Flow Cytometry Experimental Process—Spectral versus Conventional

Flow cytometry is a powerful research tool used extensively for cell identification and characterization and to establish expression analysis of proteins of interest. Principles of experimental and panel design using spectral flow cytometry are similar to conventional flow cytometry (Figure 1). Each experiment should answer a single question that is based on a clear, testable hypothesis. Determining the biological question and establishing an appropriate experimental design is paramount to generating accurate and reproducible data [1].

Figure 1. Flow cytometry experimental process overview.
 

Once the biological question has been established, the cell or tissue type needed to address the hypothesis should be determined. The biology of the system will determine the biomarkers required and metrics to use, such as the presence or absence of a cell population, relative population percentages, cell counts, or fluorescence intensity. This can usually be achieved with a thorough review of the literature. The review should focus on peer reviewed studies that are similar in nature to the biological system, cell(s), or protein(s) of interest. This will elucidate key considerations for designing and conducting the experiment, as well as any challenges with preparing the sample for flow cytometric acquisition. Evaluating multiple markers, in combination, will provide a more accurate and comprehensive assessment making spectral flow cytometry an ideal tool to provide a complete understanding of the diverse interactions of immune cells and disease dynamics.

During this early part of designing the experiment, it is useful to consult with a statistician or bioinformatic specialist to help in experimental design, sample size determination, and to explore strategies for data analyses to support the research question. At this planning stage, it is also useful to consider what information will be required for the publication of results. Relevant for flow cytometry experiments, the MiFlowCyt requirements checklist for publications, ensures minimum experimental information is adequately provided [2].

Learn more: Immunology at Work Resource Center

The use of flow cytometry requires the sample to be prepared and processed into a single-cell suspension for interrogation. Most samples will require some degree of processing in preparation for antibody labeling. A primary goal of sample processing is to achieve a suspension of live, single cells for labeling. Issues that may arise during sample preparation include cell aggregation (cell clumping), cell death, cellular activation, and changes in epitopes or loss of protein (shedding, internalization). It is important to note that the quality of the sample will dictate the quality of the data being generated.

Thoroughly optimize and standardize sample preparation, especially regarding incubation times, temperature, reagent/buffer selection, and sample processing time. Sample processing temperature should be aligned with the biological system and reagents/buffers being used. In general, processing techniques should be minimally invasive and only serve to prepare the sample for labeling. Following processing of samples, a visual inspection is recommended to determine the quality of the sample and determining a cell count or cell concentration is necessary to quantify cells for subsequent antibody staining and flow cytometry evaluation. There are specific ways to process cells depending on how they are cultured. Below are considerations for non-adherent/adherent cells and tissues:

• Non-adherent cells—Generally, these require minimal manipulation to prepare the sample for flow cytometric analysis. Density gradient centrifugation is a commonly used method for isolation and enrichment of human mononuclear cells, particularly from peripheral blood, cord blood, and bone marrow. A more recent closed system approach relies on counterflow centrifugation for cell isolation which separates cells based on size and density. Another common approach is the removal of red blood cells (RBC) with ammonium chloride-based RBC lysis buffers for human and mouse hematopoietic tissues blood or with multi-species buffer formulation. Even with non-adherent cells, filtering samples through a nylon mesh immediately before sample acquisition is always recommended. This will lower the risk of the cytometer getting clogged.
• Adherent cells—It is possible to use either mechanical means or chemical strategies to remove cells from culture vessel. Cell scraping is designed to physically remove cells from a culture vessel, while enzymes such as trypsin or other dissociation reagents such as EDTA may be used. Use of enzymes must be validated to ensure that protein detection is not altered either by the enzyme itself or by the protocol. After processing adherent cells, small clumps of cells may be present, which may be removed by filtering the cell suspension through a nylon mesh. The addition of DNase and EDTA to the media or removing/decreasing calcium concentration may also help minimize cell clumping. Another technique is trituration, where the cell suspension sample is aspirated back and forth through a small needle several times to disperse cell aggregates. Having single cells in suspension is critical for cell staining. It is preferred to remove as much cell aggregation as possible before running on the cytometer. Filtering samples through a nylon mesh immediately before sample acquisition on the flow cytometer lowers the risk of cytometer clogging.
• Tissues—Dissociation of cells from primary tissue is dependent on the tissue type and can require mechanical methods and/or enzymatic digestion to create a single-cell suspension. Using surgical scissors facilitates the generation of small tissue pieces followed by sample filtration, or tissue dissociating instruments and kits can be used. Many reagents for tissue processing require specific incubation temperatures, which range from 4–37°C and will be dependent on the tissue type and the antigens of interest. Once a single-cell suspension is obtained, the samples can be treated in a similar fashion as non-adherent or adherent cells to prepare for antibody staining. 

With any type of cells or tissue they may need to be isolated, expanded, and cryopreserved over the course of the experiment. Here are some considerations to consider when performing these techniques:

• Cell isolation and expansion—Magnetic separation technology and fluorescence activated cell sorting can both be used to isolate pure, viable, and functional cells. Cell isolation can be used with negative isolation, positive isolation, and depletion to separate cells of interest. The T cell activation and expansion workflow enables the isolation, activation, and expansion of robust T cells while maintaining cell viability for downstream applications.
• Cryopreservation—In some cases, samples may require cryopreservation after processing to store the cells for future use. In the cryopreservation process, the biological function of the cells is preserved by controlled cooling and storing at very low temperatures. When ready for use, thaw frozen cells by placing cryovials in a 37oC water bath. Cells may be used immediately or rested before use. Cell resting may impact protein expression and should be optimized as part of the standardized sample preparation process. Cryopreservation and thawing will alter cell viability compared to freshly prepared cells, highlighting the need to filter the sample and identify/eliminate dead cells during analysis. 

Learn more: Flow cytometry protocols
Explore: Cell isolation and expansion

Initial experimental planning includes identifying antigens of interest and categorizing them as having surface or intracellular expression. Researchers can use previously published literature to help elucidate the combination of antigens that identify the populations of interest. A resource of particular interest is the Optimized Multicolor Immunofluorescence Panel (OMIP), a publication type available in the journal Cytometry A featuring optimized multiparameter panels, for both conventional and spectral flow cytometry, to help other researchers in developing multicolor panels [8]. Another useful tool found in Cytometry A is a mini review called the ‘Phenotype Reports’ designed to help researchers identify different cell types [9].

Learn more: BioProbes 74: Optimized Multicolor Immunofluorescence Panels (OMIPs)
Explore: Optimized flow cytometry multiplex panels

An objective of antibody-based research is the selection of a suitable reagent for a specific target and application. Antibodies represent a critical tool in basic science research and immunophenotyping is used across many platforms including flow cytometry. Three types of antibodies can be defined.

• Polyclonal antibodies—These comprise a heterogenous mix of antibodies produced by multiple B cells, each one recognizing a different epitope on the same antigen. Because polyclonal antibodies are not specific to a single epitope, there is greater potential for cross-reactivity.
• Monoclonal antibodies—These are produced from a single B-cell parent clone and therefore recognizes a single epitope per antigen. These B-cells are immortalized by fusion with hybridoma cells which produces long-term generation of identical monoclonal antibodies. Because monoclonal antibodies specifically detect an epitope on the antigen, they are less likely to have cross-reactivity with other proteins.
• Recombinant antibodies—Recombinant antibodies are generated using antibody coding genes and are often comprised of only the heavy and light chain variable region (the portion of the antibody that binds to the target). Recombinant antibodies have tremendous utility in medicine as the lack of the Fc portion limits immunogenicity. In flow cytometry, this attribute can also be beneficial as these antibodies will be unable to bind to Fc receptors which often contribute to background signal.

To improve reproducibility in flow cytometry experiments, it is generally advised to use monoclonal or recombinant antibodies. Recent publications have noted the lack of universal standards for antibody validation, highlighting the potential for producing false negative, false positive, or inconsistent results [10–11]. Guidelines have been proposed by the International Working Group for Antibody Validation (IWGAV) for improving standards of antibody use, which provide a framework for antibody validation for use in an application-specific manner [12]. For an antibody to provide valid findings in flow cytometry experiments, it must be specific, selective, sensitive, and provide reproducible results at the optimal dilution [13]. Knowledge in the principles of antibody validation allow the researcher to select antibody reagents carefully based on the data presented by manufacturers and developers, and to source antibody reagents from a trusted supplier [14]. Testing of each antibody for both target specificity and functional application will ensure the antibodies bind to the right target and work in the specific application used. Once an antibody supplier provides validation information, it is up to the end user to verify antibody performance in their own setting [15].

Learn more: Guide to primary antibody types

Understanding antigen expression pattern and antigen density level is paramount in developing a robust panel to ensure the greatest resolution of distinct cellular subsets. Antigens have a relative density correlating directly to their amount on or in a cell, often described in the context of flow cytometry as low, medium, and high. Some antigens are highly characterized while the structure and function of others remain unknown. Resources for expression levels for some cell surface proteins have been compiled and provide useful information [16–18]. These properties have led to classification of antigens in a tiered fashion useful for flow cytometry panel design [19]. Once the antigens of interest are known, outlining a preliminary gating strategy will help define the relationships between antigens, expression patterns, and where to maximize resolution (Figure 2).

A.

B.

C.

Figure 2. Gating strategy. Once target populations and markers to identify cells of interest are selected, creating a gating strategy is necessary. This can be simply drawn (A), or you can start by using a lineage tree (B) or follow a published protocol (C). Understanding which markers are co-expressed and which have mutually exclusive expression is essential when pairing markers with fluorophores.
 

Researchers should note the cell populations that are required for the analysis, define lineage markers necessary to identify the cells of interest, and highlight markers that are co-expressed (Figure 3). Antigens and their characteristics can be evaluated at this stage. The types of antigens and characteristics of each are listed below:

• Primary antigens—well characterized and identify major subsets of cells. These are often referred to as lineage markers.
• Secondary antigens—often well characterized and may have high antigen density but may have a continuous or broad expression pattern.
• Tertiary antigens—either expressed at low levels or are uncharacterized.

Figure 3. Determine co-expression. Issues with spread affect situations where cells co-express two or more antigens. Determine which targets are co-expressed and choose fluorophores with low similarity. Reserve use of fluorophores with higher similarity for targets with mutually exclusive expression.
 

Determining the appropriate antibody clone should be considered in the specific assay and the target’s localization and sensitivity to different processing methods. An antibody datasheet from the supplier is a valuable source of information and contains details of antibody clone and isotype, species reactivity, epitope, applications tested, protocols, suggested antibody concentration to use, fluorophore conjugates, and sample data. Here is an example of an antibody data sheet. In addition, web-based platforms can provide valuable comparative information about targets and links to relevant publications [20–21].

(Top)

The configuration and capabilities of the flow cytometer or sorting instrument being used will determine the development of any panel. To determine which fluorophores can be effectively detected, verify the instrument’s optical configuration. This includes checking the lasers available, laser wavelength, and power, if lasers are co-linear or spatially separated, the optical light path, and the number and type of photodetectors capturing the resulting emissions [3]. In spectral flow cytometry, fluorescent signals derived from all available lasers and detectors, generating a more detailed spectral signature for a given fluorophore (Figure 4).

Figure 4. Spectral signature of APC, comparison of one and five laser excitation. (A) Normalized emission spectral signature of the fluorophore APC, with a single laser excitation (as also seen on conventional flow cytometer). (B) Normalized emission spectral signature of APC with five laser excitation (as seen with spectral flow cytometry). The secondary emissions obtained with additional laser excitation provide more detail in the spectral signature.
 

This added information can help distinguish fluorophores with near-identical peak emissions but different off-peak emissions. For example, comparing the spectral signatures of allophycocyanin (APC) and Alexa Fluor 647 dyes only reveals minor differences in their emissions when excited by a single red laser (Figure 5). However, when emission signals are collected from additional lasers including UV, violet, blue, and yellow/green lasers in addition to the red, unique signatures for APC and Alexa Fluor 647 dyes emerge, allowing their combined use in a spectral flow cytometry panel (Figure 5). Some instruments come with preferred instrument settings, often referred to as assay settings. These optimized detector settings aim to balance the best signal resolution across detector arrays while maintaining the spectral uniqueness of fluorophores. Manufacturer-developed instrument settings are adequate for most situations, but in some cases, they may need to be optimized to maximize detection or to prevent signal saturation.

Figure 5. Spectral signature comparison of highly overlapping fluorophores. Overlay of normalized spectral signatures of APC (hash lines) and Alexa Fluor 647 fluorophores (grey) reveal differences between the two fluorophores.

Explore: Considerations when purchasing a flow cytometer
Explore: Attune Flow Cytometers
Explore: Bigfoot Spectral Cell Sorter

After determining the instrument’s configuration, the next step is to characterize dyes on the specific cytometer. Characterizing dyes on the cytometer starts by running multiple fluorophores to see what the cytometer output looks like. It is typical to run a CD4 conjugate (or another well-known target) of every fluorophore available and examine the spectral signature to determine what is compatible. Compatibility would include finding out which fluorophores overlap or what areas in the spectral signature are available to find other fluorophores to use to expand the panel.

For this purpose, on-instrument or web-based spectra viewers can be consulted. These tools commonly display the excitation and emission spectra of fluorophores by plotting normalized intensity in relation to wavelength. Some spectral viewers provide instrument-specific configurations that display the spectral signatures of compatible fluorophores. Because of differences in instrument configuration and detector sensitivity, as well as fluorophore variation, spectral signatures may differ slightly between the data presented on a spectral viewer and what is obtained on a specific instrument. It is beneficial to use spectral viewer and panel builder tools that utilize your instrument configuration. Spectral viewers and panel builder tools can superimpose the spectra of fluorophores and may calculate a similarity index between fluorophores. To see examples of a similarity index please see the Spectral Flow Cytometry Panel Design page.

The similarity index is a comparative feature that quantitatively rates the emission curve overlap of two fluorophores, on a scale from unique to identical. Values close to zero mean the spectral signatures of the two fluorophores are very different, and there is minimal overlap between the two fluorophores under comparison. Values close to 100 indicate the spectral signatures are very similar to each other. It is recommended to avoid the use of fluorophores with similarity indices greater than 98 to limit excessive data spreading and reduce the risk of the inability to unmix the signals. High similarity values between 90–98 may correlate with high spread and should be used with caution, especially if the antigens are co-expressed. When choosing fluorophores for antigens that are co-expressed, we recommend avoiding those that have similarity measurements greater than 70. To learn more about fluorescence please see the Molecular Probes School of Fluorescence - Fluorescence Fundamentals.

Another important consideration when evaluating reagents is the relative brightness of the available fluorophores to which antibodies are conjugated (Figure 6).

Figure 6. Stain index measures relative fluorophore brightness. Freshly isolated PBMCs were labeled with anti-human CD4 antibody conjugated to five different fluorophores. Data were acquired using a 5-laser Aurora spectral cytometer using a lymphocyte gate. Histogram overlay plots of unstained (blue) and stained lymphocytes (purple) show increasing brightness and stain index from left to right.
 

Cells are stained with different fluorescent conjugates of the same antibody under identical conditions on a specific instrument configuration to calculate a stain index or separation index to evaluate fluorophores and rank their relative brightness. Typically, a CD4 antibody is used as it has well-defined positive and negative populations and is available directly conjugated to most commercially available fluorophores (Figure 7) [4–6].

Figure 7. Relative stain indices. Freshly isolated PBMCs were labeled with anti-human CD4 antibody conjugated to various fluorophores, with acquisition using a 5-laser Aurora spectral cytometer using a lymphocyte gate. The stain index of each fluorophore was calculated and used to rank the relative stain indices of all the fluorophores from dim to brightest. Download the staining index for fluorochrome brightness PDF to see all available fluorophores.

As the number of available fluorophores continues to grow, it is essential to understand key fluorophore characteristics. Relative brightness information is used when pairing antibodies with fluorophores in panel design. In general, the pairing of bright fluorophores with low-density antigens and dim fluorophores with higher density antigens is recommended. Another consideration is how unique the spectra of a fluorophore is, the more unique it is, the less spreading will occur. For more a more detailed description of spreading please see the Spectral Flow Cytometry Panel Design page.

Development of new fluorophores with unique spectra and minimal cross-laser excitation are ideal for spectral flow cytometry because they are designed to minimize spectral spillover and spreading, thus maintaining data resolution. If possible, obtaining the autofluorescence signature of the unstained sample, as a first step in the spectral panel design process, can help visualize any regions of the spectrum that should be avoided or favored when selecting fluorophores, therefore maintaining data resolution.

Explore: NovaFluor dyes for immunophenotyping
Explore: Staining index for fluorochrome brightness

In conventional flow cytometry, “spillover” is when the emission from one fluorophore spills into a detector assigned to another fluorophore. When compensation is applied, the spillover is corrected. In spectral flow cytometry, spillover is more accurately described as spectral overlap, and the data are unmixed across the multiple detectors to isolate the unique spectral signature of each fluorophore. Spread refers to a photon counting error which introduces a spread of fluorescence intensity due to inaccuracies in the detectors. Spreading is visible after compensation or spectral unmixing has been applied, resulting in the broadening, or spreading, of the of the positive population seen in the detectors receiving the spillover (Figure 8) [7]. Ideally, only fluorophores with unique spectral signatures without overlapping emission would be used in a panel. However, fluorophores with similar spectral signatures can be used together in a panel, with the understanding that spreading will occur. Taking steps to minimize the impact of spreading may be required.

Figure 8. Spreading revealed after unmixing. Lymphocytes are singly labeled with CD4-PE, with visualization of PE vs. (unlabeled) PerCP-Cyanine 5.5, before (A) and after (B) unmixing has been applied. Data after unmixing reveals significant spreading of the positive PE fluorescence that may compromise detection of co-expressed populations when combined with PerCP-Cyanine 5.5 labeling.
 

Data that has been unmixed using spectral flow cytometry or that has been compensated using conventional flow cytometry will reveal spread of the positive population in detectors receiving the spillover. Unmixing or compensation do not cause the error or change it, as the photon counting error is already present, but it does make the error more apparent by shifting it to the low end of a log-scale. Spreading can be problematic because it decreases sensitivity and resolution and becomes an important consideration as panel size increases. Common causes of spreading occur when fluorophores emit in the same spectral area, have greater similarity, and with very bright fluorophores (Table 1).

Spreading causes		Spreading minimization
Very bright fluorophores		Utilize fluorescence intensity and antigen density information
Highly similar fluorophores		Select fluorophores with unique spectral signatures
Fluorophores with multi-laser excitation		Select fluorophores with narrow excitation
Fluorophores with broad emission		Select fluorophores with narrow emission

The spreading is additive, emission is affected by the contribution from all the fluorophores in the experiment that emit at the same wavelength. While there is no method to remove spreading, a preferred approach is to evaluate spreading during instrument and fluorophore characterization to design panels that minimize the impact of spreading. Using single-stained samples, a spread matrix can be generated [7]. This quantifies the amount of signal spreading for every pair of fluorophores as measured on a specific instrument. The information in a spread matrix helps to identify the fluorophores giving or receiving signal spread. Figure 9 provides an example of a spread matrix, for further information and an interactive tool visit Spectral Flow Cytometry Assays and Reagents.

Figure 9. Spread Matrix. Careful panel design for spectral flow cytometry requires an understanding of the cytometer characteristics and fluorophore properties. This spread matrix of 20 fluorophores shows the level of fluorophore spread, data collected on a 3-laser Cytek Aurora spectral cytometer. Fluorophores in each row impact the spread of the fluorophore in the column. Although all fluorophores in the matrix can be used together, the darker red shading indicates where one fluorophore has increased spread into the other, requiring closer attention when matching targets with fluorophores.

Selecting fluorophores with unique spectral signatures, narrow excitation, and narrow emission properties will help to minimize spreading. Pairing of highly similar fluorophores is recommended only for mutually exclusive populations. Fluorophores that contribute to high spreading can be paired with low expression markers or paired with spectrally dissimilar fluorophores to improve resolution (Figure 10). Once a specific panel is designed, it is recommended that a spreading matrix be evaluated for the fluorophores within the gating strategy for that specific panel.

Figure 10. Spread and resolution. Human lymphocytes were labeled with CD56-APC and two different fluorophores with CD3 to demonstrate spread and resolution. (A) CD3 labeled with APC-eFluor780 causes significant spread into CD56-APC making it difficult to resolve any dual positive CD3+CD56+ cells. (B) Changing to CD3 labeled with eFluor 450 minimizes spread into CD56-APC and allows for resolution of dual positive CD3+CD56+ cells.

Learn more: Labroots webinar: Building to higher dimensional flow cytometry

1. Mair, Florian, and Aaron J. Tyznik. "High-dimensional immunophenotyping with fluorescence-based cytometry: a practical guidebook." Immunophenotyping. Humana, New York, NY, 2019. 1–29.
2. Lee, Jamie A., et al. "MIFlowCyt: the minimum information about a Flow Cytometry Experiment." Cytometry Part A: the journal of the International Society for Analytical Cytology 73.10 (2008): 926–930.
3. Wang, Lili, and Robert A. Hoffman. "Standardization, Calibration, and Control in Flow Cytometry.” Current protocols in cytometry 79.1 (2017): 1–3.
4. Maecker, Holden T., and Joseph Trotter. "Flow cytometry controls, instrument setup, and the determination of positivity." Cytometry Part A: the journal of the International Society for Analytical Cytology 69.9 (2006): 1037–1042.
5. Degheidy, Heba, et al. "Flow cytometer performance characterization, standardization and calibration against CD4 on T lymphocytes enables quantification of biomarker expressions for immunological applications." Journal of Biomedical Science and Engineering 2014 (2014).
6. Wang, Meiyao, et al. "Quantifying CD4 receptor protein in two human CD4+ lymphocyte preparations for quantitative flow cytometry." Clinical proteomics 11.1 (2014): 1–10.
7. Nguyen, Richard, et al. "Quantifying spillover spreading for comparing instrument performance and aiding in multicolor panel design." Cytometry Part A 83.3 (2013): 306–315.
8. Mahnke, Yolanda, Pratip Chattopadhyay, and Mario Roederer. "Publication of optimized multicolor immunofluorescence panels." Cytometry Part A 77.9 (2010): 814–818.
9. Special Section in Phenotype Reports, Cytometry Part A, March 2021, Volume 99, Issue 3, page 216–264.
10. Begley, C. Glenn, and Lee M. Ellis. "Raise standards for preclinical cancer research.” Nature 483.7391 (2012): 531–533.
11. Voskuil, Jan LA, et al. "The Antibody Society’s antibody validation webinar series." MAbs. Vol. 12. No. 1. Taylor & Francis, 2020.
12. Uhlen, M., Bandrowski, A., Carr, S. et al. A proposal for validation of antibodies. Nat Methods 13, 823–827 (2016).
13. Kalina T, Lundsten K, Engel P. Relevance of Antibody Validation for Flow Cytometry. Cytometry A. 2020 ;97(2):126–136
14. Taussig, Michael J., Cláudia Fonseca, and James S. Trimmer. "Antibody validation: a view from the mountains." New biotechnology 45 (2018): 1–8.
15. Bordeaux, Jennifer, et al. "Antibody validation." Biotechniques 48.3 (2010): 197–209.
16. Bausch-Fluck, Damaris, et al. "A mass spectrometric-derived cell surface protein atlas." PloS one 10.4 (2015).
17. Díaz-Ramos, M. Carmen, Pablo Engel, and Ricardo Bastos. "Towards a comprehensive human cell-surface immunome database." Immunology letters 134.2 (2011): 183-187.
18. Kalina, Tomas, et al. "CD maps—dynamic profiling of CD1–CD100 surface expression on human leukocyte and lymphocyte subsets." Frontiers in immunology 10 (2019): 2434.
19. Mahnke, Yolanda D., and Mario Roederer. "Optimizing a multicolor immunophenotyping assay." Clinics in laboratory medicine 27.3 (2007): 469–485.
20. BenchSci AI-Assisted Reagent Selection. Retrieved January 7, 2022.
21. The Antibody Registry. Retrieved January 7, 2022.