Reporter gene assays guide — Part 8: Controls
Experimental Controls
Controls are the parts of a reporter assay that get the least attention in published figures but determine whether the figure means anything at all. A well-designed experiment has more control wells than experimental wells; a poorly-designed one has the opposite, and the data from the latter usually cannot be interpreted regardless of how clean the numbers look. The categories below are not optional. Each one addresses a specific failure mode that you will encounter at some point.
Negative controls
Untransfected cells. The baseline luminescence or fluorescence of cells without any reporter plasmid. This tells you the background signal of the plate reader, the substrate, and the cells themselves. For luminescent reporters, this is usually very low (less than 0.1% of a typical firefly signal), but for fluorescent proteins, the autofluorescence of cells and media can be substantial, particularly in the green and red channels with standard media containing phenol red.
Promoterless reporter. The reporter construct without a promoter driving it. This measures the basal transcription from the plasmid backbone, cryptic transcription start sites in the vector, and any contribution from the bacterial plasmid DNA that may be present after transfection. A well-designed vector should give close to background levels in this configuration. If your promoterless construct gives 10% of your induced signal, you have a vector problem.
Empty vector. A plasmid that is identical to your reporter construct except it does not contain the reporter gene. This controls for any effect of the vector itself (the selection marker, the bacterial elements, the polyA signal) on the cells. If your empty vector treatment gives different cell viability or signalling than untransfected cells, that is important to know.
Vehicle / solvent control. For drug treatments, the appropriate vehicle (DMSO, ethanol, water) at the same dilution as the highest drug concentration. This controls for any effect of the solvent on the cells and on the assay. DMSO above 0.5% v/v is toxic to most mammalian cells; above 0.1% you can start to see effects on signalling.
Positive controls
Known agonist / activator. A treatment that you know will activate the pathway or promoter of interest. This is essential for confirming that your assay is working: if your positive control does not give the expected induction, the experiment is not valid regardless of what the test conditions show. Common positive controls include:
For NF-κB reporters: TNFα (10 ng/mL, 4 hours)
For p53 reporters: nutlin-3a (10 μM, 24 hours) or etoposide
For CRE reporters: forskolin (10 μM, 4 hours) or IBMX
For hypoxia reporters: deferoxamine (100 μM, 16 hours) or 1% O₂
For estrogen receptor reporters: β-estradiol (10 nM, 24 hours)
For glucocorticoid receptor reporters: dexamethasone (100 nM, 24 hours)
The specific conditions vary by cell type and promoter, but the principle is universal: have a known positive control in every experiment.
Reference compound. When running a screen or testing a series of analogues, include a reference compound with known activity in every batch. This lets you normalise across batches and detect drift in assay performance over time. For a screen, this is usually a strong positive control run on every plate.
Standard curve. A dilution series of a known activator spanning the full range of expected response. This lets you convert raw reporter signal to a meaningful biological quantity (e.g. "equivalent to 5 nM estradiol") and verifies that the assay is responding in the expected dose-dependent way.
Specificity controls
Promoter-mutant controls. If you are testing a transcription factor's effect on a promoter, include a version of the promoter with the binding site mutated. The transcription factor should not activate the mutant. If it does, the activation is not specific to the binding site.
Pathway inhibitor controls. If you are studying a signalling pathway, include an inhibitor of the pathway in addition to your agonist. The inhibitor should block the agonist's effect. This confirms that the reporter is responding through the expected pathway rather than a parallel one.
Cell-line controls. When studying a tissue-specific response, compare your cell line of interest with a control cell line that lacks the relevant machinery. If the response is absent in the control line, the response is genuinely tissue-specific. If it is present in both, the response is more general than you thought.
Time-course controls. Reporter responses are time-dependent, and a single time point can be misleading. A time course with at least 4 time points (typically 0, 1, 4, 8, 24 hours) is the minimum for characterising a dynamic response. Include the uninduced control at each time point to control for changes in baseline over time.
HTS-controls: If you are performing a high throughput screen using an enzymatic reporter, check that any hit compounds to not directly effect the reporter enzyme reaction. For example, take some lysate or media containing nano-luciferase, divide it into two equal aliquots and spike in the hit compound to one and assay. If the reporter activity changes it is likely that the compound is poisoning the reporter reaction and you are not picking up true transcriptional changes.
Assay-quality controls
Z' quality control. A positive and negative control well on every plate, used to calculate Z' and track assay performance over time. Z' below 0.5 should trigger a re-optimisation or a repeat of the plate. Tracking Z' over time is the single best way to catch gradual assay degradation: old substrate, dying cells, contamination, expired reagents all show up in Z' before they show up in your test compounds.
Edge-effect controls. The outer wells of multi-well plates behave differently from interior wells: they evaporate faster, experience different gas exchange, and are more affected by temperature gradients during incubation. Either fill the edge wells with media only and exclude them from analysis, or use specialised plates with moats and barriers. Document which approach you use.
Plate-position controls. The signal in a 96-well or 384-well plate can vary by row and column due to pipetting order, plate reader path, and incubation gradients. For HTS, randomised plate layouts and random well assignment are standard. For lower-throughput work, alternate the position of conditions across replicates.
Replicates and biological independence
Technical replicates. Multiple wells of the same biological sample. These control for pipetting and well-to-well variation. Three to five technical replicates are typical for plate-based assays. For very high-quality data, more is better, but with diminishing returns above 5.
Biological replicates. Independent transfections, independent cell preparations, or independent experiments on different days. These control for the major sources of biological variation: transfection efficiency, cell state, reagent freshness, and the inevitable small differences in protocol execution. A minimum of three biological replicates is the standard for most publications.
Independent experimental confirmation. Where possible, confirm key findings with an orthogonal method: a different reporter, a different cell line, or a different readout (e.g. mRNA, protein, functional assay). This is the most stringent control and the one most often skipped, but it is also the one that catches the largest class of artefacts.
Controls for specific artefacts
Mycoplasma testing. Mycoplasma contamination can dramatically alter reporter activity, often in unpredictable ways. Test all cell lines regularly (every 1 to 3 months) and discard contaminated lines. The effect on reporter assays is large and well-documented.
Plasmid DNA quality. Reporter assays are sensitive to the quality of the plasmid DNA. Endotoxin contamination, nicked DNA, and RNA contamination all reduce transfection efficiency and signal. Use a reliable endotoxin-free midi- or maxi-prep kit and verify the DNA quality by gel electrophoresis before transfecting.
Cell passage number. Cells behave differently at low passage versus high passage, and the difference shows up in reporter assays. Use cells within a defined passage window (typically passage 5 to 25 after thaw) and record the passage number in your methods. Do not compare data from passage 8 cells to data from passage 30 cells without testing the assumption.
Lot-to-lot variation. Fetal bovine serum, trypsin, and other media components vary between lots. For long projects, test multiple lots and reserve a sufficient quantity of a single lot for the duration of the project. This is tedious but it is the difference between reproducible and irreproducible data over months of work.