When you get a blood test, the final report usually shows you the number of different cell types, such as red blood cells, white blood cells, and platelets.
This can tell you many things, such as whether you have an infection, but what if we want to see if these cells are functioning normally? Furthermore, what if we want to isolate specific cells and do further tests on them?
What is flow cytometry?
For this purpose, a method called flow cytometry is used.
Flow cytometry is a versatile method which can identify, count, and even sort specific cell types. It is widely used in the field of immunology. Immunophenotyping is the most common application of flow cytometry, in which multiple immune cells types are analysed in a cell mixture for different parameters, such as cell surface antigen expression. Different cell types, such as T cells and B cells, have different antigen profiles on their surface, and this can be analysed by flow cytometry.
The instrumental set-up
In general, flow cytometers contain a fluidics system (essentially very narrow tubes) to guide the cell suspension through the flow cell, an optics system used to analyse the cells via scattered light, and an electronics system to convert the light signals into digital signals so they can be read by a computer.
Essentially, a laser hits the sample as it flows through the fluidics system, and the scattered light is captured and processed into readable data. The basic set-up is described below:
Figure 1: Schematic of flow cytometer hardware. The cytometer can contain one or more lasers. Most traditional flow cytometers have at least five lasers that excite at various wavelengths from ultraviolet to far-red. Photomultiplier tubes detect fluorescence from the sample. The filters direct the light towards the detectors.
How does a flow cytometer work?
Firstly, the cell suspension is prepared and stained with fluorophores, which are compounds which produce fluorescence. These fluorophores are conjugated (as in, attached) to a specific antibody which corresponds to an antigen on the cell surface.
Similarly, depending on the goal of the experiment, other antigens can also be stained. Once the staining is complete, the mixture is transferred to a flow tube and put into the cytometer where the fluidics sucks up the cell suspension.
Within the flow chamber or flow cell, the cell suspension is focused into a narrow tube where the cells flow in a single file, where they can then be interrogated by the laser. In this case, “interrogate” refers to the laser hitting the sample cell. This focusing is possible due to the sheath fluid (not shown in the figure) surrounding the cell suspension. Sheath fluid is usually a buffered saline (salt) solution, and this sheath fluid is pressurised so that it can force the cell suspension into a single file. This phenomenon is known as hydrodynamic focusing.
Once the laser hits a single cell, light is scattered. There are two main types of scattering which are measured by a flow cytometer: forward scatter (FSC) and side scatter (SSC) (see Figure 2).
Forward scatter relates to how large a cell is. The larger the cell, the higher the forward scatter. Meanwhile, side scatter relates to how granular or complex a cell is. In essence, the FSC and SSC are the two most basic data points that a flow cytometry generates, and these are usually shown on a dot plot, as seen below in Figure 2.
Figure 2: (a) Forward and side scattering of light by the sample (cell). (b) Dot plot of SSC plotted against FSC. The black “gate” encapsulates the lymphocytes (a type of white blood cell) or other live cells of interest. Meanwhile, dead cells have a high SSC and low FSC, so these remain ungated.
In addition to these two parameters, the cytometer contains several fluorescent detectors (labelled as photomultiplier tubes in Figure 1) which can detect light of different wavelengths emitted by the fluorophores attached to the cells. As mentioned earlier, different fluorophores conjugated to different antibodies are used to stain cells prior to running them on the cytometer. This allows a more complex analysis of a cell mixture.
The beam of scattered light is then directed to these detectors via mirrors and filters. In particular, these filters allow only a certain wavelength to pass through and actually reach the detector. This can be seen in Figure 1, where the scattered light is directed to each photomultiplier tube by a filter (or mirror).
Analysing flow data
Flow cytometry data can be looked at as a dot plot of all the events captured by the lasers. The data is analysed using a “gating strategy”, wherein the cell types of interest are gated to isolate them and then gated further to investigate smaller subpopulations of cells.
Below is an example of a simple gating strategy to identify CD4+ and CD8+ T cells in a cell mixture:
Figure 3: Gating strategy for analysing lymphocyte subpopulations. The cells were stained with a live/dead dye (eFluor 450), and fluorophores conjugated to different antibodies (V710-CD3, G610-CD45, V780-CD8). This figure is courtesy of Flowmetric [link].
In Figure 3, the first panel shows all the data points collected for the sample, with the gate drawn around the lymphocytes.
The second panel contains only data points from the lymphocyte gate, but the axes are changed to plot FSC-H vs FSC-A (forward scatter height and forward scatter area, respectively). These two parameters allow us to separate doublets, in case two cells were analysed as one cell within the flow chamber.
Moving on, the third panel takes all the points in the single-cell gate, plotted as eFluor450 against FSC. The eFluor450 is a dye that stains live and dead cells. A low live/dead fluorescence corresponds to the live cells, while a high fluorescence corresponds to dead cells. Here, the live cells have been gated.
The fourth panel takes all the live cells and plots them as CD3 against CD45. CD3 has been stained with V710 and CD45 with G610 fluorophores. CD45 is expressed on all lymphocytes, while CD3 is only expressed on T cells. In the panel, you can see two clear subpopulations - one which is CD3 high and one which is CD3 low. The CD3 high cells have been gated, and these correspond to T cells.
In the fifth panel, the CD3+ cells (T cells) which were identified have been plotted as CD4 against CD8. Two populations can be seen, one which is CD4 high/CD8 low, and the other which is CD8 high/CD4 low. These correspond to the two types of T cells, helper T cells (CD4+) and cytotoxic T cells (CD8+).
In a similar manner, by using other dyes and antibodies, more populations can be identified such as B cells or natural killer cells.
The applications of flow data
So, what can we do with the information gathered from a flow experiment?
We can use flow cytometry to count the number of cells of interest, such as T cell subpopulations. Cell counting gives us information about how cells are reacting to certain treatment conditions. For example, in a sample treated with an NK cell agonist (i.e. something that promotes NK cell activation), the expected result would be that NK cells would be elevated in the treatment group. This could tell us whether the agonist is effective. Another application is the identification of cell lines. We can identify a cell line by treating a sample with antibodies corresponding to the expected ligand profile (i.e. what we expect to find on the surface of that cell), hence confirming that it is correct, or if it has been mislabelled.
Some cytometers can be used for fluorescence-activated cell sorting (FACS), wherein the cell population of interest is sorted into a new tube by the cytometer. The sorted cells can then be used for downstream applications. For example, in a cell line, one may have cells of the same identity expressing different levels of an antigen. We can use FACS to sort only the “high” antigen-expressing cells and expand these in culture.
Such applications can extend into disease diagnosis, such as for leukaemia. Flow cytometry can also be used in biopsies of the bone marrow or lymphatic organs to see if they are functioning properly.
The main advantage of traditional flow cytometry is that it is quick to perform, with most machines counting thousands of cells per second, and produces a lot of data from a small sample size. It is very useful in identifying subpopulations as well.
As with any method, there are also a few disadvantages. Flow cytometers can be expensive to run and require specialised training to use. Preparing samples can be a lengthy process, and data analysis for big experiments can also be time-consuming.
The future of flow cytometry
Flow cytometry has been around for decades, with the first fluorescence-based cytometers coming up in the 1960s. The first cytometers could only detect around one and three parameters, but today’s cytometers can detect up to over 40 parameters in a single flow experiment!
The most recent breakthrough in the field has been the development of spectral flow cytometry. In traditional flow cytometry, which has been described in this article, photomultiplier tubes detect fluorescence of specific wavelength ranges. The data produced from this experimental set-up requires “compensation”, wherein the overlap between the emission spectra of different fluorophores has to be removed. To overcome this data analysis caveat, spectral flow cytometers were developed.
Figure 4: Example of overlapping emission spectra of four different fluorophores: FITC, PE, PerCP-Cy5.5, and PE-Cy7. (a) Traditional flow cytometry spectral view [link]. Grey rectangles represent bandpass filters. (b) Spectral flow cytometry spectral view [link]. Detectors do not record a filtered spectrum, but the entire spectrum.
Figure 4 shows the spectral overlap between four fluorophores commonly used in flow cytometry. In Figure 4a, the FITC, PE, and PE-Cy7 spectra are overlapping in the 550-600nm region, and the PerCP and PE-Cy7 spectra overlap in the 750-800nm region. The signal obtained has to be “compensated” so that these overlapping sections are eliminated and only the desired peak is recorded.
For example, in the 585/42 filter, we only want the PE signal, so the other signals have to be erased. This has to be done manually in traditional flow cytometry using a compensation matrix.
In spectral flow, the entire emission spectra for the sample is captured with a large range of detectors from the UV to the far-red wavelengths. This can be seen in Figure 4b, where although the spectra of the different fluorophores overlap, they can be easily “unmixed” because the cytometry captures the full spectrum. This can be called the “spectral fingerprint” of the sample.
Hence, after the data acquisition is complete, you just have to tell the software what your negative and positive controls are, and it will compensate the data automatically! This is a huge advantage especially for complex experiments involving 30 to 40 colours, where the datasets are large.
Another key difference between traditional and spectral flow cytometry is the architecture of the optical system. In spectral flow, the detectors are arranged in an array as shown in Figure 5. The beam of scattered light is directed by a lens into the detector array.
Figure 5: Detector array of a spectral flow cytometer. Each detector has a bandpass or longpass filter which allows only certain wavelengths to pass through and be detected, while the other wavelengths are reflected towards another detector. Bandpass filters allow a range of wavelengths to pass (e.g. 425 to 450 nm), while longpass filters allow wavelengths longer than the filter wavelength to pass through.
The detector array can have up to eight individual detectors, and a spectral flow cytometer can have multiple detector arrays. Hence, multiple colours can be analysed from the same sample, with some cytometers capable of analysing up to 40 colours in one experiment!
A visual comparison of spectral flow and traditional flow is shown in the figure below.
Figure 6: Comparison of conventional flow cytometry (above) and spectral flow cytometry (below). This figure is courtesy of ThermoFisher [link].
The bottom line
Overall, flow cytometry is a powerful and versatile method for multi-parameter single-cell analysis, and the more efficient and accurate the method can be, the better the results obtained.
So why should we care about advancing flow cytometry anyway?
The primary reason is to accelerate research capabilities. Flow cytometry is widely used in many fields, and as it advances, so does the speed, accuracy, and quality of data produced, and hence the ability of a research lab or industry to accelerate and advance scientific discoveries!
MRes Molecular & Cellular Biosciences
Imperial College London