Method Development and Protocol Optimization
Tissue APC subset composition is complex and requires the assessment of multiple factors to characterize cell populations.
Multicolour FC is a powerful technology that allows simultaneous the detection of multiple parameters on large numbers of cells within a single-cell suspension.
There are many potentially confounding factors that must be understood and taken into account when using FC, we developed a panel to detect the major APC subsets in human liver.
FC can also be used to determine certain functional aspects of cells. We developed and validated a protocol to assess association/phagocytosis of bacteria using an eGFP expressing coli and FC.
As explained in the introduction and background to this thesis (Chapter 1), APCs are an integral component of tissue based immune responses to infecting organisms and cancer. The role of APCs is especially relevant in the liver as it receives the majority of the blood draining from the intestine via the PT (2).
Flow cytometry allows for the simultaneous measurement of multiple variables and as a result the accurate identification of multiple subsets of cells in complex tissue.
No multicolor flow cytometry work in normal human liver has been published identifying the major subsets of intrahepatic APCs in the same panel (19). Herein we devised a panel to identify the major subsets of liver APC using multicolor flow cytometry so as to allow for a more comprehensive immunophenotypic description in human liver. We have focused on HLA-DR+CD45+ APCs following numerous exclusion gates and inclusion gates, and then identified APC subsets based on markers specific to each (and therefore known as subset specific markers).
We then devised a method using multicolor flow cytometry to assess association-with/phagocytic uptake of liver enteric bacteria by assessing eGFP following incubation of cells with eGFP expressing E.coli.
This chapter will outline considerations for the development of a multicolor flow cytometry panel and present a full panel that was used in identifying the APC subsets of the liver, finally the process involved in developing a bacterial association/phagocytosis assay used in Chapter 4 will be discussed.
Flow cytometry explained
Flow cytometry is a technique of analysis that uses fluidics and optics to assess highly complex biological samples that are in a liquid suspension(144). The fluidics system precisely funnels fluid to the point that single cells can be assessed, and the optics system allows for multiple parameters on each of these single cells to be assessed simultaneously(145). Using multiple parameters it is possible to characterize populations of cells for analysis and group them into subsets based on their relative intensity of the proteins they express (145)(see Figure 3-1). Functional assays can then be performed and changes to these subsets can be assessed(146). FACS can also be used to isolate purified populations based on their surface marker expression for further functional analysis (128, 145) or assessment of gene expression if necessary(147). These features make flow cytometry an incredibly powerful tool and further uses for it are regularly being developed(145). It is for these reasons that we decided to use flow cytometry to characterize APC subsets in the human liver.
The fundamental principle of flow cytometry relies on detecting the intensity of emission of light from particles (most commonly fluorophores attached to antibodies specific for a protein, otherwise known as an antigen, of interest) excited by a laser(148). Detectors convert the energy from light into an electrical current (a voltage) that is amplified so as to produce measurable data (using photomultiplier tubes, or PMTs)(148). This measurement of photons converted into an electrical signal allows the intensity of light emitted by a cell following exposure to a laser to be assessed using software designed to do so (see Figure 3-1)(147).
Figure 3-1 Example of data obtained from a flow cytometry experiment.
A and B present the number of cells present in a solution according to the intensity of fluorescence for CD14 and HLA-DR (DR) respectively. C shows the same data but in a two dimensional dot plot format showing the relative intensity of CD14 and DR on every cell detected, the red box indicates corresponding points in the dot plot for CD14 and the green box indicates corresponding plots for DR. As this example shows, a dot plot gives more detail regarding populations of cells and their co-expression of markers otherwise not evident – for example in this case, the fact that CD14+ APC have diverse range of DR expression.
By using light filters, the wavelengths of light that pass through to detectors can be narrowed to specific ranges. Each detector is arranged behind a dichroic mirror that only lets light pass through it above a certain wavelength, and reflects light below that wavelength. The bandwidth of the light that passes through the mirror is then narrowed further by a band-pass filter before reaching the detector. Band-pass filters only allow light within a specific wavelength to pass through them. The reflected light that does not pass through the dichroic mirror is reflected to another mirror positioned in-front of a band-pass filter and detector with a similar arrangement but that again only allows light pass through that is above a certain wavelength (this time a shorter wavelength than the first mirror). Light either passes through a dichroic mirror if it has a wavelength longer than that deflected by a given mirror or is reflected down successive dichroic mirror-bandpass filter-detector combinations until it reaches the detector of longest wavelength for the given machine (see Figure 3-2). This system controls only wavelengths within a known specific range to reach each detector – as determined by the dichroic mirror and bandpass filter.
If one is cognizant of the emission spectra of certain fluorophores, then by labelling cells with fluorophore-conjugated antibodies one can detect light emitted within a certain wavelength and can – by inference – detect the presence of proteins expressed by a given population of cells, there is no theoretical limit to the number of fluorophores that can be assessed(145).
Figure 3-2 Simplified diagram of the detector and PMT layout of a one laser system with 8 detectors
In an example modified from (145), the schematic for a sequence of detectors for light emitted from cells excited by a red laser is shown in A and how this system looks in the machine is shown in B. Light emitted from a cell is deflected towards a series of dichroic mirrors and band pass filters with associated voltage detectors and
PMTs. In this example, an 8 detector circular system is shown, light enters where the red arrow is pointing and is reflected by a dichroic mirror to pass onto the next dichroic mirror-bandwidth filter-detector set or travels through to a detector.
The size and granularity of cells is also determined by what is known as forward scatter (FSC) and side scatter (SSC) that are measures of light scatter produced by a cell irrespective of fluorescence.
However, the sensitivity of the technology and the light spectrum based mechanism that crudely detects light of a specific wavelength and then assigns that to a specific fluorophore can produce confounding data that can both over and under-represents findings. Beyond including a viability stain and excluding doublets (as described in 2.5.4) a mindfulness of other factors, such as cellular autofluoresence and a process known as compensation (described below in section 188.8.131.52) is required for adequate machine set-up and analysis of samples(149).
Flow cytometry panel development
As explained above, flow cytometry works on the principle of detecting a defined spectrum of light being emitted by fluorophores that have been excited by a defined wavelength of light (which is most often produced by a laser). Unlike lasers (that emit light at a single wavelength) fluorophores emit a spectrum of light when excited (known as the spectrum of emission), that can be broad be highly variable(145). As a result, detectors are in actual fact only detecting the percentage of light emitted from that fluorophore that has not been reflected by the dichroic mirror and has then passed through the band-pass light filter.
If one is assessing only one fluorophore or using a panel of fluorophores with sufficiently disparate emission spectra then the light that actually reaches specific detectors (once it has passed through its dichroic mirror and bandwidth filter) is likely to only have been emitted from one type of fluorophore(149) that the mirror-filter-detector has been designed to detect without any (or minimal) light being detected that has been emitted from another fluorophore used in the same panel.
However, due to the complexity of APC subsets in tissue, a minimum of 4 markers is required to analyze these cells in the liver (this includes DAPI as a viability marker, CD45, a dump gate, and HLA-DR)(150).
When numerous markers are being assessed (such as four or more, but even as few as two(149)) an issue arises in which multiple fluorophores may have an overlapping emission spectrum(149). For example the fluorophore fluorescein isothiocyanate (FITC) has a “tail” of emission spectra that is still detected by the detector with a mirror-filter combination used to detect another fluorophore phycoerythrin (PE) as is shown in Figure 3-3. This phenomenon is referred to as “spill-over” and can be partially dealt with by using what is termed compensation(149).
Figure 3-3 FITC spill-over into PE when excited by a 488nm laser.
This figure was acquired using the BioLegend Spectra Analyzer (BioLegend Inc, San Diego California) and displays the theoretical emission spectra of the fluorophores FITC and PE when excited by a laser emitting light of 488nm wavelength (blue box). FITC has an emission spectrum that extends from below 500nm to over 600nm and as a result emits light that will pass through the dichroic mirror and filter set used to detect PE. This particular filter set will theoretically allow approximately 25% of light omitted from FITC to be detected as a positive signal by the PE detector (and therefore attributed to PE). This phenomenon is known as “spill-over” and results in the requirement for the process known as compensation.
Compensation refers to the process involved in appropriately accommodating for spill-over from each fluorophore into another fluorophore’s detector in a given panel so as to improve the specificity of data collected in an experiment. In order to achieve this all fluorophores being used in a panel are assessed at their brightest level independently (this includes an unstained control to which they are compared). These spectra are then compared to each other in a matrix to determine the relative spill-over of each fluorophore into every other fluorophore being detected. Through the use of linear algebra, the compensation values are calculated(145, 151) and used by the flow cytometer (or analysis software(146)) to correct the contributions of fluorophores overlapping into a detectors that they are not designated to detect. Essentially, although not faultless(145), the process of compensation is used to improve the certainty that light detected is from the fluorophore that the given dichroic mirror- bandwidth filter-detector is designated to detect, not another fluorophore with an overlapping spectrum of emission. Compensation is essential in multi-color panels and the issues created by spill-over result in fluorophores that have a large percentage of light being emitted within a narrow range (also known as a narrow emission spectrum) being preferable(152) as it makes them easier to compensate (because they spill-over less into other channels).
Further to that, fluorophores are excited by characteristic spectra of light with stronger excitation at certain wavelengths (therefore a stronger excitation by a given laser), if lasers are set to illuminate with a known time delay then the detection of light can also be attributed to a fluorophore excited by a specific laser(151, 152). As a result, fluorophores with very similar emission spectra but disparate excitation spectra can be used on the same panel (see Figure 3-4).
Figure 3-4 Excitation spectra that have maximum excitation from different lasers reduce the burden of spill over.
This figure was acquired using the BioLegend Spectra Analyzer (BioLegend Inc, San Diego California) and displays the theoretical excitation and emission spectra of Brilliant Violet 785 and PECy7. The detectors for Brilliant Violet 785 and PECy7 both are placed behind bandwidth filters which allow through light of 750-810 nm wavelength but because they are excited by different lasers the light they emit will only be detected in the sequence of detectors for the violet (405nm) laser or blue laser (488nm) respectively.
When excited by certain wavelengths of light, molecules naturally occurring in biological samples can emit sufficiently intense light so as to mimic that of a fluorophore-conjugated antibody. This is referred to as autofluorescence and is often due to excitation of naturally occurring molecules such as NADPH and flavins (153). In tissue this can be an especially difficult problem to overcome(154).
Being aware of these issues is essential in the development of an appropriate panel of fluorophores to characterize cells in a complex mixture.
In creating a panel to assess complex tissue, fluorophore selection and machine settings are essential.
Fluorophore selection is a complex process that must consider the requirement to:-
Be aware emission spectra of all fluorophores to minimize emission spill-over if possible.
Be aware of the excitation spectra for fluorophores such that if fluorophores have a relatively similar emission spectrum to avoid an overlap of excitation spectra.
Be aware of relative antigen expression such lowly expressed antigen on rare cell populations are labelled with bright fluorophores(145).
Titrate the appropriate antibody concentration such that marker binding is saturated but not with undue background staining (145)(an example is shown in Figure 3-5).
Avoid, if possible, fluorophore-conjugated antibodies with spill-over for markers likely to be co-expressed on cells (so that the effect of spill-over does not mask a genuine shift in positive expression or produce a false-positive effect)(145).
Table of Contents
Table of Contents
List of Figures
List of Tables
Publications arising from this thesis
1 The mononuclear phagocyte system in the normal human liver.
1.1 Key Points
1.3 Systematic review of literature
1.6 Research Aims
2 Materials and methods
2.2 Cells and tissue samples used
2.3 Tissue processing
2.4 Multicolor immunofluorescence microscopy
2.5 Flow Cytometry
2.6 Functional Assays
2.7 Methods specific to Chapter 7
3 Method Development and Protocol Optimization
3.1 Key Points
3.4 Development of bacterial phagocytosis/association assay
4 Phenotypic analysis and function of APC. .
4.1 Key Points
5 Liver lobular endothelial and structural aspects
5.1 Key Points
6 APC populations by immunofluorescence
6.1 Key Points
7 Immunofluorescence assessment of ACR and HCVr.
7.1 Key Points
7.4 Patient information
8 Discussion and conclusions
8.1 Liver APC composition is complex
8.2 CD14+ APC are macrophage-like and are integral in the response to bacteria
8.3 Structure of the lobule is likely to be important in healthy liver
8.4 Characterizing the vasculature is essential to understanding the liver MPS.
8.5 The lobule appears to contain different microenvironments
8.6 APC appear to be involved in both ACR and HCVr
8.7 Limitations and future directions
8.8 Future directions
GET THE COMPLETE PROJECT
Antigen presenting cells of the mononuclear phagocyte system in the human liver.