Flow Cytometry Readouts: Yes, No, and Everything in Between

By Guest Blogger

Guest Blogger August 20, 2024

Now that you know how to read flow plots and have designed your first flow panel, you’ll load your samples into the cytometer and see one of two results for your antibody of interest: two clear populations or a huge smear across your FSH vs reporter plot. In this post, I’ll walk you through how to interpret the antibody readout on a flow plot.

Note that for the flow plots discussed in this post, voltage has been adjusted appropriately and compensation performed. Before you read out your plots, you’ll want to make sure you’ve also adjusted voltage and compensated your colors.

Here are some good practice tips on adjusting voltage: During data acquisition, make sure that the signal for the positive population is no higher than 10⁵ signal intensity. Signals any higher than this may not be clearly picked up by the sensors. You can adjust this by changing the voltage for the particular fluorophore prior to recording data. Increasing the voltage will shift the population to the right while decreasing the voltage will shift it to the left. If your signal is too high, lower the voltage to have your positive population between 10^3.5–10^4.5 for a cleaner readout.

A yes-no plot

If you are using an antibody to mark the presence or absence of a protein, a yes-no plot is the best-case result. For example, you have introduced a GFP protein into your cells (visualized via the FITC channel) and you are trying to see what percentage of total live cells are GFP+. In this plot, with adjusted voltage, cells that did not receive the GFP protein are most likely to be clustered at signals below 10³, while the GFP+ population cluster would be shifted to the right in comparison, say somewhere between a 10⁴–10⁵ signal intensity.

If your samples and controls were processed the same way, the GFP- population should overlap with the negative control population (Figure 1). Based on where the GFP- population ends, you can draw a gate which selects anything to the right of it, marking it as GFP+. Quantification for such graphs could simply be the percentage of the total cells present in the gate.

Two flow plots of GFP (x-axis) vs. FSH. On the control flow plot, a single population is visible at 10^4 on the x-axis, with an empty gate to the right of it. On the sample "GFP Introduced" plot, a negative population is in the same place as the control plot, with a positive population in the gate to the right of it. Though the edges of the populations touch each other, they are clearly distinguishable as two populations.

Figure 1: Example of yes-no flow plot of the (negative) control (left) or with the sample (right). Negative and positive samples are clearly distinguishable as two separate populations. Created with biorender.com.

Pro tip! If your samples and control underwent different treatment processes, you may see a shift in the negative population, so it no longer overlaps with the control.

In tricky experiments, these populations may overlap slightly, or the distinct populations may not be as obvious in a dot plot. If that is the case for you, switching to a contour plot should indicate where the two populations diverge.

The smear plot!

The dreaded smear - fear not! When you are looking to study the level of protein expression, instead of its presence or absence, the plots can look quite different. You might have high expression, low expression, and everything in between (Figure 2). These smears tend to occur when you are looking at proteins that are expressed regularly in cells and might be upregulated or downregulated based on your experiment. You can see from Figure 2 that determining where to place your gate based on the dot plot would be complicated.

Two flow cytometry dot plots, each PE-Texas Red (x-axis) vs PerCP-Cy5. (A) shows a large population covering the area between (0,0) and (10^5, 10^3). (B) shows the same population, but with a gate around a population from (10^4, 10^3 all the way to the right hand edge of the plot.

Figure 2: (A) a smeary dot plot showing a range of antibody readouts. (B) A gate on the smeary dot plot shows the population of interest. Created with biorender.com.

In such cases, instead of trying to quantify based on percentage positive or negative in your dot plot, looking at a histogram vs antibody plot would be better. The histogram shows peaks of high signal densities, so signal shifts due to upregulation or downregulation are easier to identify by comparing peak positions to the base expression. When compared to the control — which should be base expression — a peak to the right indicates a higher signal, or upregulation, and a peak to the left indicates a lower signal, or downregulation. Figure 3 shows a histogram for a yes-no dot plot, like the one shown in Figure 1.

A flow cytometry dot plot of CD3-PE (x-axis) vs count. There is blue peak at (0, 900) and a red double peak at (10^3.5, 900) and (10^5, 600). Though the shoulders of the peaks overlap slightly at 10^3 on the x axis, they are clearly distinct peaks.

Figure 3: An example of a histogram plot for a yes/no flow cytometry readout. Created with biorender.com.

Using histogram plots to quantify smeary plots

For smeary expressions, you can quantify shifts in signal using the mean fluorescence intensity (MFI) of the different populations. MFI is the average brightness from all cells that are positive for the marker of interest. MFI would be higher for a signal peak that shifts to the right on a histogram plot and lower for a signal peak that shifts to the left. When using MFI, you can still continue gating populations of interest with normal gates, just based on a histogram plot instead of a dot plot. However, the histogram plots look a bit different than they do for yes-no readouts.

A graph four horizontal lines dividing the Y-axis into quarters. Each quarter has a peak corresponding to a specific experimental condition. The bottom peak is at 10^4 (on the x-axis); the peak above that is at 10^3.5; the peak above that, which represents the control/baseline expression, is at 10^3.75, and the top peak is at 10^4.

Figure 4: A histogram plot for a smeary antibody readout. The orange peis the baseline expression of the protein of interest. Created with biorender.com.

The plot above is for a smeary expression and is quite different from the yes-no histogram plot shown previously. Here, the orange peak is a baseline expression of a protein while the green (top) has an upregulated expression (shift to the right). The blue (third from top) could be a downregulation of signal, though it’s hard to call from this plot. The red (bottom) peak overlaps with the baseline suggesting this sample has a comparable expression to the baseline/negative control.

Subscribe to the Addgene blog

You can see how it is easier to visualize these four populations in the histogram plot in Figure 4 compared to the dot plot in Figure 2. Quantification using MFI makes it easier to identify and study these subtle shifts. Additionally, you can gate off the histograms above, allowing you to visualize your data as a dot plot.

To summarize, if you have clear positive and negative populations for your marker of interest, a dot plot would serve you well. Percentage of cells in a parent gate (all live cells for example) that are positive for your marker would be a good statistic to compare between samples. In cases where there is no clear distinction between the presence/absence of your marker but there are subtle shifts in expression, using a histogram plot and MFI values to characterize your samples may allow for better interpretation.

Priya Prathima Headshot Priyamvada Prathima has a research background in protein mutagenesis and cancer immunotherapies. She is currently a Research Assistant in Arlene Sharpe’s lab at Harvard Medical School.