Written by Ryan Brinkman, Ph.D.

Flow cytometry is a powerful screening tool that can be used to help identify target candidates on phenotypically defined cell populations.

There are a myriad of possible assays that can be combined using flow cytometry to examine the effects on these cells. Not to mention the fact that millions of cells can be measured in one experiment, or the statistical power of flow cytometry.

Of course, for these assays to be robust and reproducible within an individual lab, adherence to standard operating procedures and current best practices in flow cytometry is required. New dyes, new protocols, and new reagents are being released on a regular basis and must be titrated, optimized, and validated to ensure high quality data is generated.

Methods sections in scientific papers are often unable to capture all the critical data necessary to accurately reproduce the results in another lab. This continues to exacerbate the concerns Begley and Ellis (and others) raise about reproducibility. Thus, there is growing need to develop methods for communicating more details about the design and execution of an experiment.

In fact, over the last several years, these concerns have surfaced as a critical discussion about the standardization and reproducibility of scientific data. Data that can not be reproduced wastes time, energy and resources.

Why Flow Cytometry Needs Standardization

In their commentary in Nature, Begley and Ellis discuss their experiences in attempting to reproduce ‘landmark’ findings and how only a very low percentage of these were able to be reproduced.

With the pressures to discover new drugs and treatments, this is a huge waste of effort by all concerned, including both academia and industry, and is one of the drivers of the increasing costs of drug discovery. These cost increases led to the National Institutes of Health developing policies to improve rigor and reproducibility in scientific research.

Collins and Tabak discussed this Nature commentary, as did the editor-in-chief of the journal, Science. Soon, as discussed here, new applications to the NIH will be required to address issues of rigor and reproducibility.

The NIH has focused on four areas of reproducibility:

1. The scientific premise of the proposed research
2. Rigorous experimental design for robust and unbiased results
3. Consideration of relevant biological variables
4. Authentication of key biological and/or chemical resources

As described below, Dr. Ryan Brinkman provides information on two specific ways in which flow cytometry researchers are effectively communicating the above information to the flow cytometry community to improve reproducibility and consistency.

These two ways include first, the use of the MIFlowCyt standard and second, sharing data using the Flow Repository. This information was recently covered in exceptional detail in a special online lecture that Dr. Ryan Brinkman, a distinguished scientist in the Terry Fox Laboratory at the British Columbia Cancer Agency, and professor of Medical Genetics at the University of British Columbia, delivered to the ExCyte Mastery Class. Enter Dr. Brinkman… 

The MIFlowCyt Standard And The FlowRepository

Flow cytometry (FCM) datasets that are currently being generated will be two orders of magnitude larger than any that exist today.

New flow cytometry instruments are available that increase the number of parameters measured for each single cell by to 30. The complexity of such datasets creates challenges in both annotation and data sharing. How do we solve this?

The rapidly expanding availability of FCM datasets through public repositories lays the foundation for researchers to integrate data from multiple research areas and diseases. This is leading to a culture where a large body of annotated and shareable data is available online to the broad biomedical research community. As a result, the development and use of data and metadata (‘data about the data’) standards are critical for achieving this goal.

An important step in curating large FCM datasets is knowing what information needs to be captured. Minimum information guidelines for reporting experiments has found broad-based support (see MIBBI) across biological and technological domains. For flow cytometry data, The Minimum Information about a Flow Cytometry Experiment (MIFlowCyt) effort is now an approved International Society for the Advancement of Cytometry (ISAC) standard and has been adopted by journals, including Cytometry A.

MIFlowCyt provides a checklist covering details including experimental overview, sample description, instrumentation, reagents, and data analysis. Almost all articles now published in Cytometry A follow this recommendation.

Data sharing is also widely recognized as critical by funders and journals including Nature, PLOS and NIH. The FlowRepository is primarily for sharing data associated with peer-reviewed publications annotated according to MIFlowCyt data annotation requirements.

The FlowRepository operates under the auspices of ISAC with guidance provided by ICCS and ESCCA. Together MIFlowCyt and FlowRepository provide a mechanism for researchers to access, review, download, deposit, annotate, share and analyze flow cytometry datasets. This article shares more on how to create a MIFlowCyt compliant manuscript using the FlowRepository.

Since methodology sections in peer-reviewed journal papers often fail to capture all the critical data necessary to accurately reproduce flow cytometry results, efforts have been taken to help flow cytometry researchers improve reproducibility and consistency. Two such efforts are the development and use of the use of the MIFlowCyt standard and second, sharing data using the FlowRepository.

To learn more about publishing your flow cytometry data, and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, Ph.D.

From instruments, to reagents, to software there are many flow cytometry innovations to pay attention to this year.

These are just a few highlights of the great things coming your way in 2016…

I. Instruments

The mid-range cytometry market is becoming crowded with a host of new instruments on the market.

Of those, we’ve had the chance to try the Novocyte, which is a very easy to use instrument built with the Accuri mindset—fixed voltage PMTs. This, coupled with easy to use software, makes it a breeze to get a researcher up and running.

We noticed some issues in the far red channel of the Novocyte, but we’ve been told this is being resolved.

I’m looking forward to testing out the redesigned Attune NxT instrument from ThermoFisher (previously Invitrogen). The acoustical focusing on this instrument makes it very intriguing for rare event analysis, as you can run faster with less spread of the data.

Moving to the higher end platforms—multidimensional data is very exciting to us. The new Helios (CyTOF III) system, from what we’ve read, is solid.

It looks like Fluidigm has been able to take what was learned from CyTOF I and CyTOF II to improve the system all around, including instrument stability, cell recovery, operator ease-of-use and more.

Of course, the trade-in price tag is a bit of a shock for those with the earlier generation instruments. We hope Fluidigm doesn’t stop supporting the earlier instruments, and we will miss the all-orange shell of the earlier CyTOF’s as well.

The Yeti is to be considered for those who want more detectors. The Yeti has 28 fluorescent detectors but maintains a small footprint.

Then there is also the BD X20 instrument, which allows for up to 20 fluorescent parameters, and (if you have the money)—there is the X50 instrument.

Finally, there is also the Sony SP6800. This spectral analyzer changes the way researchers examine the data, using spectral unmixing that allows for very closely related dyes to be distinguished.

II. Reagents

Thermo-Fisher recently acquired Affymetrix, bringing eBioscience into their family. We look forward to seeing how this acquisition affects the flow cytometry reagent market.

In terms of reagent trends, dyes are still very popular. More and more fluorescent tags are making choices for panel design easier. The range of “Brilliant” dyes continues to expand and offer new, brighter choices for the end-user.

Now, with Brilliant UV dyes, the UV laser can be used for more than just side populations and calcium flux experiments.

BioLegend has been expanding their product lineup. There is also a lot of excitement with the LEGENDPlex assay kits. Since these kits can be run on any instrument, they can be readily integrated into current workflows.

We enjoy using magnetic beads for depletion experiments before cell sorting, and we are happy to see BioLegend go-to-market with their new MojoSort product. We’ve found that BioLegend really has a good marketing team, especially when it comes to naming their new products (example, MitoSpy).

Another reagent trend to keep your eyes on this year includes the release of validated antibodies for MaxPar labeling for use on the CyTOF. If you are using a CyTOF, knowing which reagents have been validated will save you a lot of time and resources. 

III. Software

We enjoy watching and benefiting from the continued development of tools to assist in polychromatic panel design, including Chromocyte, Fluorish, and Flourofinder.

Each of these tools follow a similar path for designing the panel, and share some common features including:

•       Allowing you to design for your instrument
•       Searching vendor databases for reagents that fit your needs
•       Saving and exporting the panel to share

The first two bullet points above are the most important. After all, what good is building the best panel if you can’t run it on your instrument? With the above tools, you can build a solid antibody panel in minutes. Of course, you still have to apply your knowledge of polychromatic panel design and your instrument to build the right panel, but the hard part of searching is now easy.

In particular, Chromocyte provides a wealth of information beyond just providing researchers with a panel design tool. The website covers everything from meetings and training courses, to integrated search engines for various flow related products, to an online forum for finding more information and getting help.

When it comes to analytical software, there is a lot of continued development with the various data analysis software packages. FlowJo continues their parallel development of VX and Version 9. We are still fans of Version 9, especially with the added ability to compute the spillover spreading matrix of a flow cytometry experiment. The FlowJo SSM is another tool that can be used to evaluate and monitor the performance of polychromatic panels and instruments.

There are some interesting PC-only software packages as well. We’ve always liked the intuitive feel of FCS Express, and the add-ons that the FCS Express team offers, such as the ability to directly import experiments from DIVA. FCS Express has come a long way over the years, especially in terms of the development of their Image Analysis package. We’ve also heard positive reports of the Kaluza data analysis software package, which is also highly intuitive.

The team at Inivai continues to provide novel solutions with their Logic platform. From their lead analysis package FlowLogic, they have developed a series of other tools that make flow cytometry analysis, as well as graphing and annotating metadata and other information seamless.

What do you think of our list?

Feel free to tell us about your favorite (or least favorite) flow cytometry products on our popular product review page here.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, Ph.D.

Designing an antibody panel and running samples on a flow cytometer are not the only steps in a flow cytometry experiment.

After you run your experiment, you have to analyze the data. In particular, you need to perform statistical analyses of the data. This is especially true if you’re hoping to publish your data.

Once all the experiments are concluded and the preliminary analysis of the data performed, you must perform statistical analyses on the data to determine if there is significance in the data.

There are several different statistical tests that can be performed depending on the type of data and the comparisons being made. In the case of either making a comparison against a hypothetical mean, or comparison between two populations, the gold standard test is the Student’s T-Test.

What Is A Statistical T-Test?

The T-Test was developed by chemist William Sealy Gosset, who developed the test while working at the Guinness Brewery as a way to monitor the production of their most famous product.

Since he wasn’t allowed to publish his work directly, the paper was published under a pseudonym in the journal, Biometrika.

Before getting into the details of how the T-Test is performed and how the results are interpreted, there are several factors that need to be kept in mind…

The T-Test makes several assumptions about the data:

• The data is from a Gaussian distribution
• The data is continuous
• The sample is a random sample of the population
• The variance of the populations is equal (If not, there are variations on the theme to address this.) 

There are three major variations on the T-Test:

• One-sample T-Test – compares the mean of the experimental sample to a hypothetical mean.
• Unpaired T-Test – compares the mean of the control and experimental samples.
• Paired T-Test – compares the mean of two samples where the observations in one sample can be related to the observation in the second sample. (For example, the effects of treatment on patients where there is a before treatment and after treatment measurement.) 

The three pieces of information needed to perform a T-Test:

• The mean of both samples
• The standard deviation of both samples
• The number of observations

The T-Test compares the differences between the means of two populations to determine if the null hypothesis should be rejected. At a minimum, to perform the T-Test, one needs the means and standard deviations of both populations, and the number of measurements.

The researcher also needs to set the threshold value, also termed the α. We will compare this threshold to the P-value. If the P-value is greater than the α, there is no significance in the data. However, if the P-value is less than the α, there is significance in the data.

What Is A Null Hypothesis (H_O)?

Simply stated, this is a statement about the relationship of the above two populations.  Mathematically, this can be expressed as:

μ_A = μ_B

The null hypothesis makes the assumption that our experimental results are from random variation. If, during the statistical analysis, the data is sufficient to show that random variation is not a sufficient explanation for the data, the alternative hypothesis (H_A) must be accepted. 

A One-Tailed Versus A Two-Tailed T-Test

A T-Test can either be one-tailed or two-tailed. The above example would be an appropriate null hypothesis for a two-tailed T-Test—that is, when the investigators do not know if the treatment will cause an increase or decrease in the measurement. If the investigators expect the treatment will cause an increase OR a decrease, a one-tailed T-Test is more appropriate. 

How To Run A T-Test

In the following example, the researchers sought to determine if the percentage of CD4+ T-cells in patients who had Irumodic Syndrome was increased after treatment with Byphodine.

The percentage of CD4+ T-cells was measured on PBMCs before treatment and one week after treatment. Considering this information, this is how you would proceed to run a T-Test… 

1. Establish the null hypothesis.

“In patients with Irumodic Syndrome, treatment by Byphodine either decreased or caused no change in the percentage of CD4+ T-cells.”

In this case, since the researchers are not concerned if the treatment causes a decrease in the CD4+ cell, a one-tailed T-test will be performed, and can be written as: 

μ_A ≥ μ_B

2. Determine the alternate hypothesis.

“In patients with Irumodic Syndrome, treatment by Byphodine increases the percentage of CD4+ T-cells.” 

3. Establish the threshold.

By convention, the α is typically set to 0.05. This comes from work by R.A. Fisher who stated in his work Statistical Methods for Research Workers (13^th Edition): 

The value for which P=0.05, or 1 in 20, is 1.96 or nearly 2; it is convenient to take this point as a limit in judging whether a deviation ought to be considered significant or not. Deviations exceeding twice the standard deviation are thus formally regarded as significant.

There are cases where the threshold can be changed. Increasing the α makes it easier to show significance at the expense of committing a Type I statistical error (false positive). Decreasing the α makes it hard to show significance, and increases the chance of committing a Type II statistical error (false negative). Care must be taken, however, to ensure that the reason for the change is well-documented and spelled out.

For this example, we will set the α to 0.05. 

4. Collect the flow cytometry data.

Following all best practices, with a well-controlled instrument, all appropriate gating and reference controls used to generate the data table below. 

Table of %CD4+ PBMCs

Once this data is entered into our statistical analysis package of choice (we personally use Graphpad Prism), we can generate an appropriate graph: 

In the above case, the data is plotted, with the mean and standard deviation plotted.

When the one-way T-Test is calculated, the P-value is 0.0003, which is lower than the threshold. Therefore, the null hypothesis is rejected, and the alternate hypothesis is accepted. As a result, this data supports the conclusion.

The use of the T-Test makes the assumption that the data follows a normal distribution.  If this is not the case, there are non-parametric tests that will allow for the statistical analysis similar to the T-Test. These include the Wilcoxon test and the Mann-Whitney test. In non-parametric tests, the data is ranked according to the value (from lowest to highest), regardless of where the data comes from.

Non-parametric tests test the null hypothesis that the data is distributed at random, with the alternate hypothesis being that the data is not randomly distributed, but one population has larger values than the other.

The Student’s T-Test is an essential tool in the researcher’s toolkit to confirm that the data generated in the course of the investigation supports the hypothesis driving the research. Proper application of the T-Test (and related non-parametric tests) to determine statistical significance in the data will improve confidence in the conclusions of any published work. Following the steps outlined above will allow the researcher to correctly apply the proper statistical tool for their data.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Mike Kissner

Light scatter is unfortunately one of the more misunderstood concepts in cytometry.

When you first learn flow cytometry, you were likely told the misleading phrase, “Forward scatter signal intensity is proportional to cell size, and side scatter signal intensity is proportional to cell granularity.”

The short story here is that, yes, this can be the case when we’re dealing with typical cells (consider blood granulocytes, which are bigger than lymphocytes, have more intense FSC signals; granulocytes are more “granular” than lymphocytes and thus have more intense SSC signals). However, the truth is that light scatter complexity is belied by this perennial but unsatisfactory introductory explanation.

While particle size (particle radius) certainly does influence light scatter signal, its intensity is a function of a combination of factors, including:

•   wavelength of laser illumination
•   collection angle, and
•   refractive index of the particle and flow medium (sheath) 

It turns out that light scatter intensity has a strong dependence on the relationship between the size of the particle to the wavelength of the laser. More specifically, particles with diameters that are larger than the wavelength of the laser will scatter light with a different pattern than particles that are smaller than the wavelength of the laser.

Keep in mind that we most typically measure scatter using 488 nm excitation (and sometimes 405 nm excitation), so particles or cells with diameters larger than about 0.5 microns will behave differently than particles with diameters significantly smaller than 0.5 microns.

It turns out that there is a physical theory, named after the German physicist Gustav Mie (1869-1957), that predicts and explains the behavior of light scattering of particles larger than the wavelength of illumination. Essentially, Mie’s theory predicts that the intensity of scattered light has a strong angular dependence.

In other words, the intensity of the signal generated from scattered light depends on the angle at which we collect and direct the light towards a detector. Mie’s theory explains why we use a forward scatter (FSC) detector to measure the light scatter signal of typical mammalian cells, which are much bigger than the wavelength of the illumination source (typically the 488 nm laser).

How Small Particles Affect Forward Scatter

Forward scatter detectors collect light at small angles relative to the incident beam and can take advantage of the fact that cells preferentially scatter light in this “forward” direction.

Additionally, because cells scatter so much light in the forward direction, we can save money and use a less sensitive detector to measure this light. As such, forward scattered light is traditionally and often effectively measured with a photodiode, rather than the more sensitive photomultiplier used to measure fluorescence and side scatter.

However, the story is quite different for particles that are smaller than the illumination wavelength (<488 nm or <~0.5 um), like microvesicles and ectosomes.  It turns out that light scatter by particles of this size range is NOT dependent on the angle at which it is measured. This has very significant implications for small particle analysis. Most importantly, because small particles do not preferentially scatter light in the forward direction as do cells, the resolution in this detector may not be sufficient to measure or even identify these particles above background.

How Small Particles Affect Side Scatter

Side scatter, given its orientation in the quieter fluorescence collection path, as well as its traditional detection by the much more sensitive photomultiplier tube, will likely have much better resolution than forward scatter.

Similarly, we often use side scatter as the “trigger” or “threshold” parameter when measuring small things like bacteria, microparticles, or microvesicles for the same reason—better sensitivity allows us to better distinguish and measure small particles above background.

Interestingly, scatter gets dim VERY quickly when particles have diameters below the wavelength of illuminating light, considering that scatter intensity decreases with a dependence on r⁶ of the particle. The bottom line is that small things like extracellular vesicles can be incredibly difficult to detect using scatter signals. This can make publishing flow cytometry data on small particles very difficult.

What Is An Obscuration Bar?

Instrument manufacturers and operators often take advantage of this property of small particle light scatter by installing an adjustable obscuration bar on the forward scatter detector.

The forward scatter obscuration bar is a universal component of this detector that helps to diminish background in the FSC detector by blocking laser light from interacting with the detector. When no particle is present in the laser beam, scattered laser light hits and is blocked by the obscuration bar. On the other hand, when a particle is present in the laser beam, laser light refracted (scattered) by the particle passes over the bar and triggers signal associated with that particle.

An adjustable obscuration bar can be rotated to expose a wider or narrower surface to the laser beam, blocking more or less laser light, respectively, from hitting the detector. To resolve smaller particles from optical noise, it can be helpful to block more laser light from hitting the detector, which can be accomplished by widening the bar.

Moreover, because small particles do not preferentially scatter light in the forward direction, the proportion of signal blocked by the bar of the total signal is less significant than it would be for larger particles, which do preferentially scatter in the forward direction. This strategy is most effective when a photomultiplier tube is used for forward scatter detection than a photodiode, given that the former is much more sensitive than the latter.

This relationship between particle size, wavelength of illumination, and scatter angle can also help explain the side scatter properties of cells. Typical cellular side scatter signal is much more robustly correlated to granularity than forward scatter is to cell size. In fact, the cytoplasmic “granules” that influence side scatter signal are often smaller than 0.5 um and will thus scatter in a non-Mie pattern. Side scatter of mammalian cells and side scatter of small particles are not terribly different after all.

What Is A Refraction Index?

In addition to the wavelength of the laser and the collection angle of the scatter optics, another factor that significantly contributes to the light scattering intensity of a particle is the refractive index of the suspension medium (sheath, which is essentially water) and the particle itself.

For particles that obey Mie’s predictions, light scattering is largely composed of laser light refraction. When a particle or cell is absent at the interrogation point, the laser’s intersection with the particle or cell produces a characteristic refraction, the “ring of diffraction,” that we are most likely acquainted with on sense-in-air cell sorters. This ring of light spreads outwards from the stream in all directions, in the same plane as the laser beam, and is blocked from entering the forward and side scatter detection paths by the obscuration bars in front of each.

However, the presence of a cell in the laser beam changes the composition of the medium through which the light travels. Laser light now passes through the cytoplasm—which contains protein, lipids, and carbohydrates rather than simply the water it passed through in the absence of a cell—which therefore causes the light to bend differently than it does when it passes through the sheath fluid alone. This refraction of light by a cell is a function of the difference in refractive indices (RI) between the media through which light passes, and it causes laser light to bend in such a way that it passes over the obscuration bar and interacts with the detector, generating scatter signal.

Every type of material has an associated refractive index. When light passes from one material to another—say, from the saline of the sheath fluid to the material of the cell and then back through the sheath fluid again—the amount of light that bends due to this transition is proportional to the difference between the refractive indices between the media. The bigger the difference, the more the scatter.

Why Beads Are Not Good For Calibrating Particles

It is precisely this property of light scatter that makes it a terribly unreliable measurement of cell size.

Two particles or cells of exactly the same size may have different refractive indices, due to their composition (e.g. cytoplasmic proteins), and will therefore generate scatter signals with different intensities. This is also precisely why synthetic beads are also terrible size calibrators.

The refractive index of a typical polystyrene bead (1.59) can be significantly different than a cell’s, resulting in very different scatter signals between a bead and a cell of the same diameter. Given the fact that cells are composed of much more water than a polystyrene bead is and that water’s RI is 1.333, polystyrene beads are going to scatter a lot more light than a typical cell would.

According to studies published in Current Protocols In Cytometry, this situation can be even more severe for microvesicles. The study estimates an average and typical RI of a microvesicle to be approximately 1.39. Given the RI of a polystyrene bead at 1.59 and water at 1.333, a microvesicle’s scatter signal may be one to two orders of magnitude lower than that of the polystyrene bead. Here’s the bottom line—beads are not good at calibrating the scale of a scatter parameter in terms of particle size, regardless of whether the particle is a big one or a small one.

However, all is not lost when it comes to identifying microvesicles. Rather than use scatter as a trigger/threshold parameter to identify these kinds of particles and to measure them, the study suggests that fluorescence is a better choice. Small particles can be labeled with a universal dye that causes all membrane-bound particles in a suspension to fluoresce, allowing discrimination of microvesicles from other particles in the solution. There are nuances and caveats to this kind of labeling, but it can provide a much more robust way to identify membrane-bound microparticles than scatter alone.

Light scatter is a fundamental topic to flow cytometry and understanding how cell size and particle size affect light scatter is critical to performing proper flow cytometry experiments. By understanding how small particles affect forward scatter and side scatter, you can collect better data. Knowing what an obscuration bar is and what a refractive index is, as well as how refractive indices are affected by small particles, will help you design experiments that produce publishable data.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, Ph.D.

Flow cytometry data analysis typically involves a step where a series of gates are defined to identify the population of interest.

The development of robust gating strategies, and the communication of these strategies can be complex. Figures in papers often only tell part of the story. In the world of multicenter studies and the need for improved reproducibility, understanding the specifics of each gating strategy is critical.

As Maecker et al., pointed out in this article, the coefficient of variation due to data analysis went from 20.5% when performed at individual labs, to 4% when performed centrally. With the development of the FlowRepository, investigators can now get access to deposited data and walk through a published analysis with the raw data. This resource will continue to improve the ability to share and communicate the often complex gating strategies.

When training new users on data analysis, there are several different best practices and gating strategies you should incorporate into your analysis. There are also several misconceptions you must understand. For example, too many researchers are reliant on a simple Forward Scatter (FSC) by Side Scatter (SSC) gate to start their analyses.

In general, this basic gate is only good for high-level data reduction, as in removing debris that is on the lower edge of the axis, not for defining your entire starting population. Instead, you should rely on an antibody targeted to a specific cell type (such as CD45 if your goal is to gate on lymphocytes) as your “starting” gate, rather than define your cells based on an FSC x SSC gate alone. This will lead to a more accurate statistical analysis and better conclusions overall. 

3 Flow Cytometry Gates You Should Be Using

There are 3 gates that many researchers are not using but should be using when analyzing their flow cytometry data.

These gates are critical for good data analysis. They will help remove many confounding events that may be clouding your analysis, especially where rare events are concerned.

1. The ‘Singles’ Gate.

Proper flow cytometry data analysis requires single cells. Clumps of cells, except in very special assays, can cause major problems downstream. Likewise, coincident events (two cells passing by the intercept so fast that the pulse cannot be separated) can confound your analysis. As a result, many different ways of removing coincident events that represent cell clumps have been developed.

First, it’s important to visualize and understand the shape of the electronic pulse that comes off your flow cytometer’s detector.

As the above graph shows, the pulse is composed of three components—a pulse height (-H), a pulse width (-W) and the integral of the height and the width known as the pulse area (-A). With digital flow cytometers, the pulse area is typically used to measure and report fluorescence.  This is because pulse area is a more accurate measure of the fluorescence on the cells.

But what happens to this pulse when two cells either stick together or pass through the intercept point too closely?

Here, it’s clear that two things have changed—the pulse area and the pulse width. As the above graph shows, both the pulse area and the pulse width became larger compared to the single cells. You can take advantage of this change to create a gate that eliminates cells that show an increase in pulse area without an increase in pulse height.

As shown in the graph below, cells along the diagonal are the single cells to be gated on.  The cells off this diagonal should be excluded from the data. For this gate, use FSC-Height (FSC-H) by FSC-Area (FSC-A). SSC-H by SSC-A can also be used.

It is important to remember to turn on the -H parameter in your flow cytometer’s software package before collecting data, so that it’s included in the FCS data file.

2. The ‘Time’ Gate.

The order which the cells pass the laser intercept is integral to the FCS file. As such, it’s possible to use time as a gating parameter to help ‘clean up’ the data. There are several reasons that the time gate should be added to your data analysis workflow.

If a stable flow stream (or flow of cells) is not established, good flow cytometry CANNOT be performed. Yes, there will be data in the FCS data file, but the quality of the data will be in question. Visualizing how well the flow of cells was by a time plot will reveal flow issues such as an unstable flow of cells (see graph below).

Areas where there was poor flow can be excluded from areas of good flow by time gating (see left-hand gate versus right-hand gate in graph below, respectively), which will ensure a higher quality of data.

Clogs can happen. Back pressure can happen. Tubes can run dry. These problems can cause issues with data acquisition, often times by affecting the flow rate, and manifesting in a loss of data. Plotting data versus time will help you identify these problems and allow you to remove the questionable events from your downstream analyses (see graph below).

3. The ‘Viability & Dump’ Gate.

Good panel design includes a viability dye and a dump channel. These two markers are labeled with the same fluorochrome (or at least occupy the same channel), and serve to reduce cells that are not of interest in the analysis.

Viability dyes help eliminate cells that are in the process of dying, which is important as these cells can skew your results. Dump channels, on the other hand, contain those antibodies targeting cells that are not of interest to your downstream analysis. By visualizing your Viability dye versus your dump channel, you’ll be able to accurately gate on your ‘live’ cells of interest (see graph below).

An easy way to create a dump gate is to use a collection of biotinylated antibodies and a streptavidin conjugate of the fluorochrome of your choice. This strategy sets you up for a magnetic bead depletion assay, should you want to sort your cells later.

Altogether, the above 3 gates would result in the following gating strategy:

Adding the singles, time, and viability & dump gates to your analysis will improve the accuracy of your results by removing cells that do not belong in your population of interest. By activating the height value in your flow cytometer’s software package, you’ll be able to draw an accurate singles gate. By looking at time versus the flow of your cells, you’ll be able to evaluate whether or not the cytometer operated correctly during your collection run. By using a viability & dump gate, you’ll ensure that you’re only looking at your ‘living’ population of interest.  Using and communicating these gates in your flow cytometry experiments will help improve consistency and reproducibility of the overall field of flow cytometry data analysis.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, Ph.D.

Compensation in flow cytometry is a critical step to ensure accurate interpretation of data. It is also one of the areas that’s steeped in mystery, myths and misinformation.

Before jumping into the best practices for compensation of flow cytometry experiments, it’s good to show what NOT to do when performing compensation.

Manually adjusting the compensation values based on how the populations look, or so-called ‘Cowboy Compensation’ (thanks to Joel Sederstrom for the term), is not the correct way to determine proper compensation.

For example, review the following figure, and ask yourself what is the best compensation value? This figure shows FITC on the Y-axis, spilling into the PE channel, on the X-axis…

Without knowing the median fluorescent intensity of the positives in the negative channel, or being able to evaluate the spread of the data, it is impossible to determine which of these above plots display the properly compensated values.

4 Steps To Compensating A 4-Color Experiment

The best practices for compensation involve following some very specific rules.These best practices also involve the use of automatic compensation protocols that are available in all major data analysis software packages.

(If you’re interested in following along with this blog, you can find the data used in this experiment at this link here.)

Step 1. Choose the correct carrier for compensation.

Compensation is a property of the fluorochrome you’re using in your experiments. The role of the carrier is to bring the fluorochrome to the laser intercept point.

The choice of the carrier is up to you, but for antibodies, the use of compensation beads is strongly recommended. Using beads offers several advantages for compensation, including…

• Cells are not wasted when preparing your compensation controls.
• All the antigen is captured in your solutions, not just some of it. This results in the brightest signal possible for your controls.
• Clear positive and negative signals show up on your control plots.
• Autofluorescence is not a factor since all the beads have the same autofluorescence values.

However, beads cannot be used for some dyes, like viability dyes (such as PI, 7AAD, DAPI), fluorescent proteins, and other protein reporters (redox dyes, JC1, Ca++ dyes).

The biggest concern with preparing proper compensation controls is that the fluorescence intensity of the controls must be at least as bright as that of the cells that the compensation will be applied to. Conversely, the amount of antigen the beads are stained with is less critical.

Very often, compensation beads are stained with too much antigen and as a result, the fluorescent signal goes off-scale. When this happens, do NOT turn down the voltage to bring the signal on-scale. Instead, simply re-stain the beads with less antigen. Often times, staining the beads with 1/2 to 1/10 the concentration used on the cells will keep the signal on-scale, while keeping the signal above that of the cells that the compensation is to be applied to.

Step 2:  Collect the data and make sure there is a sufficient number of events.

After staining the carrier, it’s time to collect the compensation controls. Since compensation is a statistical calculation, the more data collected, the more accurate the compensation will be.

As shown in this data below, as the number of collected events increases, the compensation values move towards the actual compensation value.

For bead-based compensation, it’s recommended to collect at least 10,000 events. For cells, it’s recommended to collect at least 30,000 events.

Step 3. Calculate compensation correctly.

As shown in Tung et al., (2004), how compensation is calculated is based on the matrix algebra.

For the above matrix to be calculated correctly, there needs to be a positive and a negative population in each sample. Since the autofluorescence of the positive and negative carrier need to be matched, you should NOT rely on a universal negative.

All major software compensation packages allow for the use of a single control for the negative population, but again, this should be avoided. In the figure below, unstained beads are shown in red, while unstained cells are shown in blue. As the figure shows, if the experiment is being compensated with beads, and a universal negative of unstained cells is being used, compensation will be incorrectly calculated (note the excess of ‘Primary Signal’).

However, if unstained beads are used in each sample, the resulting compensation values will be correct. As such, make sure ALL of your samples contain a positive and negative fraction in them. You should also make sure that you gate around each positive and negative fraction to define each compensation control for each specific fluorochrome.

Step 4. Apply the compensation values and inspect the results.

Once your compensation values have been calculated, it’s time to apply them to your data. At this point in the compensation process, it’s important to inspect your results. For example, the below figure displays data that has been properly compensated using beads.

As you can see above, the data is compensated but the display is troublesome. The reason the data is displayed incoherently is because it has yet to be transformed.

Transformation allows the full spread of the data to be visualized, while removing events off the axis. As shown below, when the correct transformation is applied, the data around ‘zero’ on both the Y-axis and X-axis is re-plotted. Now the data is shown WITHOUT being compressed against these axes.

Automatic compensation is a flow cytometry best practice. When compensating a 4-color experiment make sure you choose the correct carrier for compensation, collect the data and make sure there is a sufficient number of events, calculate compensation correctly, and apply the compensation values and inspect the results. Failure to properly compensate the data will result in erroneous conclusions which may kill an otherwise promising project. For those who must manually compensate due to their instrument, it’s best to under-compensate the data and controls and then bring them into a third party software to finalize the compensation using the software’s automatic compensation protocols.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, Ph.D.

The most important part of executing a flow cytometry experiment correctly is actually understanding what you are doing. This means you must understand the terms and definitions that are critical to the field of flow cytometry.

As a scientist, you should not just place your faith in a specialized technician. You should not blindly agree with the data you see in front of you without you yourself knowing what ‘logicle scaling’ means, what ‘differential pressure’ means, what happens when you change the differential pressure on an instrument, and so on.

You must also be able to communicate your methodologies and results intelligently. This means, for example, knowing the difference between flow cytometry, flow cytometry cell sorting, and FACS analysis: the latter merely being a trademarked term “owned” by a flow cytometry company.

Top 12 Most Commonly Misunderstood Flow Cytometry Terms

To help resolve this confusion, we have worked with our 700+ Mastery Class members to compile a list of the top 12 most commonly unknown or commonly misunderstood flow cytometry terms here…

1. Autofluorescence

Autofluorescence is the term given to describe the natural fluorescence that occurs in cells. The common compounds that give rise to this fluorescence signal include cyclic ring compounds like NAD(P)H, Collagen, and Riboflavin, as well as aromatic amino acids including tyrosine, tryptophan, phenylalanine. These compounds absorb in UV to Blue range (355-488 nm), and emit in the Blue to Green range (350-550 nm).

The consequence of this autofluorescence is the loss of signal resolution in these light ranges and a decrease in signal sensitivity. Autofluorescence typically increases with cell size. Larger cells have more autofluorescence than small cells due to the simple fact that the larger cells often contain more autofluorescent compounds. 

2. Logicle Scaling

Logicle scaling is an implementation of biexponential scaling published by the Herzenberg lab at Stanford. The biexponential scale is a combination of linear and log scaling on a single axis using an arcsine function as its backbone.

The “logicle” implementation of biexponential was implemented in many popular software packages like FACSDiva and FlowJo. Other types of biexponential scaling exist, including Hyperlog.

Biexponential scales are more generally referred to as hybrid scales and include other variations like lin/log or log with negative. More information on logicle scaling can be found here: Parks DR et al. A new “Logicle” display method.

3. Fluorescence Minus One, or FMO Control

The Fluorescence Minus One control, or FMO control is a type of control used to properly interpret flow cytometry data.  It is used to identify and gate cells in the context of data spread due to the multiple fluorochromes in a given panel.

An FMO control contains all the fluorochromes in a panel, except for the one that is being measured. For example, in the 4-color antibody panel, there would be four separate FMO controls, as shown in the table below. The FMO control ensures that any spread of the fluorochromes into the channel of interest is properly identified.

4. Sheath Fluid

Sheath fluid is the solution that runs in a flow cytometer.  Once the sheath fluid is running at laminar flow, the cells are injected into the center of the stream, at a slightly higher pressure. The principles of hydrodynamic focusing cause the cells to align, single file in the direction of flow.

Depending on experimental needs, different formulations of sheath fluid can be used. Many labs purchase pre-mixed phosphate-buffered saline, while other labs use their own Hepes-buffered saline. The latter is particularly useful for high-pressure cell sorting as Hepes controls pH better at high pressure than phosphate buffers do.

Finally, since the sheath and sample core stream do not mix, you can use water as a sheath fluid on analyzers. Adding a small amount (0.1%) of 2-phenoxyethanol will help as this serves as a surfactant, helping keep the system flowing by reducing the surface tension.

5. Sample Injection Port

In flow cytometry, suspended cells are moved through the flow cytometer’s tubing all the way to the interrogation point and finally into the waste (or to be sorted and recovered). To do this, the fluidics components of the flow cytometry are required.

The fluidics are comprised of three components. The first component is a running fluid (or sheath fluid – see above), that runs through the system in laminar flow. The movement of this sheath can be achieved by several mechanisms, the most common method using pressure provided by pumps.

The second component of the fluidics is the sample injection port (SIP). This is where the sample is pushed through the tubing to be introduced to the sheath fluid. Based on the principles of hydrodynamic focusing, these cells are strung out, single file, in the direction of the flow, where they will pass the interrogation point.

The third and final main component of the fluidics is the flow cell, which is where the first two components and the suspended cells themselves come together. 

6. Isotype Control

The“ isotype” in isotype control refers to the genetic variation in the heavy and light chains that make up the whole antibody moiety. In mammals, there are 9 possible heavy chain isotypes and two light chain isotypes. Every antibody will have a specific isotype, and this is available on the technical information spec sheet.

For example, you might have an antibody with an isotype of IgG1, kappa. This indicates the heavy chain is of the IgG1 isotype. Where things get interesting is that these isotypes can have different non-specific binding affinity to cells, which has lead to investigators using isotype controls, as a control to identify where cells are positive or negative.

The issue with isotype controls is that they are not proper gating controls. Instead, these controls should only be used for identifying potential blocking problems. For more on this, read the following paper: Herzenberg, LA et al. Interpreting flow cytometry data.

7. Antibody Titration

Titration is the process of identifying the best concentration to use an antibody for a given assay. While the antibody’s vendor will provide a specific concentration to use, this may not be appropriate for your assay.

Performing titration is a simple process: fix the cell concentration, the time of incubation, the volume of reaction, and temperature. The graph below displays an antibody that was used to stain 1×106 cells for 20 minutes on ice. To identify the best concentration to use, the modified Staining Index (SI) was calculated and plotted against the concentration. As is shown by this figure, as the concentration increases above 0.5 μg/ml, the SI decreases, due in part to the increase in the background (non-specific staining).

At concentrations below 0.25  μg/ml, the SI decreases because the antibody is no longer at a saturating concentration. Therefore, the best concentration to use is between 0.25-0.5 μg/ml. Titration helps save money and reagents, ensures the optimal concentration of reagent is being used, and avoids background due to high concentration of antibodies.

8. Differential Pressure

Differential pressure-based flow cytometers currently dominate the market. These systems have two pressure regulators. The first is at a constant pressure that sets how fast the fluids run through the system. The second is regulated by you, the scientist.

As the sample pressure goes from low, to medium, to high, the pressure on the sample increases. This results in the volume of the sample increasing (from ~15 ml/min to ~60 ml/min). The difference between the sample pressure and the sheath pressure is the differential pressure. This controls the width of the core stream and the total number of cells passing the laser intercept point. 

9. Jablonski Diagram

The Jablonski diagram illustrates the electronic states of a molecule as well as the transitions between them. These states are arranged vertically by energy, and grouped horizontally by spin multiplicity.

In the below image, nonradiative transitions are indicated by straight arrows and radiative transitions by squiggly arrows. The vibrational states of each electronic state are indicated with parallel horizontal lines. For flow cytometry, it is important to note that the energy of the emission is usually less than that of the absorption. As such, fluorescence normally occurs at lower energies or longer wavelengths.

10. Bandpass, Shortpass, And Longpass Filters

A bandpass filter is a filter that allows light between a set wavelength to pass through it, reflecting only light above and below the set wavelength. For example, a bandpass filter with a wavelength of 550/40nm would allow light between 530nm and 570nm to pass through, but reflect light below 530nm and above 570nm.

A shortpass filter is a filter that allows light over a set wavelength to pass through and reflects light above the set wavelength. For example, a shortpass filter with a wavelength of 450nm would allow light with a wavelength less than 450nm to pass through the filter, but reflect light higher than 450nm.

A longpass filter, on the other hand, is a filter that allows light over a set wavelength to pass through and reflects light below the set wavelength. For example, a longpass filter with a wavelength of 670nm would allow light with a wavelength greater than 670nm to pass through the filter, but reflect light lower than 670nm.

11. Spectral Profile And Spectral Viewer

Every fluorophore has a unique excitation and emission profile which is usually displayed on a spectral viewer, or spectral graph. The combination of the excitation and emission profiles is the fluorophore’s spectral profile. Every fluorophore has a peak excitation wavelength (the wavelength at optimal excitation) and a peak emission wavelength (the wavelength of optimal detection). Each fluorophore will also have a much larger range of excitation and emission wavelengths at reduced optimization. This “curve” is what is displayed on a spectral viewer.

The spectral profile of a fluorophore is used to determine the excitation and detection efficiency at any given wavelength. The spectral profile aids in panel design and selecting optimal fluorophores for a given instrument. The spectral profile can also help in determining compensation considerations. There are numerous resources available to view the spectral profiles of various fluorophores, including this resource from eBioscience.

12. FACS Analysis

Flow cytometry is the science of measuring the physical and biochemical processes on cells and cell-like particles. This analysis is performed in an instrument called the flow cytometer.  FACS Analysis is the shorthand expression for this type of cell analysis. The term FACS stands for Fluorescent Activated Cell Sorting, a term first coined by Len Herzenberg in the 1970’s, and later trademarked by Becton Dickinson. Since that time, FACS has come to be used as a generic term for all of flow cytometry, even though it is a specific trademarked term.

Understanding the above terms, as well as the proper definition of each term, will help you perform better flow cytometry experiments. It will also help you intelligently communicate your methodologies and results in grant submissions and peer-reviewed paper submissions. By being careful not to misuse words such as FACS Analysis, or misunderstand words such as logicle scaling, you will be seen as competent in the field of flow cytometry and your grants and paper will stand better odds of being awarded and passing the dreaded third reviewer, respectively.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Michael Kissner

If you have experience sorting different kinds of cell types on a droplet sorter, you may have noticed that sorting efficiency often seems closely tied to the cell and sample type.

For instance, you may have experienced a better outcome, such as a higher efficiency and more cells recovered, after sorting lymphocytes versus sorting an adherent cell line. There are some fundamental concepts that underpin this phenomenon, and understanding them will help you perform better cell sorting experiments.

What Is Flow Cytometry Cell Sorting Efficiency?

Sorting efficiency, in fundamental terms, is a real-time measurement, generated by the instrument, of how successfully its sorting system is able to resolve cells that we want to sort (target events) from cells we do NOT want to sort (non-target events).

Note that we are talking about the sorting system’s ability to resolve events here (the droplets) and NOT the electronics system.

Efficiency is calculated with the following equation:

The results of this equation are highly dependent on two aspects of the sort: 1) the sort mode chosen for the sort and 2) the setup of the sorting system. Sort modes are sets of rules that instruct the instrument on what to do in situations that I often refer to as “ambiguous.”

In order for the instrument’s sort output to be acceptable with respect to the researcher’s needs, it is not sufficient to simply tell the instrument WHAT to sort (i.e. assign a sort region), but is also critical to tell the instrument HOW to sort the target population. The HOW is determined by the sort modes.

Purity, Single, And Recovery Cell Sorting Modes

If high purity is critical for the downstream application, the sorter must be instructed to exclude any target events from the sort that fall close to any non-target events (usually within one half of a droplet). Otherwise, a non-target event can haphazardly be sorted along with the target events. This kind of sort mode is often called “Purity” or “Purify” mode.

Alternatively, if extremely accurate counting of the output cells is critical for the downstream application—for single cell sorting, for example—“Single” or “Single Cell” mode is often used.

Finally, sometimes it is important to recover every single cell possible from the sort, and there’s not much concern for purity. In this case, we tell the instrument to ignore any rules and to sort everything that falls into the sort gate. These modes are often called “Yield” or “Recovery” modes and will always result in efficiencies of 100%.

Target cells that are sorted are termed sorts, and target cells that are not sorted due to violation of the sort mode rules are often called conflicts, coincidences, or aborts.

Although every type of sorter has its own way of implementing sort modes, all sorters must and do include them. In essence, these modes are comprised of combinations of masks that define where a cell can and cannot be in order for a sort to take place.

What Is A Flow Cytometry Cell Sorter “Mask”?

A certain type of mask, often called the “purity mask,” defines how close a non-target cell can come to a target cell in order to mark that target cell for sorting.

Another mask, often called the “yield mask,” defines how many drops should be sorted in order to include a target cell that may be close to a droplet boundary. In this case, the droplet to sort in order to capture this capricious cell is ambiguous and two droplets may be sorted in order to capture it.

Instruments define where cells fall in relation to droplets in relation to the cells’ passage through the lasers. The only place on the instrument where it can “see” cells is at the laser—it cannot measure where the cells are when the stream breaks into droplets—so the system effectively predicts where the cells will fall in droplets, to a certain degree of resolution that depends on the instrument’s electronics, by relating the timing of cells as they pass through the lasers with respect to the pattern by which droplets form.

It is important to emphasize here that the instrument’s determination of cell positions in droplets are predictions, so there is a degree of uncertainty here that requires a buffer zone between events, determined by the masks, to ensure that the sort outcome is as desired. Additionally, cells may speed up or slow down, depending on the sample type and instrument, between the laser interrogation point and the droplet break-off, compounding uncertainty.

Why Cell Sorters Count Droplets, Not Cells

In addition to sort modes, the sort set-up and sort conditions are tightly bound to the efficiency and sort outcome.

The relevant parameters are primarily the droplet frequency, the event rate, and the percent positive (of the total number of events) of the target population. To understand this relationship, it is critical to keep in mind that when we sort we are NOT really sorting cells, but rather, we are sorting droplets.

In other words, the fundamental sorting unit on a droplet deflection sorter is NOT the cell but is the droplets that (ideally) contain the cell. Therefore, in order to sort, the stream of sheath fluid must be partitioned into discrete sorting units or droplets under controlled conditions.

The number of unique partitions depends on the droplet drive frequency. The higher the frequency, the more droplets are generated per second. Most importantly, the rate of droplet formation, once determined at setup, never changes during the sort.

Understanding the difference between cell sorting efficiency, purity, and recovery cell sorting techniques will help you perform better sorting experiments. When performing a sort, make sure you select the proper sort mode, whether it be purity, single cell, or recovery (the latter is sometimes referred to as yield). Remember, cell sorters use masks to predict which droplets, not cells, to sort. By keeping these facts in mind the next time you sort cells, your experiment will be more successful.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, PhD

At the heart of flow cytometry is the ability to make meaningful measurements of fluorescently tagged cells.

These fluorochromes can be bound to antibodies, a fluorescent protein, a reporter fluorochrome, and the like. Free online spectral viewers are useful in a variety of ways, all of which help improve experimental design and troubleshooting.

These spectral viewers have a special place on every scientist’s browser toolbar. I refer to them regularly, at least weekly.

During the process of panel design, it is useful to have these open to check and compare different fluorochromes. In fact, with recent upgrades to FluoroFinder, there is an integrated spectral viewer based on the filter configuration.

There are a host of different spectral viewers available online. Each one has its strengths and provides specific information. This often necessitates having to use two or three of them to get the information you want. These are the spectral links I use (in alphabetical order):

1. AffymetrixFluorPlan Spectra Viewer
2. BD Bioscience Spectrum Viewer
3. Biolegend Spectra Analyzer
4. ThermoFisher Fluorescence SpectraViewer

All the spectral viewers listed have several common and important features. These all allow the investigator to specify the laser excitation lines, the filter configurations and the fluorochromes. The one caveat with fluorochrome choice revolves around proprietary spectra and some will only be available with specific vendors.

The spectral viewers also output similar information, as illustrated below. This information can include how much of a given excitation curve is found in a given filter, the percentage of maximal emission, and more.

What Is A Spectral Viewer And Why You Should Use It

One of the primary benefits of spectral viewers is that they are useful in learning more about fluorochromes. For example, the popular tandem dye PerCP-CY5.5.  

This shows the excitation at 488 nm, and the nice emission. The BD Bioscience Spectrum Viewer gives you a nice additional feature–the %Max excitation, in this case it’s 98.4%. The excitation curve (shown in dashed blue) ends at about 450 nm. So, how can the spectral viewer help us explain the following data?

In this experiment, the beads were stained with either PerCP-Cy5.5 or Qdot705 and measured with either blue laser excitation (Blue B 710/50) or violet laser excitation (Violet B 710/50). As you can see from the graph, it’s clear that there is significant PerCP-Cy5.5 signal in the Violet B channel (Blue line).

Where is this signal coming from?

Seeing that the PerCP-Cy5.5 excitation profile ends, the following figure shows the excitation profile from PerCP.  Our answer is revealed in the full excitation spectrum.

PerCP, it turns out, can be excited by a 405 nm laser, at about 27% efficiency. Coupled to an efficient transfer to the Cy5.5 acceptor on the tandem, it explains why we see this signal in the violet laser.

This is significant if you were going to be using those two dyes in a polychromatic panel.

How To Evaluate New Fluorochromes On The Market

The Brilliant Violet™ dyes, produced by Sirigen, have been a boon to users of the violet laser. These dyes are extremely bright and have become very popular. The Brilliant Violet™ series of dyes include both polymer dyes (BV421™ and BV510™), and tandem dyes with the polymer core (BV570, BV605™, BV650™, BV711™, BV785™). While they may not be named as tandems, they may have some issues that the spectral viewer can reveal.

Here is the spectrum of BV605™…

Notice that there is a second excitation (here shown in the green 561 nm line). This excitation means that this tandem may have issues affecting the spillover of this dye into a PE and PE-Texas Red®-like channel.

Anytime a new fluorochrome comes out, it’s good to learn about it using these spectral viewers.  The Biolegend Spectra Analyzer is quick and easy to use, and they even have an app for the program, so you can go mobile with it. This is useful when you’re reviewing data with someone and don’t want to access your traditional computer browser.

How To Identify Areas Of Spectral Spillover

Another powerful use of the spectral viewers is to understand what channels on an instrument a given fluorochrome will spill into. The following example is using Alexa Fluor® 488, with three instrument filters placed on the graph.

With the ThermoFisher Fluorescence SpectraViewer, if you hover over the filter, it will report the percentage of the curve that is contained within that filter. In this case, the 530/30 bandpass filter captures about 49% of the curve.

About 12% of the Alexa Fluor® 488 fluorescence is captured with the 585/42 filter, and about 1% with the 630/20 filter.

It is important to note that this is not the amount of compensation that needs to be applied, just the amount of the curve that is present in the filter.

Here is an example of another use for these spectral viewers, courtesy of the AffymetrixFluorPlan Spectra Viewer. In the results tab, it shows in table form the percentage of a given fluorochrome’s emission curve found in the filter in question.  Looking at this figure, with the PE and PE-Cy7 curves plotted, what can be said about the PE-Cy7 fluorochrome?

It turns out that the FRET between the PE emission and the Cy7 excitation is not optimal.

How To Optimize Flow Cytometry Filters

Another great use of the spectral viewers is to optimize the filters for a given fluorochrome on a specific instrument. Take for example, the following data. A researcher noticed some sensitivity issues off of two detectors when an instrument was installed. Beads were stained with FITC (488 nm excitation) and QDot545 (405 nm excitation) and run on the instrument. The data looked like this:

When the 532 nm laser was on, it was clear that the dim signals were shifted to the right. The instrument came with the vendor supplied filters. In modeling this issue, putting these two filters in, along with the laser lines, we see the following:

Notice that the 532 nm laser line was directly in the middle of these two filters. The cause of the loss of resolution was a result of the scatter from cells as they pass through the 532 nm laser. A fraction of this scatter wound up in the fibers leading to the blue and violet detectors.

The lesson learned from this was to always model your filters before ordering an instrument.  A tweak of the filter solved this problem, and the experiments continued.

What Is The FluoroFinder Spectral Viewer?

Recently, FluoroFinder released a new version of their panel design package.  As part of that package, when the cursor hovers over a given fluorochrome, the system will provide the researcher information based on the instrument configuration and filters on the machine. This is shown below.

This is an excellent resource when designing polychromatic panels. It is nice to get a view of the spectra of the fluorochrome choices on the instrument being used and quickly get a feel for how well a given filter/detector combination will capture the photons emitted from the fluorochrome. It also helps to see what other channels might be affected by the fluorochrome choice.

For example, on this instrument, there are several possible filters that can be used for QDOT565. Based on the filter configuration and emission profile of the fluorochrome, this emission may impact three other detectors. For that reason, it may be better to choose a different fluorochrome. This addition to FluoroFinder is a great feature to help make those critical fluorochrome choices during the design process.

Fluorochrome emission is the lifeblood of flow cytometry. The use of in silico tools can save a lot of effort and missed opportunity by allowing for the modeling of excitation and emission profiles in the context of what filters a given instrument is equipped with. Using these tools, it is easy to identify where a new fluorochrome will be measured on an instrument, where a fluorochrome may cause issues with other fluorochromes, and what filters are best for detection. These tools can save a lot of troubleshooting at the beginning of an experiment, and also help understand when issues do pop up. Bookmark them and use them at every opportunity.

To learn more about spectral viewers and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, PhD

Cell proliferation is a critical component in biological systems.

While normal cell proliferation keeps the body functioning, abnormal proliferation (such as in cancer) can be a target for therapy.  

Measuring how cells proliferate in response to a stimulus is a time-honored assay in science. This can be as simple as a cell count between untreated and treated cells. More sophisticated assays can include the use of ³H-thymidine or colorimetric assays (MTT assays).

While these all can measure proliferation, they lack the finesse that flow cytometry can bring to the assay – which allows the phenotypic identification of which cells are actually dividing, as well as allowing for calculations of values such as the precursor frequency, the percentage of cells that have divided, a proliferation index, and more.

Proliferation calls for cells to make a trip around the cell cycle, and there are many ways to measure cell division.

The focus here is on the long-term measure of cell division—the ‘temporal’ dimension for measuring such biological processes as:

• Proliferation of immune cells in response to stimulation
• Self-renewal of stem cells
• Biological homeostasis
• Tumor cell proliferation

To this end, there are several critical components in developing, validating and optimizing an assay to make these measures using flow cytometry. These 4 components are…

1. Pick the right cellular proliferation dye.

Determine which dye you want to use for proliferation. The qualities of a good dye for proliferation include:

• It is taken up by live cells
• It stains brightly
• It is well retained by the cells
• It segregates equally between daughter cells

There are two major classes of these dyes, based on where the dye is retained by the cells.  

The first class is the intracellular dyes that enter the cell, are acted upon by cellular esterases which cleave the compound into a fluorescent form that can also interact with intracellular molecules, thus binding inside the cell.

The most common of these dyes is carboxyfluorescein diacetate succinimidyl ester (or CFDA-SE).  When this enters the cell, it is cleaved to the active form CFSE, which is amine reactive, binding to intracellular proteins, and has been used extensively for cell tracking and proliferation.

The second class of dyes for proliferation are lipophilic dyes that bind to the cell membrane.  

These dyes are typically not fluorescent until they are incorporated into the cellular membrane, and over time distribute over the whole cell. There are a host of these dyes, one of the most popular is PKH26. Table 1 below lists some of the most common cell proliferation dyes available.

There are many more dyes available, and a quick search of your favorite vendor’s catalog will reveal one that will work for your needs. The Molecular Probes Handbook is especially useful in selecting a dye for your experiment needs and instrument capabilities.

Table 1:  Some Common Cell Proliferation Dyes

2.  Validate your proliferation dye—make sure your cells like it!

After you have selected your dye, it is critical to optimize the labeling reaction for the dye.

This will include optimizing the labeling solution (typically between 0.1 to 10 μM), the cell concentration (between 1-20 million cells per ml), the incubation time and temperature, and the quenching step.

In the case of succinimidyl dyes, this would involve adding protein (BSA) and letting the cells rest for 5-10 minutes before washing. As part of this assay development, it is important to make sure that the dye does not kill the cells.

After labeling, a viability check is critical.  

Shown below are data from Dr. Andy Filby (head of flow cytometry at Newcastle University): cells were labeled with increasing amounts of CellTrace Violet (CTV), CFSE and eFluor670 (EPD) and the viability measure.  At 1 μM, the cells have about the same viability, but this rapidly changes, especially for CFSE, where increasing amounts of the dye increased cell death.

More dye is better, in so much as the brighter the signal, the more generations can be measured in the experiment, but not at the expense of increased dead cells.

Dead cells tell no tales.

3.  Optimize your flow cytometry instrument.

In an ideal world, as the cells divide, the fluorescence signal would decrease precisely by ½ and calculating the proliferation metrics would be easy.

As shown in this figure, modified from a lecture Dr. Andy Filby presented for Expert Cytometry (and available to Mastery Class members here), the reality is not that pretty.

Several factors influence the spread of the data, including:

• The true biological variance inherent in the cell systems
• The intrinsic spread because of fluorochrome labeling and speed of the cells through the system
• The extrinsic spread because of the instrument (such as optics, laser power, alignment)

The consequences of these factors can be reduced in several ways:

• Careful planning of the experiments
• Run the cells at low differential pressure
• Monitor the laser alignment
• Keep the instrument (flow cell) clean

Before running samples, run a standard bead set to validate the instrument. Checking linearity, sensitivity, and alignment before the actual samples are run is a good way to help minimize the causes that can be controlled.

4.  Analyze your proliferation data correctly.

In designing a proliferation experiment, as with any flow cytometry experiment, it’s important to develop a data analysis plan before beginning the experiments.

There are three informative parameters that can be measured from a properly constructed experiment. These values are:

• The Percent Divided – the measure of the percent of the input cells that entered division.
• The Division Index – a measure of the average number of divisions which includes the undivided cells.
• The Proliferation Index – the average number of divisions that exclude the undivided cells.

There is a clear temptation to manually gate on the populations and calculate these values, because of the overlap between peaks. It is impossible to easily set the gates to avoid the overlap. This is where modeling of the data comes into play.

Take, for example, the below data (courtesy of Dr. Andy Filby)…

On the left, the populations have been manually gated, while on the right, they have been fitted to a model using the FlowJo™ proliferation package.

Manually fitting the data gives a percent divided at 78%, while the model fitted data shows a percent divided of only 59%.

It’s clear that the above difference has a major consequence on the interpretation of the data.  

It’s also important to remember dyes like CFSE have a 24-48 hour proliferation-independent loss of signal that must be taken into account before measuring proliferation.

Another thing to remember is that as the cells divide, it will become over-represented in the data by 2^{division round}.  To determine the true percentage of dividing cells, you have to do some math. This starts by correcting the frequency of each generation by dividing by 2^{division round}.  

These values are then added together, and the data normalized by that value to determine the real frequency in each population. Furthermore, by subtracting the undivided fraction from 100, it is possible to get the precursor frequency.

Making sure that you know the data needed to answer the biological questions, and the power (and limitations) of the analysis of the data is critical.

Define the question, define the statistics needed and then, and only then, foray into the lab and begin the experiments.

Proliferation assays are a powerful tool for understanding and monitoring this important cell process. Understanding this process, how it can be dysregulated and what ways it can be controlled (chemically), is a critical process. The development of new treatments for cancer, for example, target the proliferation of the cancer cells. Failure to properly design and analyze the data will result in missed opportunities and false leads. Knowing the steps to optimize these assays and properly interpret the results, as discussed here, will help ensure the best data and best opportunities are pursued.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Mike Kissner.

Cell sorters have become more sophisticated to rival the multicolor capabilities of analytical cytometers, offering up to seven lasers and a score or more of detectors.

Likewise, cell sorting experiments have also become more complicated in terms of the number of markers and fluorophores utilized to identify target populations, and more complicated in terms of understanding an increasing canon of flow cytometry and cell sorting definitions.

These days, sorts of up to 12 or more colors are not uncommon.

While multicolor sorts are very feasible and can yield excellent results, success is always a product of very careful planning and optimization.

Attempting a 12-color sort of precious samples without trial runs will almost always result in failure and require a return to the drawing board to perform the optimization steps that were initially foregone.

Additionally, all setup steps, especially compensation, must be performed on-the-spot rather than during offline data analysis, as is commonly done in analytical experiments.

Satisfactory results require that all controls and settings must be perfected during the setup phase immediately before the sort.

The following 4 core technique will help you through the planning, optimization, and trial process to give you successful multicolor sorting experiments.

1. Choose your fluorophores more intelligently.

One of first steps to designing a multicolor experiment is to determine which fluorophores to use to detect target markers.

When going about this task, keep in mind that compensation in and of itself is not something to be feared.

Compensation is unavoidable when venturing into the world of multicolor experiments, and the methodology for choosing fluorophores is to do so in a way that maximizes signal detection above background.

First, obtain the optical configuration – the lasers, their power, and the filter sets –  of the instrument that you will be using to sort.

Once you have determined which spectral bands can be detected by your instrument, utilize an online spectral analyzer tool to find optimal fluorophores for that instrument.

There are several options, and they are all useful. Here are some to try…

BD Spectrum Viewer

ThermoSpectraviewer

BioLegendSpectraanalyzer

Use these tools in combination with a reagent search on the manufacturers’ websites to find conjugates that are optimal for the instrument.

This can be facilitated using panel design tools like Chromocyte, Fluorish, and Fluorofinder.

When choosing reagents, the general rule is to match fluorophore brightness with antigen density.

Keep in mind that all fluorophores are not created equal.

Some are much brighter than others: specifically, they emit more photons and have higher probabilities of detection than others.

For example, Brilliant Violet 421, PE, and APC/Alexa 647 are brighter than PE-Cy5 and FITC, which in turn are brighter than Pacific Blue.

With the advent of polymer dyes like the Brilliant Violets, there are more choices for bright fluorophores than there were in the past.

Reserve bright fluorophores, like APC or Brilliant Violet 421, for low-expression markers like CD25 or CD56.

Dimmer markers, like FITC, can adequately be used to detect bright antigens like CD45 or CD3.

Additionally, choose fluorophores to minimize the amount of overlap between them as much as possible, while not neglecting the importance of the relationship of brightness and antigen density.

On multilaser instruments with multiple beam spots and detection paths for each laser, it can often be very helpful to choose fluorophores so that detection is spread across the lasers and detection paths rather than saturating a single laser with several fluorophores.

On these kinds of instruments, fluorophores are excited and detected at different points in time as the cells pass through the interrogation points, which minimizes overlap and permits detection of fluorophores with the similar emission peaks but different excitation maxima (e.g. PE-Cy7 and APC-Cy7).

Online spectral analyzer tools permit visualization of fluorescence spectra based on laser line and will also overlay typical filter sets to get a sense of how much overlap there may be between detectors.

Keep in mind that overlap depends on laser power, filter sets, and reagent quality (especially in the case of tandem dyes) so the actual overlap on your instrument may be quite different.

The topic of choosing dyes in relation to compensation is a very tricky and difficult one.

In essence, compensation itself is not a problem.

Instrument acquisition software using the proper controls does an excellent job of calculating and correcting for spectral spillover between detectors: the overlap alone is not a problem that cannot be handled.

The big problem, which can result in severe loss of signal-to-noise resolution between stained and unstained populations, is something called spillover spreading.

Spillover spreading is a visual consequence of the process of compensation.

Compensation does not cause it but rather, reveals it.

The detection of every fluorophore is associated with a degree of error that is inherent to the measurement.

In other words, if the same cell stained with a particular dye was passed through the instrument 1,000 times, the fluorescence from the dye would be detected slightly differently during each measurement, which gives rise to a distribution or population.

The amount of variation in this distribution depends on many factors, including the power of the laser, the spectral region being measured (red fluorophores are lower energy, which gives rise to more variation in how many photons are detected at the photocathode during the measurement), the filter set used to detect the dye being measured, and the dye itself.

As with all measurements, dyes detected in channels other than their primary channels (spillover) are subject to the same variation.

During the process of compensation, the population measured (the FITC+ population in the example below) in the spillover channel (the 585/40 channel in the example below; typically used to measure a dye like PE) is shifted down on the log scale so that its median matches that of the negative population.

This process causes the FITC population to appear wider or more “spread out” than it appeared to be before compensation.

This phenomenon is essentially a visualization artifact of logarithmic scaling.

The portion of the scale in which the data resided before compensation is more compact (contains more data “bins” to which measurement values can be assigned) than the portion of the scale that results after compensation, so the population appears to be wider compensated compared to uncompensated.

The critical point about this phenomenon is that the process of compensation has absolutely no impact on the actual variation of the population being compensation; all it does is reveal this error.

Spillover spreading can have significant impact on the resolution between positive and negative populations and this specifically is the reason why avoiding spillover, when possible and reasonable, is important.

This topic can become very complicated very quickly, but the main takeaway is to be mindful of channels that receive spillover from other channels.

The degree to which spillover spreading occurs depends on many things – laser power, filter set, and the fluorophore itself – but, generally, the more spillover from other fluorophores that occurs in a particular channel and the more compensation required for that detector, the more spreading there will be.

Interestingly, a channel used for PE detection can receive significant spillover from other channels and can be problematic in this regard.

So even though PE is a terrific fluorophore in terms of brightness, it may be wise to be wary of using this channel.

While this is not necessarily reason to avoid using PE, it is something to be aware of and can help design smart and well-crafted panels.

For example, utilizing markers that are not co-expressed with the marker that PE is used to detect will help mitigate the extent to which spillover spreading is a problem.

There are some excellent and thorough resources out there that can assist in both better understanding and dealing with spillover spreading. Two of these are listed below:

Nguyen R et al. Cytometry Part A.

Perfetto S et al. Nature Reviews Immunology.

2. Titrate your antibodies correctly.

This step of setting up a panel should be performed regardless of whether the experiment is a sorting experiment or a solely analytical one.

In flow cytometry, it is always important to keep in mind that almost everything we do is geared towards maximizing signal-to-noise resolution to provide the best possible chance to distinguish stained and unstained cells.

The better that we can distinguish positives from negatives, the more robust our identification of target cells and the better the sorting results will be.

Titration is a critical first step in ensuring that signal-to-noise and staining index are as high as they can be.

Although understandably daunting and arduous, with panels that include large numbers of fluorophores, the initial time investment will be well worth it.

When a suboptimal amount of antibody is used, the ultimate result is the same regardless of whether the concentration was too high or too low – compromised population identification.

When the amount of antibody is too high, non-specific binding of the antibody to cells that do not express the target antigen will occur.

When this happens, the negative population may both expand and increase in fluorescence intensity as compared to a non-stained sample.

Both of these effects influence how well the positive population can be discerned from the negative population and are components of the staining index, which is a an excellent measurement of separation.

Staining Index (SI) = MedianPositive-MedianNegative2 x Standard DeviationNegative

When too little antibody is used, the fluorescence intensity of the positive will decrease, yielding compromised separation and determination of positivity.

Be aware that the manufacturer’s recommended antibody volume “per test” is often much higher than what is needed.

The correct volume may be orders of magnitude lower than the recommended volume.

There are many nuanced methods for performing titrations, but they all rely on serial dilutions of a starting amount of antibody, often the manufacturer’s recommendation or slightly more.

However you titrate, make sure to choose the amount that generated staining with the highest staining index.

Keep in mind that only the Staining Index takes any widening of the negative into account compared to the S/N (Median_Positive – Median_Negative), so SI can, but not necessarily, be a better choice for calculating the right titer.

3. Test and optimize the instrument being used.

After the markers, fluorophores, and titrations have been worked out, be sure to run a few pilot experiments before the actual sort.

There is always a chance that some unforeseen issue may crop up even after everything has been optimized.

Even more critically, make sure to test the stain on the instrument that you will be using to sort.

It is very tempting, given the typical price difference in core facilities between time on an analytical cytometer compared with time on a cell sorter, to use an analyzer to test the stain, however, there is significant danger in this approach.

The robustness of a stain and the cardinal ability to distinguish stained populations from unstained populations can be highly dependent on the fundamentals of a cytometer like laser power, filter set, laser geometry and even the detectors themselves.

Sometimes, two cytometers are configured to detect the same fluorophore with different laser paths and excitation sources.

For example, PE can be detected either under 488 nm excitation conditions or 561 nm excitation conditions, and the stain may look significantly different under each of these conditions.

Additionally, the stain may look very different under the typical, more sensitive cuvette-driven interrogation points in analytical cytometers than under the jet-in-air interrogation points in many cell sorters.

4. Use more and better controls.

This can not be emphasized enough, especially when sorting, as all of the setup must be performed immediately before the sort.

There are two primary kinds of controls to that are essential: compensation controls and controls that determine positivity in a channel.

Proper compensation can only be calculated properly by using high-quality controls.

In order for a control to be acceptable, a clear positive and negative population must be discernible, and there must be sufficient cells in each population to collect a file with at least a few thousand events in both the positive and negative gates.

Correctly calculating compensation is not dependent on signal intensity.

The stain will be compensated correctly as long as the control is brighter or as bright as the staining intensity of the actual stain.

Both stained cells and beads can generate high-quality compensation controls.

If you anticipate that one of your stains will be dim or contain a low frequency of positive (or negative cells), beads are a better option.

As far as beads are concerned, be sure to use antibody capture beads, which are available from a variety of sources (BD™ Comp Beads, eBioscienceOneCompeBeads, Thermo Fisher Flow Cytometry Compensation Beads, among others), and not hard-dyed beads.

Antibody capture beads contain both a population of beads that bind antibodies so they can be stained with your experimental fluorophore as well a population of negative beads that will not bind antibodies.

Be sure to choose the product appropriate for the antibodies you are using in your experiment – some are species and/or isotype-specific and may not bind every antibody in your panel.

What about the autofluorescence component of compensation?

What happens if the autofluorescence of cells being used for compensation is different than the autofluorescence of the cells being measured in the experimental conditions?

It turns out that autofluorescence does not affect the compensation as long as the autofluorescence of the positive and the negative in each control is the same.

If this is the case, autofluorescence will not factor into the math of compensation, so it has no impact on compensation.

This can be tricky when compensating for a marker like CD14.

CD14+ cells, primarily monocytes, display a different autofluorescence profile than most of the CD14- cells (e.g. lymphocytes), so autofluorescence may impact the compensation calculation. In this case, antibody capture beads are the better choice.

It is perfectly acceptable if some compensation controls are generated with beads and others with cells.

Be sure that each control contains a negative population so that autofluorescence of the positive population and the negative population is the same.

“Universal negatives” are not appropriate in this case, as autofluorescence will be different among the compensation control samples.

What about compensating for fluorescent proteins?

Because it is critical that the fluorophore used in a compensation control is the same fluorophore being used in the full stained sample, a sample of FITC-stained cells is not a good compensation control for a GFP signal.

Clontech offers beads that are spectrally matched to GFP or mCherry.

For other fluorescent proteins, it is optimal to utilize a cell line that expresses each singly.

This can be arduous, especially if a transfection must be performed, but it really is the best way to properly compensate.

Controls that help determine what is positive and what is not positive are trickier.

We don’t have an extensive arsenal of controls at our disposal to do this, but the ones we do have can be very powerful.

Unstained cells, while often used, are not the best choice.

They may be passable for smaller panels, but they don’t tell us anything about background staining or spillover spreading, so they can often result in false positives, which can be devastating for sorting.

Isotype controls, while perhaps helpful to determine whether there is a general non-specific binding problem with the sample, can display different non-specific binding patterns than the antibody being used in the full stain, so they may not be terribly relevant either.

When staining is bright with a particular marker in a particular channel, we really don’t need (or have) controls to determine what is positive and negative.

When the separation is good, it is clear which populations are stained and which are unstained.

Keep in mind that staining above background itself does not necessarily mean that a population is positive in that it expresses the target marker.

Non-specific binding of antibodies to dead cells, for example, may result in a signal that appears positive for a marker but in reality is not.

What is more critical are controls that help us to determine what is stained and what is not stained when fluorescence is dim and there isn’t a clear bimodal distribution in the channels being measured.

These controls are called fluorescence minus one, or FMO controls, and are generated by staining the experimental sample with every antibody conjugate in the panel except for one.

For example, the FITC FMO control will be stained with every other marker except FITC.

FMO controls help us to determine the effect that our old friend spillover spreading has on the stain, which is an effect that no other type of control can reveal.

Below is a famous figure from a 2004 paper from the NIH that very clearly shows why FMO controls are necessary.

Perfetto S et al. Nature Reviews Immunology.

This figure shows, with real data, that bounds of positive staining are different depending on whether an unstained sample is used or an FMO control is used.

The unstained sample overestimates the amount of positive staining because it fails to take into account the background expansion as a result of spillover spreading.

In other words, the brighter fluorescence of the FMO control, compared to the unstained sample, in the PE detector in the figure above is due to the combined effect of FITC, Cy5PE, and Cy7PE.

The latter two are not represented on these plots but, regardless, have an impact on the fluorescence measured in the PE detector.

While it is often not necessary to prepare FMO controls for every single channel, it is wise to do so during optimization steps and the first time that an experiment is executed on the sorter to get a sense of the extent that spillover spreading is a problem.

Channels that often do not require FMO controls are those in which there is significant separation between stained and unstained populations and those which received little spillover from other channels.

These properties of a stain can be predicted to a certain degree, but even though it can certainly be arduous, be sure to cover all bases by preparing FMO controls for all channels the first few times the experiment is run in-full.

Choosing fluorophores to maximize signal detection and utilizing compensation controls to minimize spillover spreading are important first steps in planning your execution for this kind of cytometry experiment. Titrating your antibodies and optimizing your equipment while maintaining proper controls that you’ve set up and performed in advance are critical first steps to performing accurate multicolor sorting experiments.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, PhD

“It would be possible to describe everything scientifically, but it would make no sense; it would be without meaning, as if you described a Beethoven symphony as a variation of wave pressure.”

― Albert Einstein

The goal of any scientific process, as you know, requires the communication of the data that supports or refutes the hypothesis under testing.

Before it is deemed worthy of publication, it must survive the process of peer review ― where the data is laid bare before a group of experts in the field who judge the material impartially (usually) and in secret ― then pass judgment on the suitability of the information for publication.

The presentation of your data must be clear.

As such, choosing the right flow figures to communicate your data is essential.

Good handwriting (formerly known as proper penmanship) and drawings might have been enough to convince peers in the distant past, but not today.

Today, the expectation is that you’ll choose the right flow figures from all that are available, selecting the ones that reflect your data accurately and without confusion.

There is so much data and so little time that it is essential to present information in the clearest, most concise way.

As Einstein once said: “Everything must be made as simple as possible. But not simpler.”

Presenting the data in the best possible format, highlighting your results while avoiding glitz that can make the integrity of your data suspicious, is key.

At first glance, flow cytometry data is very visual.

Analysis techniques rely on presentations using univariate (a.k.a. histograms), bivariate (a.k.a. dot plots) and even higher order plots (3D plots, SPADE trees, etc.).

The huge caveat with falling in love with any of these types of plots is in knowing the plots used for flow analysis are more often than not a means to an end.

Their purpose is to extract numeric values (such as percent positive or median fluorescent intensity) from the data ― the real value of the data to be presented.

Here are the benefits and drawbacks of popular flow figures to consider when presenting your data:

1. Histograms.

Histograms tend to be the most abused of figures for presenting flow cytometry data.

These plots show the intensity of expression versus the number of events.

Typically, figures are shown with data from different conditions shown on one graph, often with an offset as below…

Histograms are useful for cell cycle and proliferation analysis, but are less useful for presenting data for several reasons:

• No relationship between different markers (can’t identify double positive cells)
• Subtle populations lost in larger distribution (no rare events)
• Shape is dependent on binning (different for different instruments and analysis tools)
• Peak height is a function of the number of events and spread of the data

2. Scatter Graphs.

The real data that is important are the numbers extracted from these graphs. As such, scatter plots should be seen as a way to summarize the real data.

The power of the scatter graph shows several things:

• The number of the experiments that were performed in generating the data
• The average of the data
• The spread of the data
• The significance of the data

3. Bivariant plots.

Bivariant plots have some utility in presenting the manner in which the populations of interest were identified.

Bivariant plots show the relationship between two different markers, allowing for more complex phenotypes to be identified and important populations of interest to be isolated via gating.

The original bivariant plot was the ‘dot plot’, a figure that showed the relationship between two variables, but lacked detail in terms of the intensity of the number of events in a given region.

4. Density Plots.

The dot plot led to the development of the ‘density’ plot ― a way to show not just expression levels, but the relative number (i.e. density) of events in a given region.

Three such density plots are shown below (generated in FlowJo v9)…

Each of these plots show the same thing, just in slightly different ways, so pick the one you are most comfortable with and use it.

5. Contour Plots.

The other way to show the density of your data is to use a contour plot.  Like the above density plots, these show the relative intensity of the data using contour lines. In this case, each line contains x% (as defined by the plot).

In the plot below, the lines are at 5% of the population, so the outermost line contains 95% of the cells, the second line 90% and so on.  

The closer the lines are together, the steeper the ‘island’ of cells. Unfortunately, contour plots are not good at showing the outliers. The best strategy here is to couple a contour plot with a dot plot, allowing your rare events to be displayed (shown below in the plot on the right).

One concern reviewers may have over the contour plot that can prevent your data from being published is that these plots do not convey a sense of the number of events on the plot. This is a common criticism of all bivariate plots.

As shown in this figure, only a few points make a very compelling plot (or seemingly compelling plot)…

The solution to this problem is to indicate the number of events on a given plot. This will give reviewers and all readers an indication of the magnitude of the data involved in the analysis.

6. Gating Strategy (All Plots).

The gating strategy used is of great interest to the reader of a paper or grant. It is also a common criticism of flow cytometry data in general. Why? Because…

Gating is a subjective art form.  

At least, gating can be a subjective art form. In a Nature Immunology paper, Maecker and other researchers performed a series of studies concluding that…

In other words…

Since the conclusions from the study will be based on the populations of interest as defined by the gating strategy, getting this consistent, and communicating how the gating strategy was established, is a critical piece of data to share.

An excellent example of this can be seen in any of the published OMIPs, such as OMIP-3 by Wei et al. (see below).

The above presentation of the gating strategy is valuable for dispelling that myth that gating is a subjective art form.  

As new automated analytical techniques become more widespread, they will also help in addressing this issue while adding a level of confidence that the data extracted for downstream statistical analysis has come from a robust, vetted process.

When preparing figures for publication, the scientific question and hypothesis that forms the basis of the paper must be central and all the figures must be in support of that. The flow cytometry data that forms the basis of the conclusions should be presented clearly and concisely. While it provides pretty pictures and colorful layouts, the meat of the data are the numbers ― percentages of populations, fluorescent intensity levels and the like ― these are what will convince the reader that the hypothesis tested is valid and well thought-out.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

By EditaMotyčáková, Ph.D.

Editor’s note: This is based on Edita’s experiences implementing the proposals reported in Perfetto et al., (2012) Nature Protocols 7:2067.  She submitted this to the ExCyte Mastery Class and developed the Excel sheet (attached) to assist others in tracking their QC data.

Developing and implementing a new Quality Control (QC) protocol can sometimes be a daunting task.

With the continued emphasis on reproducibility in science, QC programs are an essential step that cytometrists are encouraged to both develop and implement.

This QC program comprises several assays that are focused on three important characteristics of the flow cytometer:

(1) Optimal instrument setting (e.g. instrument optimization)

(2) Cytometer sensitivity (e.g. instrument calibration)

(3) Monitoring of day-to-day variability in measurement (e.g. quality assurance)

QC Program Step #1 — Instrument Optimization

Instrument optimization assays include protocols useful for determination of laser power, photoelectron efficiency, testing of filter characteristics, evaluation of signal synchronization, and laser delay determination.

These are central characteristics of the flow cytometer and understanding these values at installation help ensure that when changes are made, the system is performing as well as when it was first brought into service.

In addition to making these measurements at installation, instruments should also be optimized whenever the optical pathways are changed, including: changing a filter installation, installation of a new laser, and/or realignment because of a new flow cell. Basically, whenever an optical pathway is changed. This gives the user a baseline to know how the instrument is performing and a reference for when there are issues.

Instrument optimizations do require some specialized equipment, namely a laser power meter and super-reflecting mirror to perform any of these tests.

QC Program Step #2 — Cytometer Calibration

There are two separate protocols necessary for cytometer calibration:

• Determining the sensitivity of PMTs
• Validation of PMT sensitivity

To complete the first protocol, three bead sets are needed. For this work, you should use:

• Single peak beads – BC Flow-Set Pro Fluorospheres
• Multipeak beads – DakoFluorospheres
• Unstained beads – DakoFluorospheres

To generate the top graphs in the below figure, the voltage was plotted against the Signal-to-Background (S-T-B). This is calculated by dividing the median fluorescent intensity (MFI) of a well-separated peak by the background MFI. In this case, I chose the 4th peak, as it was nicely displayed and recognized throughout most of voltage range and on most detectors.

These data also allow for the determination of PMT linearity, which is measured as difference between MFIs of two adjacent peaks from multiple-peak beads divided by MFI of lower peak from selected peak pair. Like the PMT calibration, this value is plotted over the voltage range.

To interpret the top graphs in the above figure, you need to determine the voltage with the highest S-T-B.  As shown above, the FL1 detector (BP525/30) is most sensitive at 450 V, whereas the FL3 detector (BP620/30) is most sensitive at 600 V. Equipped with this information, when you run a new experiment, you should use the voltage with the highest S-T-B ratio for primary detector as your default voltage.

The PMT linearity showed that the FL1 detector gives linear response throughout the tested voltage range, while in the case of FL3 the lower voltage (below 450V) setting should be avoided because the response to the fluorescence intensity is not linear.  

Remember that compensation cannot be correctly calculated if the signal is not in the linear range of the PMT.

Validation of the PMT voltages is performed to confirm the results of the optimization step above. The protocol for this requires particles (CompBeads) that give a single peak when stained with antibodies. For this experiment, three antibodies were chosen that had been previously characterized (titrated) to ensure the optimal signal.

Here, polystyrene microparticles were used, which bind any mouse kappa light chain-bearing immunoglobulin, as well as three fluorochrome labeled antibodies (mouse CD16-PC5, CD45-PC7 and CD36-FITC), which are suitable for three different detectors. The titration curve showed that the CompBeads can be stained in ratio 1:20, 1:20 and 1:5 for CD16-PC5, CD45-PC7 and CD36-FITC, respectively.

As a second step, CompBeads were mixed with a negative control (without any binding capacity) and labeled individually with an appropriate marker. Finally, the primary detector (FL4 for PC5, FL5 for PC7 and FL1 for FITC) was set to gain highest fluorescence response (see below).

The highest fluorescence for primary detectors was determined as 670 V at FL4 (for CD16-PC5), 500 V at FL5 (for CD45-FL5) and 500 V at FL1 (for CD36-FITC).

The next step was to set the secondary channels to a minimal MFI. For this, a wide range of detector voltages (400-700 V) were tested. The voltage with the lowest MFI found in PMT linear region was chosen (see below).

As you can see in the above graphs, decreasing MFI values were found on the primary channels (yellow bars), despite being set on a single voltage during the course of all measurements.

Notice how drastically the setting of secondary channels can influence results on primary channels.

When calibrating your cytometer, the final step is to measure rainbow single-peak beads using the same primary/secondary channel settings that were used for your individual fluorochrome.

The repeated measurements (n=20) serves for determination of the target value range for CV, which is the highest value found within +- 1 SD or +- 10 % of mean value.

In the above figure you can see that the FL5 detector has the lowest sensitivity in comparison to the other detectors, particularly when compared to the FL1 detector. Even if FL5 detector is set to achieve the highest MFI, its response is much lower than the response achieved by the FL1 detector.

QC Program Step #3 – Implement QC Checkpoints

After spending the time and effort to perform the above measurements to optimize your instrument, it’s important to implement and monitor how the system is changing over time.

Implementing the proper Quality Assurance (QA) checkpoints is the only way ensure that your flow cytometer is functioning properly over time. It’s also the only way to determine what the issue is and how to fix it if there are deviations in these checkpoints.

To implement these QA checkpoints, data from three different bead sets must be recorded and analyzed. The beads include:

•      Single peak bead
•      Multiple peak bead
•      Unstained bead

The parameters to be monitored or “checked” include:

•      PMT voltage
•      CV of the single peak bead
•      MFI of a defined peak (peak 4)
•      MFI of the unstained beads

Altogether, this allows for three calculations that can be used to assess the overall quality of the instrument over time. These calculations are:

•      Accuracy (voltage setting as a function of time)
•      Precision (CV as a function of time)
•      Sensitivity (S-T-B ratio as a function of time)

Once the above data are collected, they must be plotted for analysis. The most common plot is the Levey-Jennings plot, which shows the daily data, a running average and lines representing tolerance ranges.

These plots are commonly defined relative on the stringency needs of the investigator to either +/- 2 standard deviations from the mean, or +/- 5% of the mean value. Typical plots are shown below.

If your instrument does not perform this analysis automatically, or if you’re interested in having a second QA protocol that is independent from your vendor’s protocol, you can download this QA protocol spreadsheet and use it to monitor your data over time.

The above QA protocol spreadsheet has been developed for a FC500 instrument, but the logic can be applied to any number of detectors for any system.

The spreadsheet is organized as follows:

• Tab 1:  Quality control data – this is the place to put the data that will be the basis of the calculations. Make sure to include the bead type and LOT number.  This is especially critical when there is a change to the bead lot. When coming to the end of a lot of beads, it is good to order the new lot and perform an overlap experiment where the old and new lots of beads are run in parallel and any changes to target values can be identified and noted.
• Tab 2: Accuracy – here the data is used to calculate the change in voltage over time.
• Tab 3: Precision – here the coefficient of variation (CV) is calculated using a single bead over time.
• Tab 4: Sensitivity – here the changes in S-T-B are calculated over time.

By adding your data to the first tab, all the graphs and tables will be updated, allowing for a rapid check and confirmation that your instrument is performing within acceptable ranges. A Levey-Jennings plot for detector is shown with both the +/-2 SD and +/-5% of the mean for each value. This helps you visualize the changes over time, and identify trends before they become problems.

Implementing a system of quality assurance protocols of this nature lends confidence to the data collected, especially for those researchers performing longitudinal studies. Optimal instrument setting, cytometer sensitivity, and monitoring of day-to-day variability in measurement leads to improved assurance for those using this instrument to collect their critical data. QC programs will continue to be prudent measures for cytometrists to take as they align with the current emphasis on quality and reproducibility.

To learn more about how to create a flow cytometry quality assurance protocol for your lab and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, PhD

“Every block of stone has a statue inside it and it is the task of the sculptor to discover it.” — Michelangelo

When sitting down to perform a new analysis of flow cytometry data, it is much like Michelangelo staring at a piece of marble. There is a story inside the data, and it is the job of the researcher to unravel it.

The critical difference between sculptor and scientist is that while the sculptor is guided by a creative vision, the researcher is guided by very particular laws of nature and a specific method of working through a biological hypothesis to avoid shaping the results to his or her whims.

Science must be objective, or it is simply an exercise in creative sculpting, which does nothing to move science forward.

Thankfully, there are many ways to avoid shaping the results, and instead sifting for the real and actual data that is relevant to the flow cytometry experiment at hand.

Communicating the results of a flow cytometry experiment is where the researcher has the power to make new or subtle findings instantly comprehensible to the audience. This is also where science becomes an art form.

5 Gating Strategies For Publishing Flow Cytometry Data

Gating is a data reduction technique.

While actual cells will not be lost in trying various gating strategies, data points can be eliminated from your population. In other words, you can reuse and refine your gates and plots over and over again without actually losing cells, but you and you alone will determine which events you are displaying. Hopefully, you will objectively choose the right events to display.

To this end, the following hierarchy was created to help you gate your events correctly…

• Flow stability gating — to capture events once the flow stream has stabilized, eliminating effects of clogging, back-pressure, and other instrument issues.

• Pulse geometry gating — to remove doublets from the dataset.

• Forward and side scatter gating — to remove debris and other events of non-interest while preserving cells based on size and or complexity.

• Subsetting gating — to rely on expression of markers and what they identify. Using viability dyes and dump channels further narrow to the cells of interest. This is where Fluorescence Minus One (FMO) controls become critical in defining the populations of interest.

• Backgating — to provide visualization of cells in final gate at higher level.

The details on this hierarchy, including how each fits together sequentially to produce the optimal flow cytometry figure for every experiment, are outlined below…

1. Flow stability gating.

The principle of this step is to ensure a good and even flow stream during the instrument’s run.

Clogging, back-pressure and other instrument-related issues can affect the flow, so eliminating cells that may have been affected by such problems is an important step to cleaning up the data. An example of this is shown in the below plots.

These plots show the sample running evenly over the time of acquisition. The data are plotted against a time parameter versus a scatter parameter. Either forward scatter or side scatter are good choices, as they are both intrinsic measurements of all events passing through the laser intercept.

The red gate on the right-handed plot was used to remove the first seconds of the run where the instrument was in the process of stabilizing the run and not yet ‘flowing’ evenly.

A recent paper published by Fletez-Brant et al., introduced an automated program in R called “flowClean”, which can do this process in an unbiased, automated fashion.

Interestingly, when this program was run on over 29,000 files in the FlowRepository, the authors showed almost 14% had fluorescent anomalies.

Failure to address this problem reduces the sensitivity of all experimental measurements and may result in inaccurate data and results.

2. Pulse geometry gating.

This gate is used to remove doublets from the dataset and is particularly useful with digital data.

When cells pass through the laser intercept and fluoresce, the photons are converted to an electronic pulse in the photomultiplier tube.

The instrument can measure three characteristics: the height of the pulse, the width of the pulse (or time of ‘flight’), and the area of the pulse (see below figure).

In the case of clumps of cells, the transit time increases, thus the area will also increase. 

In a plot of the area versus the height measurement, the single cells typically fall along a diagonal, while the clumps of cells will show up with increased area relative to the height.

Using this pulse geometry gate removes these clumps, which is important because flow cytometry analysis is based on single cell analysis, not doublet cell analysis or ‘clump’ analysis.

Another example of pulse geometry gating is shown below. Here, the pulse geometry gate is applied to 487,000 cells and, as shown, over 93.7% of them are single cells. This reduced the initial dataset by 30,000 cells.

3. Forward and side scatter gating.

Forward and side scatter gating is one of the most common gating strategies used in flow cytometry analysis.

The goal is to identify the cells of interest based on the relative size and complexity of the cells, while removing debris and other events that are not of interest.

It is recommended that this gating strategy be as generous as possible, to eliminate ONLY those events that are absolutely not of interest.

As shown in the figure below, the major density of events is captured by this gate. The events with very low FSC and SSC, as well as those with low FSC and high SSC are eliminated. These events represent debris, cell fragments and pyknotic cells. As a result, approximately 45,000 more events have been eliminated from the analysis.

4. Subsetting gating.

This is where the major work of data analysis is done. Subsetting gates rely on the expression levels of markers in the analysis, and what those makers identify.

Using tools like viability dyes and dump channels, more unwanted cells are removed to reveal the data relevant to the experiment and the overall hypothesis behind the experiment.

When designing a panel, adding a viability dye is critical to ensure that dead cells, which can non-specifically take up antibodies, are eliminated from the analysis.

The dump channel is useful for ‘mass’-eliminating specific markers that represent cells that are not of interest to the researcher. For example, when performing a T-cell analysis, one might add markers for B-cells, macrophages, monocytes, and the like into a single channel.

As seen in the above figure, plotting a viability marker against a dump channel eliminated another 189,000 events.

Moving forward, additional analysis to identify the specific cells of interest, in this case CD3+CD4+ cells, continues (as shown below). These additional gates have eliminated over 70% of the events that were initially collected on the flow cytometer.
Of the 487,000 cells that were present in the first plot, there are only 127,785 cells remaining — that is to say, a total of 127,785 CD3+CD4+ cells are present.

Knowing what the relative percentage of your final population is will ensure that sufficient cells are collected for meaningful statistical analyses.  

At this point, you should consider your FMO controls, which are used to define the final cellular subsets, in this case CD25+FoxP3+ cells. The FMO control is very useful in addressing issues of how spectral spillover from other fluorochromes in the panel affect the spread of the data in the channel of interest. In the case of the data being analyzed, a gate is drawn on the fully stained sample, and applied to the FMO controls to confirm positioning.

Based on the FMO controls applied to the two right-hand plots below, it is clear that the gate in the left-hand plot is in a good position. Now, the analysis is down to only 2,800 cells.

From here, the researcher would be able to extract additional information, in the form of median fluorescent intensity values, or percentages of cells expressing markers of interest on the identified ‘regulatory T-cells’, as defined by CD3+CD4+CD25+FoxP3+.

5. Backgating.

Backgating is a technique that should be applied at the very end of your gating analysis.

This technique allows for the visualization of the cells in the final gate at a higher level. The goal of this gating strategy is to determine if any cells are being missed by the gating strategies that have been previously applied.

As the red dots in the above right-handed plots show, the FSCxSSC gating could be tightened up, reducing the ‘noise’ downstream. Likewise, the viability gate clearly shows why this particular gate is valuable, as there would be a fraction of cells that would be included if this gate was not used.

Once the gating strategy has been developed and validated, it is time to move to extracting the necessary statistics that will be used to answer the biological question. With careful application of the gates discussed above and the proper experimental controls, the researcher should have freed ‘the statue in the stone’. Michelangelo would be proud.

To learn more about how to get your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written By Mike Kissner

Multicolor cell sorting is a complicated process and certain scientific errors can be common.

Unsuccessful multicolor sorts can result in erroneous data and inconclusive results. Successful multicolor sorts, on the other hand, can give excellent results and lead to dynamic conclusions.

Successful multicolor cell sorting requires special attention to planning.

Using specific setup strategies for your experiment can create a streamlined system for an otherwise complicated process. For example, these critical steps and strategies for multicolor sorting experiments can save you time and maximize your results.

When setting up a multicolor experiment, the most common mistakes are failing to set PMT voltages properly, failing to use a viability dye, failing to address doublet discrimination properly, and failing to set the right sort regions and gates. Eliminating these 4 mistakes is important for any kind of flow cytometry experiment, but particularly for flow cytometry cell sorting experiments.

The following 4 mistakes should be avoided prior to the setup phase, which should be executed immediately before the sort. This setup phase should be included as part of the planning, optimization, and trial process of the experiment to give you the best cell sorting results possible.

Here are 4 common multicolor cell sorting mistakes you should avoid…

1. Failing to set the PMT voltages properly.

When setting up a multicolor experiment, the most saliently critical step is to set PMT voltages and to do so properly.

The overarching theme to this portion of experimental setup, as with most anything in flow cytometry, is to maximize signal-to-background resolution.

As such, setting voltages using an unstained sample to place the negative peak in the first log quadrant (or any other desired position in the plot) may not, and often doesn’t, accomplish the goal of maximizing sensitivity in each channel.

Keep in mind that PMTs do not perform maximally (i.e. convert photons to electrons as efficiently as possible) at every single voltage setting. Moreover, in order to ensure that a detector is operating at peak performance, a sample that contains both negative and positive populations must be used.

An unstained sample provides perspective with respect to the negative population only, so it cannot be used to determine how well a stained population will be resolved from an unstained population.

In general, the danger arises when the voltage is set too low, which may result in suboptimal photoelectron generation and signal detection.

When measuring signal in channels in which cells tend to autofluoresce, like the green region of the spectrum, setting voltages based on the position of the unstained may result in a PMT voltage that is too low. Conversely, setting voltages based on the position of the unstained in red channels, in which cells autofluoresce very little, may result in the voltage being set too high, which in turn may result in the positive population to be off-scale once the full-stain is acquired.

If using BD instruments controlled by FACSDiva (e.g. FACSAria, LSR II, LSR Fortessa) the CS&T system can help to determine minimum baseline voltages, or the minimum voltage at which that detector should be operated. There are some excellent references that provide extensive and thorough methods to accomplish the same goal.

In general, there some useful rules of thumb that can help guide you along the most optimal path for setting your PMT voltages properly.

First, voltages must be set so that no stained population is off-scale. This is critical both from a visualization perspective (no one likes to look at data where staining is smashed up against the high end of the scale), and from a measurement one. The very high portion of the scale may not be in the linear range of the detector and may not facilitate proper signal measurement. Again, this goal can only be accomplished by running a sample with a clear positive population.

Be wary of using compensation beads to set voltages.

Staining can be very bright, which may result in a tendency to reduce voltage to possibly suboptimal settings. After checking to make sure no staining is off-scale, adjust the voltages, usually by increasing them, so that the separation between positive and negative populations is clear and maximized as best as possible.

One common practice in flow cytometry is the tendency to adjust PMT settings with the specific aim of minimizing percent overlap in the compensation matrix. Remember, the primary goal in setting voltages is to ensure that the resolution between positive and negative is maximized.

The percent overlap is not a particularly good indicator of whether separation is maximized.

As long the voltages are set so that no populations are off-scale, the detectors are operating in linear range, and that positive and negative are well separated, do not worry about the compensation percentages, assuming that compensation was set up properly. Instead, let the data speak for itself.

Always ensure that the PMT voltages are the same for each control. Compensation will not be calculated correctly if voltages in all channels are not consistent between controls.

2. Failing to use a viability dye.

Antibodies have a tendency to stick to dead cells, which will result in false positives that may drastically compromise purity.

This can be devastating for a sort, especially when the cells will be used for downstream molecular applications that rely on high-integrity sort purities. Moreover, while false-positive dead cells including in the sort fraction may not grow in cell-based assays, their presence may affect cellular processes of other cells present.

There are many nice choices for viability dyes and there are two kinds whose mechanisms are different: DNA-binding dyes and amine-reactive dyes.

DNA-binding dyes, like propidium iodide (PI), DAPI, and 7-AAD, are typically positively charged molecules with strong DNA-affinity that cannot pass through intact cell membranes. Thus, they only stain the DNA of dead or dying cells with compromised membranes. These dyes are good choices because staining is very rapid, so the dye can be added very soon before the sort and does not require a separate staining step.

Because the stain is present in the buffer in excess, DNA-binding dyes provide a “real-time” indicator of cell death; cells that die during the sort will allow dye into the nucleus and will begin to fluoresce.

While typically not a concern for sorting, these dyes cannot be used with fixed cells because the dye-DNA binding is non-covalent and equilibrium-driven.

If cells are fixed after staining, dye that dissociates from DNA in cells that were dead before the fixation may stain DNA of cells that were live before the fixation, given that fixation disrupts cell membrane integrity. Alternatively, amine-reactive dyes, often called “fixable” dyes, bind covalently to free amines on and in the cell. Staining with these kinds of dyes must be performed during an independent staining step.

Dye will enter and stain cells that are dead and have compromised membranes, so staining intensity of dead cells will be much higher than that of live cells, which permit binding of the dye to only those amines on the cell surface. After staining, cells can be fixed if desired, due to the fact that the dye-amine bond is covalent and not equilibrium-driven, so staining integrity will be preserved after fixation.

In general, DNA-binding dyes are preferable to amine reactive dyes for sorting, given their ease of use and “real-time” properties, so stick with the many choices available for these when designing a panel.

Reagent manufacturers have devised DNA-binding viability in many flavors, so finding one that fits into your panel should not be a difficult task. The SYTOX^Ⓡ dyes, manufactured by ThermoFisher, can be a good choice.

One nice thing is to combine a viability dye with a dump channel, or a channel used to gate out cells that are positive for a marker or multiple markers, to remove “lineage-negative” cells from analysis, for example. Since both the dump and viability dye channels are used to gate out cells that are stained, both can be combined into one channel, which can free up another channel on the cytometer for a another marker.

3. Failing to discriminate between doublets and single cells.

Doublets occur when two cells pass through the interrogation point so close together that the instrument treats them as one event.

When this occurs, the pulses from a doublet event measured in the FSC detector look like those illustrated in the figure below.

The height of a doublet pulse will generally be equivalent to the height of a single-particle pulse.

However, because a doublet pulse is essentially the merger of two single-particle pulses, the area and width of such a pulse will be larger than that of a single-particle pulse. We can take advantage of the disparity between the pulse parameters of single particle and doublet signals to distinguish the two from each other.

Typically, area is plotted against height, height is plotted against width, or area is plotted against width. All cells must have a measurable signal in the parameter chosen, so forward scatter or side scatter are usually utilized.

Also, two doublet discrimination gates, one utilizing FSC and the other utilizing SSC, can be included for more robust doublet identification. While doublet discrimination is important for any kind of flow cytometry experiment, it is especially critical for cell sorting.

Failure to discriminate doublets from single cells can severely compromise the purity of a multicolor cell sort.

A doublet event may incorporate one cell that fulfills the sort logic AND another cell that does not fulfill the sort logic. Because the sorter has identified both of these cells into one event, the entire event — both the target cell and the non-target cell — will be sorted, resulting in both a target and non-target cell in the collection fraction.

The presence of doublets does not necessarily indicate poor performance of a sorter.

Doublet events are a normal and expected aspect of a flow cytometry experiment and whose frequencies are dictated by how the cells are dispersed into a stream. Denser suspensions and sticker cell types can certainly influence dispersion, so do not be dismayed if doublets are observed. The most important thing is to find and eliminate them.

4. Failure to set the right sort regions and gates.

Setting the right sort regions and gates is especially critical for sorting, given that all set-up must be perfect before the sort begins in order to achieve results of the highest caliber. Gates should be set based on the boundaries of positivity determined by FMO controls to ensure that only true positive cells are sorted.

Keep in mind that populations in flow cytometry are distributions with inherent variances or widths.

The width of a population is primarily a function of both the number of fluorophores bound to (immunofluorescence) or expressed by (fluorescent proteins) the cell as well as the measurement variation. The fluorescence of a single theoretical cell passed through a cytometer 1,000 times will be measured differently each time and will give rise to its own “population”.

The degree to which this is the case depends on many factors, including laser power, collection efficiency of the instrument, and wavelength of detection. The lower the population falls on a log scale, the more this error will be revealed in the same way that error is revealed by compensation resulting in spillover spreading.

Lower decades on a log scale contain fewer bins, or fluorescence intensity values, than decades higher on the log scale, so a distribution with the same variance will look broader in the second decade of a log scale than in the third decade.

In the above example, the GFP+ population falls very close to the GFP- population and the two populations overlap.

As such, it is critical in this case to position the region that classifies events as GFP+ far enough away from the negative population to ensure that no GFP- cells fall into the GFP+ region as a result of measurement imprecision. Moreover, the distribution of the negative population reflects no fluorescence signal whatsoever, and there is no meaning to where a cell falls in that distribution.

For the most part, assuming the autofluorescence of all cells in the negative population is the same, a cell on the left side of the negative distribution is no different than a cell on the right side of a distribution. As such, do not expect a “pure” population if the sort region encompasses a specific portion of the non-fluorescent population.

When run back through the instrument for a purity check, the entire negative distribution will be repopulated, given that there is absolutely no difference between unstained cells with regards to where they appear on the scale.

As a tip, it is often better to distinguish dim GFP signal from background on a two-dimensional dot-plot than on a histogram, as illustrated below.

By plotting GFP, or any other signal for that matter, on a plot against another parameter that is not being utilized in the experiment, low-expressing cells can be distinguished from the autofluorescence of non-expressing cells due to how the cells are distributed in both channels, as illustrated above.

The figure below, from Arnold and Lannigan, clearly emphasizes the importance of setting sort gates conservatively when signal is dim. Failure to do so can severely impact purity by permitting non-expressing cells into the sort region.

The above figure is from a paper published in Current Protocols in Cytometry by Arnold et al. Here, Panels A and C shows the effect when the sort gate, R1, is placed too close to the negative population (R2). Because this gate encroaches on the negative distribution, it does not distinguish non-expressing cells from expressing cells. Purity is poor using this gate.

When the sort gate is positioned more conservatively, purity is much higher.

Keep this in mind when setting gates for dimly expressing cells. It can make the difference between a successful sort and a suboptimal one.

Multicolor flow cytometry sorting experiments, while sometimes challenging, are not unsurmountable. When setting up a multicolor experiment, the most saliently critical step is to set PMT voltages properly. In addition, using a viability dye and addressing doublet discrimination and setting the right sort regions and gates is important for any kind of flow cytometry experiment, but particularly for cell sorting. Utilizing the tips described here as well as the abundant other resources available to help optimize multicolor staining, should help clarify some of the more difficult aspects of setting up and executing this kind of cytometry experiment.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written By Tim Bushnell, PhD

Prior to the mid-1960’s, the ability to study a defined cell type was severely limited.

Researchers had to use centrifugation methods, such as differential centrifugation, rate zonal centrifugation, or isopycnic centrifugation, to define cell types.

All of these methods would allow separation of cells based on the property of the particles within different separation medias, but didn’t allow for very fine resolution of the cell populations. 

That all changed starting in the mid-1960’s, when Mack Fulwyler published the first true cell sorter, which combined the power of cell characterization by the Coulter principle with the electrostatic separation of droplets developed by Richard Sweet (and used in inkjet printers).

For the first time, researchers could rapidly isolate individual cells based on more precise physical characteristics.    

4 Common Questions About FACS Analysis

Early cell sorting technology eventually found its way into the Herzenberg lab at Stanford University, where a talented research group added lasers and developed what is now known as the “Fluorescence Activated Cell Sorter”, or ‘FACS’ machine.

This first instrument had a single laser and two detectors, capable of measuring one fluorescence and ‘forward scatter’.

With advances in areas of electronics, lasers, optics, and fluorochromes, instruments are now available that can measure as many as 15+ simultaneous fluorochromes and sort at rates of 20,000 events per second.

Cell sorting technology has come a long way, but many scientists still struggle to answer basic questions about FACS analysis. Here are the 4 most common FACS-related questions…

1. What is FACS and how does it work?

The term FACS is held as trademark by BD Bioscience, but the word has become accepted as a reference for any cell sorter, regardless of vendor.

FACS combines the traditional power of flow cytometry and couples it with the ability to isolate the cells of interest.

The most common FACS systems on the market use electrostatic separation, although there are some systems that use a physical or microfluidics design for isolation of the cells.

Just about every cell sorter is also a standard flow cytometer. As such, cells are stained following standard methods and introduced into the sorting machine by gentle pressure.

From there, the cells undergo hydrodynamic focusing and flow, single file, towards the laser intercept point(s), as the below figure shows. 

Next, the flow stream is vibrated at some frequency, breaking it into many thousands of droplets. Some of these droplets contain the cells of interest. It is to these droplets that an electric charge is applied.

As the droplet flies free, it enters an electrostatic field and based on the applied electric charge, is deflected to a collection tube. Those droplets that do not get a charge are discarded as waste.

There are some technical differences between the various electrostatic sorters on the market. These differences are predominantly based on where the cells are interrogated.

2. What are the range of cell types that can be sorted by FACS?

The cell type that can be sorted is limited to the size of the cell, the quality of the instrument, and the ingenuity of the investigator.

Cell sorters have a nozzle, and the size of the nozzle dictates how large (or small) a cell can be sorted. Most often, cells should be 4-5 times smaller than the nozzle being used.

Most sorters on the market today can sort from very small cells (bacteria) to very large cells. There is even a special sorter that can sort very large clumps of cells and even small organisms.

3. How fast can a FACS instrument process cells?

When it comes to the processing speed of a cell sorter, there are two points to consider.

The first point to consider is the inverse relationship between the size of the nozzle and the frequency of droplet generation that will produce a stable stream.

The below table shows the frequency of sorting for several different nozzle sizes. You can see that there is a range of frequencies, which are related to the pressure of the system. The pressure of the system has to be balanced with the nozzle size to produce a stable stream.  

The second point to consider regarding the speed of the cell sorter is related to how many events per second the system should run. This relates the need for purity of the sorted product and the poison distribution of events within the fragmented stream.

If there are too many events based on the frequency, this leads to the decreased purity and loss of recovered cells. 

As the above figure shows, there is a greater chance of having two cells next to each other, or multiple cells in one drop, when the event rate approaches the frequency of droplet generation. A good rule of thumb is an event rate at ¼ the frequency, as the below table shows.

Now it becomes possible to calculate how long a sort might take. For example, sorting at 60 kHz, at a rate of 15,000 events/second, if one needs 100,000 cells for a downstream application, and the cells are at a frequency of 1%, will take at least ((100,000 cells)/(frequency))/15,000 about 667 seconds or 11 minutes for this sort.  Assuming a 50% recovery would double the number of input cells needed, thus increasing the time to 22 minutes or so.

4. What topics should someone new to cell sorting consider?

There are several important tips that can help a researcher who is new to cell sorting and help ensure the best possible outcome for the experiment…

1. Talk to the operator(s) of the cell sorter. They are friendly and will be able to provide a wealth of information on planning and executing the experiment. Enter into their good graces by making them part of the process to ensure they care about your cells as much as you do.
2. Review the protocol. Go over the staining protocol and make sure everything is ready before beginning the process. Do the back of the envelop calculation to make sure you know how many cells will be needed. Always assume a 50% loss from the cell sorter (due to electronic aborts, coincident events, cells dying post-sort, etc.).
3. Coat the tubes. Coating your experimental tubes goes a long way to ensure that the charged droplets don’t stick to the plastic of the catch tube. Neutralizing that charge by coating with some protein can improve recover post sort.
4. Filter the cells. Nothing ruins a sort like a clog. Remember Howard Shaprio’s First Law of Flow Cytometry – “A 51 𝞵m particle clogs a 50 𝞵m orifice.” Filtering the cells just before they are put on the sorter is a good way to minimize this issue. Another great trick is to add some DNAse (10 units per ml of sample) to help reduce clogging caused by dead cells releasing DNA (the biological equivalent of duct tape).
5. Use the right controls. As with every flow cytometry experiment, controls are critical. Bringing a tube and saying ‘sort the green or red ones’ doesn’t endear one to the sort operator. As such, consider the following controls…
   1. Compensation controls
   2. Any gating specific controls (i.e. FMOs)
   3. Any controls necessary for setting gates
   4. (Paper control) – A copy of the gating strategy

6. Be on time. Sorting facilities often have back-to-back bookings, and need to get each one started on time. Be considerate to everyone and be on time.

In the end, cell sorting is a powerful tool that can be used to phenotypically identify cells of interest, from GFP+ transfectants to rare stem cells, and isolate them to homogeneity for downstream applications ranging from culturing, to genomics and NGS sequencing, to proteomics, etc. From the humble beginnings of a hybrid technology to the instruments available today, FACS analysis is now the entry point for many experiments. Understanding the inner workings of FACS instruments and the best practices for preparing samples will lead to more successful experiments.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Ryan Brinkman, Ph.D.

Editor’s Note:  Reproducibility continues to be a critical area that all researchers need to be aware of. From the NIH’s focus on reproducibility in grant applications, to a renewed focus by reviewers on the way data has been analyzed and presented, it is imperative that researchers keep up on best practices to ensure they pass these hurdles. 

One area that flow cytometry researchers should be focusing on is the emerging changes in the area of automated data analysis. Over the last five years there have been dramatic changes and improvements in these programs and workflows. As Dr. Brinkman discusses below, the automated analysis of flow cytometry data is coming into its own. 

Flow cytometry (FCM) datasets that are currently being generated will be two orders of magnitude larger than any that exist today, and new instruments, both flow and mass cytometry, have increased the number of parameters measured for each single cell by 50% (to 30).

Even in 14 dimensional datasets there are 3¹⁴ possible cell populations of interest pre-sample (1). The information contained within large and complex single cell datasets can only be realized with approaches to effectively curate, integrate, analyze, interpret, and share these datasets.

What Is Reproducibility And Automated Analysis?

While there are many steps in the analysis pipeline that can benefit from automated approaches for which approaches have been developed (Figure 1), a major bottleneck in the analysis of flow cytometry data is in the identification of cell populations.

Manual analytical techniques lack the capacity and rigour to bring out the full potential of signals latent in the data (1, 2) and its subjectivity has been identified to be the primary source of variation between analytic results (3, 4).

Figure 1: Typical flow cytometry automated analysis workflows.

Analysis usually starts with several pre-processing steps (blue boxes) followed by identification of cell populations of interest (orange boxes) and visualization.

To address this problem, the computational cytometry community has developed a collection of widely used approaches for the high throughput analysis of FCM and Mass Cytometry (CyTOF) (5). Methods have been extensively evaluated against manual analysis through the Flow Cytometry: Critical Assessment of Population Identification Methods (FlowCAP) project (6,7,8) and have been found to meet and in many cases exceed the performance of manual analysis.

Only by taking advantage of cutting-edge computational abilities will we be able to realize the full potential of data sets now being generated and be able to keep up with the quick rate of progress and advancement in our fields.

Further Reading (References hyperlinked above)…

1. Aghaeepour N, Chattopadhyay PK, Ganesan A, O’Neill K, Zare H, Jalali A, … Brinkman, RR. Early immunologic correlates of HIV protection can be identified from computational analysis of complex multivariate T-cell flow cytometry assays. Bioinformatics 2012, 28(7):1009-16.
2. O’Neill K, Aghaeepour N, Špidlen J, Brinkman RR. Flow cytometry bioinformatics. PLoS Comput Biol 2013. 9(12):e1003365.
3. Maecker H, Rinfret A, D’Souza P, Darden J, Roig E, Landry C, … Sekaly R. Standardization of cytokine flow cytometry assays. BMC Immunol 2005. 6:13.
4. Qiu P, Simonds E, Bendall S, Gibbs KJ, Bruggner R, Linderman M, … Plevritis S. Extracting a cellular hierarchy from high-dimensional cytometry data with SPADE. Nat Biotechnol 2011. 29:886-91.
5. Kvistborg P, Gouttefangeas C, Aghaeepour N, Cazaly A, Chattopadhyay PK, Chan C, … Maurer D. Thinking Outside the Gate: Single-Cell Assessments in Multiple Dimensions. Immunity 2015. 42(4):591-92.
6. Aghaeepour N, Chattopadhyay P, Chikina M, Dhaene T, Van Gassen S, Kursa M, …, Brinkman RR. A benchmark for evaluation of algorithms for identification of cellular correlates of clinical outcomes. Cytometry 2016, 89(1):16-21.
7. Aghaeepour N, Finak G, TheFlowCAPConsortium, TheDREAMConsortium, Hoos H, Mosmann T, … Scheuermann RH. Critical assessment of automated flow cytometry data analysis techniques. Nature Methods 2013. 10(3):228-238.
8. Finak G, Langweiler M, Jaimes M, Malek M, Taghiyar J, Korin Y, … McCoy J. Standardizing Flow Cytometry Immunophenotyping Analysis from the Human ImmunoPhenotyping Consortium. Scientific Reports 2016. 6:20686.

As mentioned above, FCM datasets will soon be two orders of magnitude larger than those that exist today. As such, researchers must keep up on best practices for data reproducibility, especially in the area of automated data analysis. This will ensure that the field of flow cytometry and scientific research overall maintains its integrity while continuing to advance rapidly.

To learn more about how to improve reproducibility through automated analysis, and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, PhD

To reproduce reliably in flow cytometry, one must control the gate.

The identification of the target cells of an experiment is the critical first step to performing the secondary analysis that will be used to judge the biological hypothesis and is done by peeling away the layers of cells that do not meet the criteria.

This involves the data reduction method of ‘gating’ with the researcher as gatekeeper, controlling what may pass and what shall not pass, based on the controls designed for the specific experiment.

It is disappointing to realize that in the paper, Maecker et al., the authors evaluated different models for conducting clinical trials and found that individual labs experienced a ~20% CV in the data analysis whereas a central lab showed only a ~4% variance in data analysis.

One of the best ways to improve gating is to ensure the most appropriate controls are identified and collected in the experiment.

How these controls are used to identify the population of interest is also critical to improving this process. There are 4 common gating controls that can be used for improving gating consistency and reproducibility:

1. Fluorescence Minus One (FMO control).

The term Fluorescence Minus One (FMO) was first introduced in this Cytometry paper in 2001. The FMO control is designed to identify the effects of spectral overlap of fluorochromes into the channel of interest.

This overlap can reduce the sensitivity of measurement in the channel of interest and make identifying the true positive population difficult. The FMO control is performed by staining the cells of interest with all fluorochromes except one. When the data is displayed, the spread of the data in the channel of interest becomes apparent, as shown in the figure below.

Here, human PBMCs were stained with FITC, PE, CY5.5 PE and APC. The left panel shows the unstained sample and the right panel, the fully stained sample. The middle panel shows the PE FMO control.

If the unstained control was used to set positivity, as shown by the red line, it would appear all the cells would be PE positive. However, when the same cells are viewed in the context of the FMO control, it becomes clear that there is spread of the signal, and based on the blue FMO bound line, it is clear these cells are not PE positive.

The FMO control is a valuable control and should be run with all combinations during panel development. Through this development cycle, the researcher will be able to identify the critical FMO controls that are necessary for proper gate placement.

The FMO control is especially essential when attempting to measure rare events, identify emergent markers, or where there is a continuum of expression.

2. Internal Negative Controls (INCs).

Internal Negative Controls (INCs) are those cells in the staining sample that do not express the marker of interest. Unlike the FMO control, where one reagent is left out, the INC is exposed to all the markers, but biologically does not express the marker of interest.

In this case, the INC can help identify and address proper gating when there is non-specific binding of the antibody. This control takes advantage of the fact that we know a bit of the biology of the system and do not expect that the INC cells will bind with the target marker. This, of course, needs to be confirmed in the literature and through experimentation, but leads to a powerful control for proper gate placement. 

In this figure, the data on the left comes from the identified INC cells. They are plotted against CD4+, which is our population of interest.

To help set the gate, a quadrant marker can be used to help track the boundary of the INC. As can be seen, the target cells are clearly positive for the marker of interest, and the INC helps ensure we have identified the correct gate.

3. Unstimulated control.

A third control, useful for stimulation experiments, is the unstimulated control that Maecker and Trotter discuss in their paper from 2006.

The unstimulated control again relies on the biology of the system to assist in setting the proper gate. The unstimulated control also takes into account the background binding of the target antibody, since the unstimulated cells should not be expressing the target. 

As shown in this figure, there is some background binding of the Activation Maker target on the un-stimulated cells. The FMO (left panel) is used to correct for issues of spectral spreading into the Activation Maker channel, but alone does not allow the proper gate placement. It is only when the FMO is combined with the unstimulated control that the best gate placement identified.

4. Isotype control.

The final control to consider is the isotype control. The concept is that one stains cells with an irrelevant antibody that has the same isotype as the target antibody and labeled with the same fluorochrome. This is supposed to allow for identification of the background binding caused by the specific antibody isotype.

The use of this control remains controversial.  

Several papers, such as this one from Keeney et al., call into question the use of isotype controls for setting gates. Maecker and Trotter caution on reliance of the isotype control, and show an excellent figure (Figure 2) where PE-labeled isotype controls show wide variability of staining on small lymphocytes.

When using an isotype control, one makes several assumptions:

1. That the affinity of the variable region on the isotype has similar characteristics for secondary targets as the target antibody.
2. There are no primary targets for the isotype Ab to bind to (and do you know what the primary target is for the isotype?).
3. The fluorochrome to protein (F/P) ratio is the same (and how do you titrate an isotype control?)

We cannot easily know the answer to #1 or #2 and must trust the vendor that the Ab target will not bind to the cells of interest.

Other than with large fluorochromes (PE, APC, etc.), where the F/P is usually 1:1 (due to the size of these fluorochromes), antibodies can have very dramatic optimal F/P ratios for FITC and the Alexa dyes (for example), that have to be optimized out during labelling.

This information therefore has to be collected by the vendor during QC and provided to the customer, something not always readily available on websites. 

The isotype control becomes another variable to be tested, validated, and optimized for marginal gain as a gating control. As Maecker and Trotter state,

“…It is thus a hit-or-miss prospect to find an isotype control that truly matches the background staining of a particular test antibody. And, remembering that we are using the isotype control to help us define the true level of background staining, this becomes a circular proposition…”

Where isotype controls can assist researchers is in assessing the success of the blocking of the cells.  In this case, if the cells are poorly blocked, the isotype control can reveal that, but should not be used to set gates.

In the continuing efforts to ensure consistent and reproducible data, the proper use of controls to establish the boundaries of gates is critical. With the exception of the isotype control, each of the controls discussed above serve a specific role in that process, and should be part of every experiment. This will help reduce the variability in the data in a given experiment, and when the use is communicated (or demonstrated) in publications, it will assist researchers seeking to reproduce the data in achieving similar results, while helping to reduce data analysis variability between institutions.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, PhD

The optical system of a flow cytometer consists of an elegant coordination of many components that function concordantly and synchronously to generate the signals that we need to measure in order to shed light on the biology at hand.

Understanding the optical system of a flow cytometer may seem unnecessary for performing a typical experiment, but the more you know about your instrument, the better you will be at understanding nuanced aspects of your data, as well as troubleshooting any potential issues that may arise during an experiment.

A cytometer’s optical system can be broken down into two major parts: A) lasers, and B) lenses, mirrors, and filters.

Lasers and some lenses comprise the excitation optics that generate optical signal, while other lenses, mirrors, and filters form the emission optics which collect optical signal. A brief description of the roles of each of these components is listed below, followed by more detailed descriptions.

Part I of this article will focus on lasers. Look for Part II for the remaining discussion on lenses, mirrors and filters.

• Lasers illuminate the stream with coherent, focused light of specific wavelength (energy) and power. This illumination facilitates the generation of fluorescence signals from cells labeled with fluorophores and light scatter signal from redirected laser light.

• Lenses focus laser light and collect light scatter and fluorescence optical signal, and direct this signal to the optical detection path.

• Mirrors are responsible for directing light through the detection path and partitioning it so that fluorescence and scattered light are directed to the appropriate detectors.

• Filters, placed in front of detectors, function to restrict the light that is introduced to the PMT detectors so that each detector can be dedicated to measure fluorescence from a specific set of fluorophores.

Before diving deeper into the optical components, it is worth discussing some fundamental concepts about electromagnetic radiation, or light.  While we typically think of light as something that is visible, the electromagnetic spectrum spans a very large range, of which visible light is only a small portion.

Figure 1 below from NASA illustrates the limited range occupied by visible light in the entire spectrum.

Electromagnetic radiation is characterized by its wavelength. Wavelength is inversely proportional to energy, so the longer the wavelength of light, the lower the energy.

Electromagnetic radiation with very short wavelengths, like gamma (~10 picometers) or X-rays (0.01 – 10 nanometers), has very high energy — high enough to break covalent bonds and wreak havoc on biological systems. Those with longer wavelengths, like microwaves (1 mm – 1 m) and radio waves (whose wavelengths can be in the kilometer range), are lower in energy.

Flow cytometry is primarily concerned with the visible spectrum, which occupies a portion of the spectrum in about the middle, with wavelengths of about 380 – 700 nanometers or so.

The spectral range that is utilized in flow cytometry is actually a bit wider than the true visible spectrum, typically between ~350 nm to ~800 nm. The wavelength of visible light determines its “color”. Ultraviolet light, the highest energy light used in flow cytometry of wavelengths below about 400 nm, is not visible. The “cytometric spectrum” can be very roughly conceptualized as follows:

• ultraviolet light occupies the mid-to-high 300 nm range
• violet light occupies the low 400 nm range of the spectrum
• blue light occupies the mid-to-high 400 nm range
• green light occupies the low 500 nm range
• yellow light occupies the mid 500 range
• orange light occupies the high 500 nm range
• red light occupies the range above about 600 nm
• light above ~700 nm is not visible.

When it comes to your flow cytometer’s lasers, there are 4 factors that you should understand. These 4 factors are…

1. Coherence.

In order to measure fluorescence from labeled cells, a light source is necessary to produce this fluorescence. Light can be generated in several ways, but the most effective way for the conditions and configurations of flow cytometry is by utilizing the laser.

Lasers, whose name is actually an acronym (Light Amplification by Stimulated Emission of Radiation), are especially suited for flow cytometry for two primary reasons…

First, laser light is coherent. Second, laser output is of a very narrow energy range — the wavelength of light can be specified with high precision.

Coherence is the best thing about lasers as far as flow cytometry is concerned. In technical language, this means that all of the light that is emitted by the laser, according to Shapiro, is “in phase with and propagating in the same direction.”

In practical terms, this means that all of the photonic power of a laser can be directed and focused onto a very small spot. Unlike microscopy, particles flowing through a cytometer spend a very short amount of time in the illumination spot. Given a stream velocity of 20 meters per second, a beam spot of 20 micrometers, and a 15 micrometer cell, the cell is illuminated for only 0.015 microseconds. That’s not a whole lot of time.

To maximize the likelihood that sufficient fluorescence events are produced by a labeled cell, and in order to best measure that fluorescence, it is necessary to bombard that cell with as many photons as possible.

The coherence property of lasers ensures that the photon density at the illumination point is high enough to allow us to precisely measure the fluorescence necessary to glean useful biological information.

In contrast, other kinds of light sources, such as arc lamps or LEDs, which are commonly used in fluorescence microscopy, are not coherent. Their light output travels in all directions from the origin, and specialized optics are required to gather this light and direct it onto a measurement point.

Another benefit of lasers is that they can be designed to produce light of a very narrow spectral, or wavelength, range.

Unlike a typical fluorescent or arc lamp, which output white light, or a wide spectral band of photons, laser output can be tuned, depending on the construction and materials of the laser, to produce a particular color, or wavelength, of light. Laser lines commonly used in flow cytometry are: 355 nm, 375 nm, 405 nm, 488 nm, 530 nm, 561 nm, and 640 nm.

2. Spontaneous and stimulated emissions.

The way that lasers work is interesting. The laser consists of material, called the lasing medium, typically through which electrical energy is pumped. This causes electrons in the medium to be excited, or transition to higher energy states. When the electrons fall back to lower energy states, a photon is generated.

The emission of these initial photons results from spontaneous emission — they are not in-phase or polarized and are not necessarily of the same energy (wavelength), as reported by Shapiro. However, the incredible thing, which Einstein showed, is that when electrons of a molecule or atom are excited to a higher energy state, the presence of a photon nearby with a particular energy will increase the probability that the excited molecule will EMIT a photon with the same energy.

This is called, in contrast to spontaneous emission, stimulated emission, and is the logic behind the word “stimulated” in the acronym “laser”.

In other words, photons have a “mob mentality.” If one if doing something (i.e. propagating in a certain direction with a certain wavelength), other photons like to follow-suit and do the same.

By equipping cytometers with multiple lasers, each outputting a specific wavelength, a spatially-separated system can be constructed in which fluorophores are illuminated and excited at distinct points on the stream.

In this kind of system, each laser is focused on its own spot, or interrogation point, on the stream, so only fluorophores with excitation spectra in the range of the laser’s wavelength will be excited at each interrogation point.

By constructing a system like this, it is possible to simultaneously measure and differentiate fluorophores which have very similar emission spectra but different excitation like, for example, PE-Cy7 and APC-Cy7.

Both of these fluorophores emit photons at the same wavelength (Cy7’s emission spectrum, whose maximum is in the high 700 nm range). However, PE-Cy7’s excitation spectrum is largely restricted to the 488 nm or 561 nm laser lines, while APC-Cy7 excitation spectrum is largely restricted to the 640 nm laser.

Spatially-separated systems can differentiate between these two fluorophores because each beam spot, or interrogation point, is associated with its own dedicated collection path. In other words, fluorescence signal from 561 nm excitation is routed exclusively into the 561 nm collection path, while fluorescence signal from the 640 nm collection path is routed exclusively into its own, separate path.

The simultaneous detection of both of these fluorophores would not be possible if PE-Cy7 and APC-Cy7 were excited at the same point in time and space.

This is only possible when the illumination source of a single interrogation point consists of a very narrow range of wavelengths.

3. Colinear systems.

In colinear systems, on the other hand, multiple lasers are focused on the same spot. This can be a cost-effective and convenient way to accommodate two excitation lines when two separate beam spots (interrogation points) are not practical or possible.

However, these systems carry the caveat of the inability to simultaneously measure two fluorophores with very similar emission spectra, as described above.

For example, it is very challenging to measure and distinguish Brilliant Violet 786 from APC-Cy7 simultaneously using a 405-640 nm collinear system. Both Brilliant Violet 786 and APC-Cy7 fluorescence over largely the same range of wavelengths but excite at very different wavelengths (~405 nm for Brilliant Violet 786 and ~640 for APC-Cy7).

Since both of these dyes will be excited at the interrogation point in a collinear system and then be collected into the same optical path, both will be measured by the same PMT. In contrast, Brilliant Violet 786 and APC-Cy7 can be measured simultaneously on a spatially-separated system.

4. Lasing mediums.

There are a few different kinds of lasers, with respect to the lasing medium, and while a detailed discussion of this aspect is beyond the scope of this article, most lasers in current cytometers are solid-state lasers.

In these kinds of lasers, the lasing medium is a solid, as opposed to a gas or plasma. Lasers these days are much smaller, have lower power requirements, and do not require the amount of warm up time that they used to in the past.

The evolution of laser design is one of the reasons that cytometers and sorters have gotten so much smaller in the last decade. The era of the “benchtop” cytometer has in large part been facilitated by the development of smaller lasers without sacrificing output power.

In a flow cytometer, lasers must be shaped by the excitation optics before they reach the interrogation point and interact with cells. This shape is typically elliptical which results in a Gaussian energy profile. In other words, the photons are most “dense” in the middle of the beam and taper off towards the edges.

Figure 2 below illustrates this property of the elliptical laser shape.

Because energy is densest in the center of beam, cells must flow through this portion of the spot in order to generate most fluorescence. When flow rates are high, the core stream through which cells flow widens, which results in more variation in cells’ positions in the beam.

Cells located towards the edge of the beam will be exposed to fewer photons and generate less fluorescence, while cells that flow through the center of beam will be exposed to more photons and generate more fluorescence.

Figure 3 below illustrates how flow rate can affect cells’ positions in the laser beam.

Finally, another helpful property of lasers is that their output, or power (in milliwatts), can also be specified. Laser power, which is essentially a measurement of how many photons are output per unit time, is usually adjustable on a cytometer but sometimes not.

Typical powers range from 20 mW to 100 mW, although some cytometers are equipped with very high-powered and considerably dangerous lasers of up to several hundred mW. Although not always apparent in practice, the more power the better (when it comes to lasers).

The more photons that a fluorophore sees, the higher the chance that that fluorophore will fluoresce and the higher the chance that a useful biological measurement can be made.

Regarding power and safety, with the exception of UV lasers (~355 nm), lasers on cytometers in the typical range of powers are not terribly dangerous, as long as care is taken to not look directly into the beam. That being said, leave laser alignment to your service engineers unless you’ve had formal laser safety training. Any laser can be dangerous in the right context.

For further reading, check out the following: Shapiro, H.M. Practical Flow Cytometry. New York: John Wiley & Sons, 2005.

Part I of a look into your cytometer’s optics focused on the dynamics of the lasers and their application in flow cytometry. Understanding the coherence property of lasers and how they impact fluorescence, along with principals of emissions, use of colinear systems and lansing mediums helps you understand the intricacy of the equipment you’re using while providing you with the opportunity to troubleshoot during your experiments.

To learn more about getting your flow cytometry data published and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.

Written by Tim Bushnell, Ph.D.

The importance of the optical system of your flow cytometer was established in Part I of this series, noting in particular the benefits of foundational knowledge to gain a comprehensive view of your data, as well as troubleshooting throughout your experiment.

In Part I, the role of lasers was broken down. This follow-up article investigates the lenses, mirrors, and filters in your flow cytometer.

To review these three elements…

• Lasers illuminate the stream with coherent, focused light of specific wavelength (energy) and power. This illumination facilitates the generation of fluorescence signals from cells labeled with fluorophores and light scatter signal from redirected laser light.
• Lenses focus laser light and collect light scatter and fluorescence optical signal and direct this signal to the optical detection path.
• Mirrors are responsible for directing light through the detection path and partitioning it so that fluorescence and scattered light are directed to the appropriate detectors.
• Filters, placed in front of detectors, function to restrict the light that is introduced to the PMT detectors so that each detector can be dedicated to measure fluorescence from a specific set of fluorophores.

1. Lenses.

As the lasers interact with particles and cells at the observation point or the interrogation point, scattered and fluorescence light is generated. In order to measure this light, the cytometer needs to collect as much of it as possible. This is the job of lenses.

Howard Shapiro phrases this duty nicely when he says “The lenses provide spatial resolution, enabling us to collect a great deal of light coming from a very small region of space (i.e. the interrogation point) and relatively little of the light coming from other regions a very small distance away.”

In other words, good lenses allow us to collect the light we’re interested in (scattered and fluorescence light) while avoiding irrelevant light (e.g. errant laser light).

The optical collection system of a cytometer must accomplish two goals. First, it must gather as much light as possible from the interrogation point. Second, it must collimate that light so that all rays propagate parallel to each other and can travel through the collection path without diverging. The lenses on a cytometer are designed to do these two things and do them well.

The collection lens system, which usually consist of multiple lenses, is placed directly in front of the interrogation point. Collimating lenses may be positioned some distance away from the collection lens, depending on the optical design of the cytometer.

The collection lens dedicated to the detection of both fluorescence and side scatter signal is typically positioned 90° relative to the angle at which the laser beam interacts with the stream.

Forward scatter signal, on the other hand, is collected at 180° relative to the angle that the laser hits the stream (the FSC path is in front of the laser, looking at it straight-on).

Figure 4 below illustrates a generalized configuration of the optics at the interrogation point, as seen from above, looking down at the cytometer.

One prominent feature of the forward scatter detection system is the obscuration bar. This device prevents laser light from hitting the forward scatter detector. Because of its position, the forward scatter collection paths “look” directly at the laser beam. If there wasn’t any device that blocked non-scattered laser light, any relevant forward scatter signal that found its way to the FSC detector would be entirely drowned out by laser light.

The obscuration bar is a horizontal piece of metal that blocks laser light but allows scattered light to pass over it and into the detector.

Figure 5 below shows how the forward scatter obscuration bar interacts with laser light.

Many cytometers use optical fibers to direct collected light to the fluorescence and side scatter detection system. In these types of systems, the output of the collection lens is focused on the ends of fibers, which are routed to the detection path. This can be very advantageous in overall cytometer design.

Detection paths can be integrated in spaces in the instrument that they would not be able to otherwise thanks to the flexible path that fibers offer.

In some systems, the lens and the fibers are directly coupled using optical gel which may minimize light loss due to refraction.

As light passes through different types of mediums (water, quartz, and air), it bends at the media interfaces. The degree to which this occurs depends on the difference in refractive index between the two mediums: the greater the difference, the more refraction occurs. By coupling the lens, which is typically glass or quartz, to material with a similar refractive index, like gel, there may be less loss as light transitions between the mediums.

The downside of gels is that they can crack and uncouple the lens from the fibers, which will prevent most collected light from entering the fibers and require a service engineer to repair.

Some cytometers use optical fibers to deliver lasers to the interrogation point. This strategy also provides a space-saving benefit in terms of where the lasers can be positions in the instrument. However, a downside to this approach is that there can be significant power loss between the laser output and the interrogation point as laser light travels through the fiber.

Additionally, fibers are not compatible with higher energy light, especially UV wavelengths, which can degrade the material of the fiber over time and require frequent replacement.

2. Mirrors.

Once light has been collected and collimated from the interrogation point, it must be partitioned by wavelength so that each detector can be dedicated to the measurement of a specific spectral band.

Again, Shapiro phrases this very elegantly: “Optical filters (and mirrors) provide spectral resolution, allowing discrimination between scattered, fluorescent, and background light.”

Mirrors generally direct and partition light through the detection path while filters are placed directly in front of each detector to ultimately determine the band or wavelength range of light that interacts with that detector.

Dichroic mirrors are pieces of glass that are coated on one side with a material that allows light above or below a certain wavelength to pass through while reflecting the rest. Placed at 45° relative to the direction of incident or oncoming light, dichroic mirrors come in longpass and shortpass flavors. A 600 LP (longpass) mirror, for example, reflects light shorter than 600 nm while allowing light longer than 600 nm to pass through. A 600 SP (shortpass) would do the opposite.

The activity of a dichroic mirror is best illustrated using a graph of percent transmission (how much gets through) as a function of wavelength.

Figure 6 below is a transmission graph from the product information of a 590 LP mirror manufactured by Chroma Technologies, one of the primary manufacturers of optical filters and mirrors used in flow cytometry.

https://www.chroma.com/products/parts/t590lpxr#tabs-0-main-2

At wavelengths below 600 nm, the transmission of light drops off precipitously using this 590 LP mirror. At 590 nm, transmission is 50% and continues to drop quickly. All of the non-transmitted light is reflected.

Dichroic mirrors are positioned in the optical detection path so that the coated surface faces the incident beam of light. You may be wondering what the effect would be if the mirror were installed backwards, so that the uncoated side were facing the direction of oncoming light. Interestingly, it probably wouldn’t have much effect at all.

The light that passes through the mirror would not be affected. However, the reflected light may bend slightly by the time it is reflected. If the filter is installed “backwards”, the incident light would travel through glass twice — once to reach the coating and once after it is reflected from the coating —which may result in some refraction.

In practice, this usually has little impact on fluorescence measurements. Certainly, installing dichroics backwards isn’t recommended but it is interesting that the effect of doing so is not as severe as it may seem.

3. Filters.

Filters are pieces of glass coated on both sides that allow light of a certain collection, or band, of wavelengths to pass through while absorbing or interfering with photons of other wavelengths.

These come in bandpass, longpass, and shortpass flavors. Bandpass filters are the ones that are most commonly used in flow cytometry. Positioned in front of the detectors, these components determine what collection of wavelengths, and ultimately which fluorophores, will be measured by each detector. Bandpass filters are named based on the center and width of the band of light that will pass through.

For example, a 525/50 filter allows light to pass that is of a range of wavelengths of 500-550 nm (525 +/- 50 nm). Note that the entire band’s width is 50 nm — the range is not 525+/-50 nm but is 525+/-25 nm (25 nm on either side of the center wavelength).

Figure 7 below illustrates the transmission curve, from Chroma Technologies, of a 525/50 bandpass filter.

https://www.chroma.com/products/parts/et525-50m

As shown, the transmission of this filter drops almost asymptotically at 500 nm and at 550 nm. This particular filter, given its transmission band, is ideal for measuring fluorescence of FITC, GFP, or any other fluorophore with similar emission spectra.

Finally, Figure 8 integrates both dichroic mirrors and bandpass filters to illustrate how they cooperate in a detection path. The arrows represent the direction of light as it passes through the path.

One final comment about filters. While they are, for the most part, very good at letting relevant light pass and keeping out irrelevant light, there are certain circumstances when the wrong kind of light — especially laser light — can sneak past the guards and sabotage detection.

This is most typically a problem in the “PE” channel measured off the 561 nm laser.

This channel’s bandpass’ center is usually centered around 575-590 nm and its longer (wavelength) edge can be precariously close to 561 nm. There is some variability in filters as well, that result in laser light leakage.

Finally, filters are only able to block up to a certain point. If enough light — say, high-powered laser light — is directed on them, a certain proportion of that light will pass through. The effect of all these extraneous photons can be severe.

Excessive background light in a detector can cause a drastic loss in sensitivity. If measuring 8-peak beads under conditions of high optical background, you will see both the dim peaks much higher on the scale than they would be otherwise, as well as merging of peaks.

Knowing this can be useful for troubleshooting. If you are having trouble resolving a population in a channel, especially one close to a laser line, it may be worth investigating a laser light leakage issue into that channel. 8-peak beads can be a useful first-line diagnostic tool in this regard.

This article outlined some of the major components of the optical systems used in flow cytometry. While it is certainly possible to explore these topics in much more depth, what is presented here should provide enough insight to understand what happens before a signal is produced from the PMT detectors. Additionally, this article may also equip you with a knowledge toolkit that can help troubleshoot problems you may encounter when performing your next cytometry experiment.

To learn more about flow cytometry system components and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.