One of the most important considerations for flow cytometry of lung cells is how best to obtain single-cell populations for analysis. There are an estimated 40 cell types in the lung including mesenchymal lineages (vascular endothelial cells, lymphatic cells, pericytes, and fibrocytes), hematopoietic lineages (dendritic cells, lymphocytes, macrophages, and neutrophils) and epithelial lineages (neuroendocrine, basal, club, ciliated, goblet, type 1 and type 2 alveolar epithelial cells) along with several putative progenitor cells recently discovered [2, 3, 10, 11]. This diversity is further compounded by the multiple different compartments that make up the lung such as the bronchoalveolar space, vasculature, lymphatics, interstitium, and epithelium. These compartments can further be partitioned in respect to spatial dimension (i.e., proximal vs. distal) [12].

Obtaining cells for flow cytometry from certain lung compartments is easier than others. For example, cells from a bronchoalveolar lavage (BAL) consistent of mainly myeloid cells already in single-cell suspension. Cells can be obtained from this compartment by inserting a cannula via the airways and instilling a solution such as saline into the airspace and aspirating back to lavage the lung and obtain the cells located at the alveolar space. This compartment has been the most extensively studied compartment in both animal models of acute lung injury and human flow cytometry experiments due to the availability of obtaining cells for analysis [12].

Other areas in the lung such as the interstitium, vasculature, and epithelium must first be either enzymatically digested or mechanically dissociated to obtain cells in suspension [12–15]. In some instances either method may have similar yield of cell types [16]; however, with enzymatic digestion the protease(s) used may cleave proteins of interest affecting identification and analysis [12, 13, 17]. This is an important aspect that should be considered when initiating a flow cytometry study and should be evaluated for each molecule of interest. For example, enzymatic digestion with one protease such as Dispase could disrupt antibody specific recognition of a particular surface receptor compared to another protease such as collagenase (Fig. 7.1). Table 7.1 provides a list of references describing the numerous methodologies for isolating and identifying specific populations from the lung and its different compartments before, during, and after injury.

Signal intensity is directly related to the voltage applied to the PMT detector. For most applications the voltage is set to place the nonfluorescent cells (negative population) within the lower part of the measured scale. The intensities measured by the PMT are recorded and can be displayed in either a linear or logarithmic scale. The linear scale is generally used for depicting forward scatter (FSC, i.e., size), side scatter (SSC, i.e., granularity) or DNA content. A logarithmic scale is utilized for fluorescent marker applications given the wider order of magnitude with fluorescent signals [18].

There are several ways to present flow cytometric data. One of the more common methods is by displaying data for one parameter as a histogram in respect to measured fluorescence intensity (Fig. 7.2a). This graph displays the intensity of a signal (fluorescence) on the x-axis with the number of events (cells or count) on the y-axis, and can be used to obtain quantitative data about a specific marker. Two or more populations from different groups can be displayed by superimposing individual histograms assigned as different colors (Fig. 7.2a; red—whole lung single cell suspension versus blue—trachea single cell suspension only).

A second common method is by a dot plot to display information for two parameters. This type of presentation graphs the two parameters for each single cell and displays the data as a single dot (Fig. 7.2b; fluorescence intensity vs. SSC). Two parameter acquisition is often used to identify and measure subpopulations from a larger population in the single-cell suspensions. These subgroups are often further evaluated in histogram plots for analyses. The FSC versus SSC dot plot is frequently the starting point for most flow cytometric data analysis. The FSC versus SSC dot plot can provide morphological data for cell samples and identify subpopulations such as lymphocytes, neutrophils, and macrophages such as in a mouse bronchoalveolar lavage sample (Fig. 7.3).

Areas of signal intensity can be identified and subdivided by selecting specific areas on the data plot (region gating) either in one parameter histograms or two parameter dot plots. The other setting to provide information on subpopulations is gating. A gate is assigned by the computer after a region is marked. A region can only include one or two parameters; however, a gate can encompass numerous regions, and can be used to select for a smaller specific population or subset within the sample passed through the cytometer.

One important step in data acquisition is to discriminate against events where two or more cells have been interrogated by the cytometer and measured as a single event. This can lead to misleading data. Evaluation for cell doublets (two cells attached to one another) is performed by measuring the cell’s area versus either the cell’s peak width or height and plotting both parameters against one another. This gating strategy allows for identification of a single spherical cellular subpopulation, allowing for only apportioning just the single cells population to be further analyzed. An example of doublet exclusion is provided in Fig. 7.4 with the gating strategy demonstrated for identification of single cells by measuring FSC area vs. FSC height to discriminate against events where two or more cells have been interrogated by the cytometer simultaneously.

Most fluorochromes after excitation can emit light over a range of wavelengths, and this spread can lead to signal overlap in data analysis when more than one fluorochrome is used. A flow cytometer can select out specific wavelengths before analysis by the PMT by passing the light emitted by the fluorochrome through a bandpass filter covering a specific wavelength range; however, even with these filters in place the PMT may register the overlap of undesired emission spectra from other fluorochromes. This overlap can result in the data being susceptible to incorrect assumptions [19]. Therefore, when more than one fluorochrome is used it is important to take into account the potential overlap of all the fluorochromes’ emission spectra into the desired fluorochrome’s spectra that will be measured by the PMT. Correcting or subtracting out this overlap between emission spectra must be performed for each experiment. One method is by performing single fluorochrome (color) staining of cells or synthetic beads and measuring overlap or spillover into the other PMT channels being used in the experiment. Using beads which can bind antibodies instead of cells is another method for compensation analysis and can avoid issues surrounding potentially low marker-density on a single-stained generic cell. A compensation sample will be needed for each fluorochrome used in an experiment (i.e., seven fluorochromes used necessitates seven individual tubes containing only one of the seven fluorochromes). Therefore one fluorochrome’s (single-stained positive control) emission spectra can then be recorded by all the detectors in use for the experiment (primary detector and then the other peripheral detectors). Subtracting out the percentages of signal registered in the primary detector from the signal in the peripheral detector is termed compensation. Importantly, compensation values depend on the PMT voltages values for an experiment, which are set by the user, and any changes in voltages will negate current compensation settings and require them to be determined again. Most compensation algorithms also compare positive singly stained samples with a universal unstained negative cell population. This unstained sample may also assist when cells have background fluorescence (autofluorescence) and this can be identified and compensated for also by measuring fluorescence values of unlabeled (unstained) cells through the flow cytometry as a negative control.

Alveolar macrophages are the most numerous cells in BAL fluid of uninjured animals and humans. One characteristic relevant to flow cytometry of lung injury and repair is that this cell population can have a high level of autofluorescence (AF) [12]. AF is defined as natural emission of light from absorbed light by biological structures. Common AF molecules include: NADPH, flavin, collagen, lipofuscin, and metallo-proteins [12]. While AF can help identify a population, more often AF can complicate flow analysis. AF of alveolar macrophages occurs at an excitation wavelength of 488 nm which is a standard laser in cytometers, and can obfuscate signals from fluorescence dyes such as fluorescein isothiocyanate (FITC) or phycoerythrin (PE). Smoking can increase AF in alveolar macrophages [20]. AF of alveolar macrophages is lower at higher wavelengths used in flow cytometry, and fluorochromes such as allophycocyanin (APC) that are excited and emit light at higher wavelengths can assist in overcoming the complication of AF [12, 21]. Other methods include the use of tandem dyes or the use of quenching dyes (crystal violet or trypan blue) to reduce AF; however, the later requires the cells to be fixed and permeabilized before their application [12, 22, 23].

Some cells types express Fc receptors (CD16, CD32, CD64) such as macrophages and monocytes. Fluorochrome-conjugated antibodies can bind to both the specific antigen epitope and nonspecifically to Fc receptors. Blocking Fc receptors with an antibody specific for Fc receptors before application of the fluorochrome-conjugated antibodies can help negate this effect and increase antibody specificity. Fc receptor blockade is most useful when determining signal backgrounds with isotype fluorochrome-conjugated antibody controls so as to block nonspecific binding of the isotype antibody to a cell’s Fc receptors [8]. Without blocking this potential interaction there can be a falsely high estimation of background fluorescence to fluorochrome fluorescence signal. This application step is also useful if cells are prepared in protein-free buffers.

The expansion of novel fluorochromes makes building a multicolor flow cytometry panel challenging. Adequate panel design and antibody and instrument optimization are critical to obtain quality results. There are several principles to consider and include familiarizing with your instruments (e.g., lasers, PMT/filters, etc); using fluorochromes based on their relative brightness (e.g., choosing brighter dyes for lower abundant cell markers), minimizing spillovers and using appropriate biological and staining controls.

Given the potential for nonspecific staining when using multicolor flow cytometry, particularly for low abundant markers, the inclusion of Fluorescent minus One (FMO) should be considered. It represents a sample that contains all the fluorochromes in the panel except the one you are trying to measure (e.g., if a panel contains W, X, Y, and Z; the FMO for Z will only contain W, X, and Y). It is used to appropriately identify and gate cells for that specific marker measured.

Flow cytometric data can be reported as cell-population data such as size and percentage (i.e., dot plot parameters) or may be quantitative based on fluorescent signal analysis (i.e., histograms of a fluorescent signal parameter). Estimating population size and determining the percentage of a population compared to the total cell population are common uses of flow cytometry. Of note, determining absolute numbers of cells is not a design of flow cytometry; however, there are methods to do so such as including fluorescent beads in the sample [24]. The histogram analysis can provide information on the percentage of a population in the region of interest along with quantitative data for the fluorescence emitted in the form of arithmetic or geometric means along with median values. Quantifying the fluorescence signal for a fluorochrome is proportional to the amount of fluorochrome bound to the cell; however, this measurement is dependent on the voltage setting which can be changed in each experiment. Choosing between arithmetic mean, geometric mean, or median values depends on the type of analysis [18, 25]. For log-amplified data using the geometric mean is more common, as it accounts for data distribution while arithmetic mean is used for data displayed on a linear scale. The median value of a population is often used, as the median is less influenced by skewed populations or outliers.