Multiplex flow cytometry experiments need the right combination of fluorophores

(See a list of the products featured in this article)

Advances in both flow cytometry reagents and instrumentation allow researchers to run increasingly complex multicolor experiments. The advantages of multiparameter flow cytometry include the ability to perform single-cell interrogation with multiple markers, to correlate cell data using multiple analytes, and ultimately to more accurately define cell populations (Figure 1). Additionally, multiparameter experiments can improve efficiency by requiring fewer samples and smaller sample volumes and by increasing sample throughput. Increasing the number of colors and antigens detected, however, increases the complexity of the experimental design, requiring significantly higher attention to optimization, controls, and other details [1].

One of the biggest challenges in multiparameter flow cytometry is selecting the right combination of fluorophores and antibody conjugates so that the need for compensation and spillover adjustments is kept to a minimum while the quality and accuracy of the data are not compromised. There are some excellent resources available for the beginner, including the Molecular Probes flow cytometry webinar “Multicolor Flow Cytometry Panel Design” by Dr. Holden Maecker of Stanford University. Additional publications on this topic are available [2–4].

When designing a multicolor flow cytometry panel, there are several key points to consider:

  • Know the configuration of the instrument being used (laser and filters) before you begin.
  • Use a tool like the Molecular Probes Fluorescence SpectraViewer to visualize the spectral overlap of fluorophores.
  • Titrate and optimize each antibody; building the right panel is an iterative process.
  • Use bright fluorophore labels on antibodies for low-abundance antigens and dim fluorophore labels on antibodies for highly expressed antigens.
  • Use fluorophores that are spectrally similar for different cell subpopulations that will be gated and analyzed separately.
  • Include a cell viability dye in the panel to exclude dead cells and debris from the data.

Figure 1. Ten-parameter immunophenotyping of human peripheral blood mononuclear cells (PBMCs) with the Attune NxT Acoustic Focusing Cytometer. Lymphocytes and monocytes were gated based on forward- and side-scatter profiles (A). Within the lymphocyte gate, T cells can be isolated based on their expression of CD3 (B), and further subdivided into CD4+ and CD8+ subpopulations (C). In addition, regulatory T cells express CD4 and CD25 (G) and are important mediators of dominant peripheral tolerance. CD62L identifies naive CD4+ and CD8+ T cells (TN), whereas HLA-DR is expressed by activated T cells (TA) (D, H). Conventional dendritic cells found in peripheral blood are generally negative for T and B cell lineage markers and co-express the integrin CD11c and HLA-DR (F). Monocytes are located just above lymphocytes in the scatter profile (A), and express both CD14 and CD33 (E).

Determining fluorophore brightness

In flow cytometry, fluorophore brightness is a function not only of the quantum yield and extinction coefficient of the fluorophore itself, but also of the effects of background contributions. Background fluorescence— e.g., from nonspecific staining, cellular autofluorescence, and instrument noise—can affect the ability to resolve the fluorescence of the antibody conjugate–stained cell population (positive) from that of the unstained cell population (negative).

The signal-to-noise ratio (S/N) is one measure of the sensitivity of an assay and its ability to detect differences between stained and unstained populations. To calculate a simple S/N, divide the median fluorescence intensity (MFI) of the positive cells by that of the negative cells (Figure 2).

However, the relative brightness of a fluorophore-conjugated antibody is determined not only by the intensity difference between stained and unstained cells, but also by the intensity distribution spread of the unstained cell population. Proposed by Maecker et al. [2], the Stain Index (SI) takes these two parameters into account (Figure 2). The SI can be useful for comparing histograms of cell populations stained with different fluorescent conjugates of the same antibody (Table 1, Figure 3).

Figure 2. Comparison of Stain Index (SI) and signal-to-noise ratio (S/N). An illustration of two fluorophores with the same S/N but different SI due to different widths of the negative peak (narrow W1 vs. wide W2). Because the width of the negative peak affects the separation of the positive and negative signals, SI is the preferred statistic when comparing fluorophore brightness.

Table 1. Staining Index for different fluorophore conjugates of an anti-CD4 antibody (clone 53.5).

Brightness Fluorophore component of conjugate (Conjugate Cat. No.) Ex max* Em max* Laser line BP Em filter† Stain Index
High APC (MHCD0405) 645 660 633 660/20 200.31
PE (MHCD0404) 496, 565 575 488 585/42 158.46
APC-Cy5.5 (MHCD0419) 650 690 633 710/50 108.97
PE-Cy5.5 (MHCD0418) 496, 565 690 488 695/40 105.91
Alexa Fluor 488 dye (MHCD0420) 495 519 488 525/50 91.72
Medium PE–Alexa Fluor 610 dye (MHCD0422) 488 628 488 620/10 70.71
FITC dye (MHCD0401) 493 525 488 525/50 56.40
PE-Cy7 (MHCD0412) 496, 565 774 488 780/60 53.70
PE–Alexa Fluor 700 dye (MHCD0424) 496, 565 723 488 720/30 52.45
TRI-COLOR dye (PE-Cy5) (MHCD0406) 496, 565 670 488 695/40 50.31
PE–Texas Red dye (MHCD0417) 496, 565 613 488 695/40 40.85
Qdot 605 nanocrystal (Q10008) 350 605 405 605/20 35.17
APC–Alexa Fluor 750 dye (MHCD0427) 645 775 633 780/60 31.91
Alexa Fluor 700 dye (MHCD0429) 696 719 633 710/50 24.85
Low Qdot 655 nanocrystal (Q10007) 350 655 405 655/20 20.62
Qdot 705 nanocrystal (Q10060) 350 720 405 720/20 18.38
Pacific Blue dye (MHCD0428) 410 455 405 450/50 14.61
Alexa Fluor 405 dye (MHCD0426) 401 421 405 450/50 10.01
PerCP (MHCD0431) 482 675 488 695/40 8.75
Pacific Orange dye (MHCD0430) 400 551 405 585/42 6.06
*Approximate fluorescence excitation (Ex) and emission (Em) maxima for conjugates, in nm. BP Em filter = bandpass emission filter, in nm. Staining Index was determined on the BD™ LSR II Flow Cytometer with FACSDiva™ version 6.1 software.

Figure 3. Representative histograms for cells stained with anti-CD4 antibody conjugates. Ammonium chloride–lysed human whole blood was used to evaluate the performance of 20 different mouse anti–human CD4 antibody direct conjugates (see Table 1). Each conjugate was titrated and optimized to produce a maximum signal-to-noise ratio. Cells were analyzed on a BD™ LSR II Flow Cytometer with FACSDiva™ version 6.1 software. Histograms represent 10,000 cells collected in a lymphocyte gate: (A) high brightness from anti-CD4 antibody, APC conjugate; (B) medium brightness from anti-CD4 antibody, PE–Alexa Fluor 700 conjugate; (C) low brightness from anti-CD4 antibody, Pacific Blue conjugate. The Stain Index (SI) for each conjugate is listed in the left corner of the plot.

Online tools: Fluorescence SpectraViewer and Flow Cytometry Panel Builder

The Molecular Probes Fluorescence SpectraViewer is an online tool that displays the excitation and emission spectra for fluorescent dyes and proteins, facilitating selection of appropriate dyes for your multicolor experiment. You can enter your instrument laser and filter configuration (Figure 4), then select the fluorophores under consideration. Figure 5 shows an example of a five-color panel. An additional feature of the SpectraViewer is the spillover table function, which shows fluorescence overlap (or spillover) for each dye in each channel (Table 2). You can find the Fluorescence SpectraViewer at thermofisher.com/spectraviewer.

Our Invitrogen Flow Cytometry Panel Builderbeta can help you choose fluorescent antibody conjugates for your flow cytometry panel. This online tool guides you through flow cytometry panel design, providing a simplified, customizable experience to fit your flow cytometry panel design needs. Similar to the SpectraViewer it displays the excitation and emission spectra for fluorescent dyes and proteins, but also shows available antibody conjugates to facilitate selection of appropriate antibodies for your multicolor experiment.

Overview of Molecular Probes SpectraViewer components

Figure 4. Overview of Molecular Probes SpectraViewer components. (A) Excitation spectrum and (B) emission spectrum for the same fluorophore. (C) Laser excitation wavelength. (D) Bandpass emission filter wavelengths.

 
Five-color panel depicted on the Molecular Probes SpectraViewer

Figure 5. Five-color panel depicted on the Molecular Probes SpectraViewer. Emission curves for Pacific Blue dye (purple) excited by the 405 nm laser, LIVE/DEAD Fixable Yellow stain (yellow) excited by the 405 nm laser, R-phycoerythrin (PE, orange) excited by the 488 nm laser, Alexa Fluor 647– PE (red) excited by the 561 nm laser, and Alexa Fluor 750–allophycocyanin (brown) excited by the 633 nm laser. Lasers shown are 405 nm (violet), 488 nm (blue), 561 nm (yellow), and 633 nm (red).

Table 2. SpectraViewer spillover table for five fluorophores (Figure 5) with emission filter configurations found on the Attune NxT cytometer.

Fluorophore Bandpass (BP) emission filter (nm)
440/50 574/26 670/14 780/60
Pacific Blue dye 48.2% 0.9% 0.0% 0.0%
LIVE/DEAD Fixable Yellow stain 0.2% 20.3% 2.2% 0.0%
PE (R-phycoerythrin) 0.0% 53.7% 1.2% 0.0%
Alexa Fluor 647–PE 0.0% 3.3% 28.9% 0.1%
Alexa Fluor 750–allophycocyanin 0.0% 0.0% 2.9% 66.1%
Percentages represent the relative fluorescence signal detected with the indicated emission filters (independent of excitation light source). The 574/26 nm BP filter collects 20.3% and 53.7% of the LIVE/DEAD Fixable Yellow stain and PE fluorescence, respectively. However, there should be no actual spillover because the LIVE/DEAD stain is excited at 405 nm but not 488 nm, and PE is excited at 488 nm but not 405 nm; i.e., they will be excited using different lasers.