, 570 nm for the MTT reaction. Plastic and media can differ from assay to assay. If you have measured several experiments you can better compare the 570 nm data between the experiments.

I have a follow up question to this...I will occasionally get negative values for my background wavelength. Does anyone know what this means and/or why they would be negative?

Ya it is good to take reading at two wavelengths. 630nm is refrence seeting to get the actual results.

You do not need two wavelengths, unless your OD is too high so that it will be better to use another wavelength. What most people forget is to use a reading at the start of the experiments (d=0), which is the background reading of your cells. When the final reading is below this background you really have cell kill. The MTT assay is often referred to as a cell survival assay, which is often not correct since the majority of scientists just look at growh inhibition. See the original NCI protocol for details on this. http://dtp.nci.nih.gov/branches/btb/ivclsp.html

Why are my MTT assay results low ? The numeric values are always less than 0.5.  I choose 595 nm as detection wavelength.

Dear Wei

May be you cell count is low. You can seed more cells in each well..I suggest, at first you try to setup number of cells for the period time of treatment. For example if you are going to study the effect for 3 days , you should seed soe different cells from 5000- 12000 cekks/well and then at day 3 you choose which well has 60-8o-% cell conflency for next MTT assay.

Be succeed

Dear Noor

Reference wavelength in colorimetry is a correction wavelength for different reasons, for example say for well to well variation, treatment well to control well variation etc. For MTT assay, reference wavelength is usually taken at 630nm. But for instance if you dont have a filter of 630nm then you no need to worry, you can escape this step of taking reference wavength unless and until your work is for a research purpose only and not in a professional sector where you face audits visits by the professionals. Not taking the reference wavelength is absolutely fine since you are not taking the reference wavength for any well in the plate. 

The reference absorbance at greater than 650 nm in the MTT assay and at 630 nm - 690 nm  in the XTT assay is used to correct for nonspecific background values. Nonspecific readings include well plate variability, fingerprints etc. If the reader is capable of a measuring the absorbance at 630 nm to 690 nm, it is recommended that you take this reading. A properly functioning plate reader should give a value of 0 for the wells of a nondefective plate.

Accidental fingerprints on a 96-well plate can be avoided by holding it on the long sides between your thumb and fingers. Multi-well plates like the 96-well plate should also always be checked by visual examination for any possible defect before use.

If the plate reader returns a value greater than 0 for an absorbance reading at 630 nm - 690 nm, you should check that your plate reader is functioning properly.

If the microtiter plate or ELISA reader lacks 630-690 nm wavelength, you may skip the requirement for a reference wavelength without a significant change in the accuracy of the assay.

https://www.atcc.org/Global/FAQs/9/A/630nm%20-%20690nm%20reference%20absorbance%20readings%20for%20MTT%20amd%20XTT%20assays-995.aspx

Hi

What is the difference between the absorbance of blank well and positive control ( untreated cells)

Background is different in all wells not because of fingerprints or defects but because of different amount of cells. After dissolving in DMSO, a white plaque remains in wells. In control wells, it can be uniform, but in experimental wells it can cover the bottom in random way. Most of the cell structures scatter light at near IR which is used in some cell counters. That is why you need to eliminate this difference. From my experience on Tecan Spark M10, the difference with/without reference at 690 nm can be up to 10% viability.