This post was contributed by guest blogger, Scott Messenger, Technical Support Scientist 2 at Promega Corporation.
It’s always an exciting time in the lab when you find a new assay to answer an important research question. Once you get your hands on the assay, it is always good to confirm it will work for your experimental setup. Repeating the control experiment shown in the technical manual is a great way to test the assay in your hands.
After running that first experiment of your assay, it looks pretty good. The trends of control and treatment are consistent. Time to get on with the experiments…but wait—the RLUs (Relative Light Units) are two orders of magnitude lower than the example data! I can’t show this data to my colleagues; it doesn’t match. What did I do wrong?
This is a concern that we in Technical Services hear frequently. The concern is real, and I had this same thought when doing some of my first experiments using luminescence. When a question like this comes in, a Technical Service Scientist will make sure the experiment was performed as we described, and in most cases it is. We then start talking about RLUs (Relative Light Units).
What Is a Relative Value?
You have likely heard of RLUs or RFU (Relative Fluorescent Units) during your time in the lab. If you were like me, you didn’t really know what they were but, you also assumed that it is science so, RLUs and RFUs must have some standard value that is highly defined. Not so for RLUs and RFUs—they are not standardized, and for this reason they should not be compared across instruments.
When you are measuring output from luminescent or fluorescent assays, you are essentially measuring the number of photons emitted. It turns out accurately quantifying photons is hard. Our GloMax® instruments, as well as most high-quality luminometers, contain a photomultiplier tube (PMT) that amplifies the signal so it can be measured. Very simply, when photons strike the detection window, electrons are produced which generate an electrical signal that gets amplified. The more photons, the more electrical signal. The software in the instrument then attempts to quantify the electrical signal and reports a numerical value.
As you can imagine, there are different methods that may convert the electrical signal to numerical values. Since the numbers are calculated in proportion to the electrical signal, they are relative to the amount of electrical signal produced: the relative in the RLU.
Some instruments may use numbers from 0-10,000,000,000 while others may use numbers from 0.1 to 10,000. Moreover, since the PMT relies on electrons to be produced in response to a photon and the signal is amplified, even the same model of an instrument can give different RLU values for the same sample. Therefore, you should never compare raw RLUs.
If It’s All Relative, How Do I Evaluate the Results?
So just how do you know if the assay worked? The best way to confirm this would be to perform the control experiment typically described in the technical manual including all the positive, negative, background and any other recommended controls.
The background control (e.g., medium or buffer alone) raw RLU values can be subtracted from all of the sample values for better accuracy. Subtracting the background control allows you to see how a true signal for the positive control compares to just background noise. The background subtracted raw RLU values can then be normalized by calculating a fold change compared to the negative (untreated) control, ideally on the same plate. This normalized ratio or percentage is unitless and can now be used for comparisons. If you have a robust assay, the fold change should be comparable across different plates, experiments, analysts and instruments. For dose response assays, you can also calculate and compare the half-maximal eﬀective/inhibitory concentrations (EC50/IC50) for full response curves. For genetic reporter assays using transient transfections, we also recommend including a second internal control vector with a secondary reporter to help account for transfection efficiency and other sources of variability.
If you are doing something completely new that doesn’t have a known treatment and response, one of the best things to do would be to run a standard curve with a known purified material. If available, you can include a serial dilution of a standard and use the raw RLU values to interpolate the concentration of the analyte in the samples (ideally include a standard on each plate). If it is a cell-based assay, the alternative to this is to run a cell number titration. Both control experiments will give a range of RLUs that you can then use to determine the sensitivity, dynamic range and find the linear portion of the assay in order optimize your experimental conditions. These calculations all rely on internally controlled data that assess the change in RLUs compared to a control rather than using the raw RLU values. Luminescent assays offer superior sensitivity with low background and a wide dynamic range. Just remember, with RLUs it’s all relative, so you need the controls to gain some meaningful perspective.
Here are some helpful tools and examples for calculating a fold change and analyzing cell health or reporter assay data:
- ApoTox-Glo™ Triplex Assay Data Analysis Worksheet
- Using input data from viability, cytotoxicity and apoptosis assays, the worksheet subtracts background, calculates averages and standard deviations of replicates and calculates the fold change from untreated cells (UTC). This worksheet can be used for other cell health assay data if the same plate layout is followed.
- Designing a Bioluminescent Reporter Assay: Normalization Options
- Optimizing control-reporter experiments and analyzing dual-reporter assay data.