Tips: Sample Prep

If you work with tissues, disaggregation steps can be mincing, pushing through steel mesh, rubbing between frosted glass slides, or more standardized commercial sample preparations systems. Using enzymatic treatment is also common for tissues, and adherent culture cells might require enzymatic treatment as well. Enzymatic treatment might make it necessary to test the sensitivity of the epitopes to the enzymes. 

Cell clumping should be reduced by filtering the samples already prior to the staining procedure, and then again immediately prior to analysis/sorting. This can be done using FACS tubes with cell strainer caps (35µm), or cell strainers fitting onto Falcon tubes (various sizes), or filters that can be attached to the pipette tip. The mesh size needs to be selected keeping the target cell size in mind!
Please see Sample Buffer for information on which buffer to use to keep cells from sticking together during staining and analysis/sorting.

Cell clumping will result in poor staining, and in reduced cell counts during analysis/sort (because these aggregates are excluded by the software algorithm or later during doublet discrimination gating), poorer resolution of the data, and poor sort purity and yield.

Reagent selection must begin with knowing the instrument configuration and thereby knowing which fluorophores can be used. The available fluorophores should then be ranked according to their brightness. Many vendors specify the brightness of their fluorophores.

Thermo Fisher Fluorophore Selection
BD Brightess Chart
BioLegend Brightness Index

When the fluorophores have been selected, they are matched with the antibodies targeting the antigens of interest. In general, the brightest fluorophores are reserved for the dim/least abundant antigens, and vice versa. However, at the same time, clever panel design needs to avoid spillover from bright populations/dye-antigen-combinations into detectors that require high sensitivity for the same population (due to data spreading; please also see the SSM matrix available for our LSRFortessa on PPMS).
In order to determine the spillover, many vendors provide tools to overlay various fluorophore spectra. As you can define the excitation source and the bandpass filters, the most handy of those tools will give you a rough idea of how the spectral overlap will look like with the given instrument configuration (not taking into account how the detector voltages are set etc.).  

Thermo Fisher Invitrogen Spectra Viewer
BD Spectrum Viewer
BioLegend Spectra Analyzer
Omega Curvomatic 
Chroma Spectraviewer (includes many fluorescent proteins) 
Fluorescent protein database

The Panel Design process can be long and painful, which has brought up companies that offer help, such as Fluorish or  FluoroFinder (free for academic users). I have listed the instruments of the Flow Cytometry Core Facility BMC (CFFlowCyt) and their configuration on FluoroFinder, facilitating the panel design for you. 

Always use a viability dye to exclude dead cells. When staining cells with conventional dead cell exclusion dyes (targeting DNA), e.g PI, 7-AAD or DAPI, add the dye just before running your stained sample. These stains cannot be fixed.
There is a broad range of new viability dyes available for many fluorescence channels that often have a more narrow emission range compared to traditional dyes, such as the Sytox Dye range (Thermo Fisher) or the Zombie Dye range (BioLegend). 

There are also fixable versions (Amine-reactive dyes), for example Fixable Viability Dye eFluor 450 and eFluor780 (eBioscience/Thermo Fisher) or the Fixable Viability Stain (BD Horizon) range. Stain the cells with these viability dyes using the manufacturer's recommended buffer before surface staining. 

The antibody manufacturers provide a specific concentration to use of each antibody, but this may not be the appropriate concentration for a given assay. Too much antibody will lead to increased non-specific binding and increased background, while too little antibody leads to insufficient separation of populations. Antibody titration seems like additional work, but it will in fact safe time on post-acquisition troubleshooting and can safe money on antibody in the long run.
To perform an antibody titration, the cell concentration, total reaction volume and incubation parameters are kept constan, while only the antibody concentration is changed. The best concentration is identified by calculating the Staining Index (SI) (taking into account the separation of neg/pos, and spread of the neg population) for each condition.

To determine and account for the spectral overlap (meaning the 'spillover' of a fluorophore into a channel that is designated to detect another fluorophore) when using multi-color panels, single-stained controls are needed. There are several rules for a good compensation:

1. Compensation control samples must show as bright, or brighter fluorescence than the real samples. Markers such as activation markers, cell subset markers, cytokines etc. can show a low antigen abundance.
2. The fluorophores in the compensation controls and the real samples have to be identical. If using GFP in the real sample, compensation cannot be done with FITC etc.  When using tandem dyes, the antibody used needs to be exactly the same (due to lot-to-lot variability and time-dependend disintegration of the tandems).
3. A considerable fraction of events needs to be positive for the respective marker. This can cause problems when using cells as compensation controls and is sometimes circumvented by switching the fluorophore to another antigen, for example compensating a CD34-PE staining with CD45-PE antibody. This has to be done with caution! The antigen used for compensation must always be higher expressed than the original one (see rule 1). In additon, this approach is not recommended for tandem dyes.

Using compensation beads is the method of choice for many applications. These beads bind antibodies of multiple isotypes and species and result in a bright positive signal (UltraComp beads, eBioscience/Thermofisher = broad range of species; Comp Beads Plus, BD = species specific). Even beads coupled to GFP are available now (Spherotech). The amount of antibody used on beads can be reduced to up to 1/10 of the amount used on cells. Plus, one can save the precious sample. 

Below you can download a Technical Bulletin by BD, explaining some basics about compensation and setting up compensation controls in BD FACSDiva software. Also FlowJo provides information in their Biologist's Guide to Spectral Compensation.  
Come talk to the staff if you need further advice on how to compensate your panel. 

The Fluorescence-Minus-One (FMO) control is a control sample necessary for proper interpretation of flow cytometry data. It is used to determine positivity and accurately gate populations, allowing to control for autofluorescence and data spread due to spectral overlap / applied compensation.
An FMO control contains all the fluorochromes in your staining panel, except for one. And this (strictly seen) needs to be applied for each fluorophore. For example, in a 6-color antibody panel, there would be 6 separate FMO control samples. In reality, it depends on the quality of the overall panel and the particular antigen if an FMO control is necessary for accurate gating. Antigens such as CD3, CD4, CD8 etc. when combined with the right fluorophore result in very clear separation of negative and positive population and don't require an FMO control. But as soon as one is interested in low-abundance markers (such as activation markers) or dim fluorophores, FMO controls for these markers are necessary.
The same holds true for fluorescent protein expression - non-transfected/mock-transfected cells can be used as FMO control for the respective signal, and as biological control at the same time.

For any service analysis or service sorting, talk to the staff in advance about the necessary FMO controls.

Different antibody isotypes may have different non-specific binding affinity to cells, which has lead to researchers using isotype matched control antibodies to determine positivity. Isotype controls are not any longer considered proper gating controls, being replaced by FMO controls in the case of extracellular stainings. However, they are still helpful in revealing potential blocking problems and the 'stickiness' of a cell type in specific cases (for example when performing intracellular staining).

For further reading and insight into the ongoing discussion about the (lack of) usefulness of isotype controls versus FMO controls: 

ExpertCytometry Blog: https://expertcytometry.com/when-to-use-and-not-use-flow-cytometry-isotype-controls/
Stelzer et al. https://www.ncbi.nlm.nih.gov/pubmed/9383095
Hulspas et al. https://www.ncbi.nlm.nih.gov/pubmed/19575390
Maecker et al. https://onlinelibrary.wiley.com/doi/full/10.1002/cyto.a.20333

As with any other best-practice experiment, in addition to the staining controls you need a variety of biological controls for your flow cytometry experiment depending on the application. Here are a couple of examples:

• For transfected cells expressing fluorescent proteins, un-transfected cells that have undergone the same treatment are necessary as controls.  
• For uptake experiments (phagocytosis etc.), controls must include not only the cells incubated without the particle to be taken up, but also incubated at 4°C together with the particle. Only this allows to control for the residual fluorescence in the sample due to unbound or non-specifically bound fluorescent particles. 
• For activation markers, cytokine expression and the like, un-stimulated cells are needed as control in addition to isotype controls.

Come talk to the staff if you are not sure what else to include as controls in your experiment. 

We recommend to use PBS supplemented with 1% FBS or 0.5% BSA. Using a lower concentration of protein can reduce optical interference and ehance resolution. Adding 2mM EDTA to the buffer is necessary to prevent formation of aggregates, in particular for larger sample volumes, primary cells or sorting applications.

The high pressure during sorting can cause the sort buffer to become basic. You can add HEPES buffer (final conc 25mM) to maintain a neutral pH. There is also the possibility of adding 25-50ug/ml of DNAse I and 5mM magnesium chloride hexahydrate if cells are clumping due to cell death.

Do NOT use cell culture medium as sample buffer. Wash cells well after staining and avoid phenol-red containing media for culture if increased sensitivity is needed.