Imaging the Neutrophil Phagosome and Cytoplasm Using a Ratiometric pH Indicator

Neutrophils are crucial to host innate defense and, consequently, constitute an important area of medical research. The phagosome, the intracellular compartment where the killing and digestion of engulfed particles take place, is the main arena for neutrophil pathogen killing that requires tight regulation. Phagosomal pH is one aspect that is carefully controlled, in turn regulating antimicrobial protease activity. Many fluorescent pH-sensitive dyes have been used to visualize the phagosomal environment. S-1 has several advantages over other pH-sensitive dyes, including its dual emission spectra, its resistance to photo-bleaching, and its high pKa. Using this method, we have demonstrated that the neutrophil phagosome is unusually alkaline in comparison to other phagocytes. By using different biochemical conjugations of the dye, the phagosome can be delineated from the cytoplasm so that changes in the size and shape of the phagosome can be assessed. This allows for further monitoring of ionic movement.


Introduction
The neutrophil is the most abundant innate immune cell in the body. Its main function consists of patrolling the bloodstream and engulfing and digesting the foreign particles that it may encounter in a process known as phagocytosis 1,2 . The particles are degraded in an intracellular compartment called the phagosome. The activation of neutrophil NADPH oxidase, the isoform NOX2, initiates a cascade of biochemical reactions that culminates in the death of the pathogen. NOX2 protein subunits proceed to form an electron transport chain complex in the phagosomal membrane 3 . Once activated, it transports electrons from NADPH across the membrane to molecular oxygen inside the phagosome, producing superoxide anions and further reactive oxygen species. This is known as the respiratory burst, and it is thought to be essential for efficient microbial killing and digestion 2 . However, this exclusive movement of negative charge across the membrane would soon inactivate NOX2 if it were not compensated for by positive charge moving in and/or negative charge moving out of the phagosome. It has been well established that the majority of charge compensation in the neutrophil is carried out by the proton channel HVCN1 4,5 . This channel allows for the passive movement of protons down their electrochemical gradient from the cytosol into the phagosome. Proton concentration is reflected by pH, so for a given level of oxidase activity, measuring the pH in the phagosome can provide information on the relative participation of protonic and non-protonic pathways in charge compensation.
The human neutrophil phagosome has an alkaline pH of approximately 8.5 for 20-30 min after phagocytosis 5 . This implies the existence of additional non-proton ion channels in NOX2-induced charge compensation, as the fusion and release of the contents of the acidic granules and sole compensation by HVCN1 would maintain an acidic environment, in contrast to that observed. The movement of ions to compensate this negative charge may also exert changes in phagosome size via osmosis. These may be ions present in the neutrophil at high physiological concentrations: potassium ions have been shown to move into the phagosome 6 , and chloride ionic movement is another candidate important for neutrophil function 7 .
The regulation of pH in the phagosome is vital for antimicrobial protease activity 5 . Myeloperoxidase (MPO) appears to have optimal activity at pH 6, while for cathepsin G and elastase, the optimal levels are pH 7-9 and pH 8-10, respectively 5 . Therefore, transient change in phagosomal pH may provide activity niches for different enzymes to function. Understanding how pH is involved in neutrophil microbial killing may provide useful information for the design of novel neutrophil-augmenting microbial agents.
The neutrophil phagosome is a highly reactive environment. This makes it difficult to accurately assess pH, because dyes may be easily oxidized, leading to technical artefacts. Historically, fluorescein isothiocyanate (FITC) has been the dye of choice to measure intracellular pH 8,9 . However, there are some disadvantages for its use in measuring neutrophil phagosomal pH. It has a pKa of 6.4 10 , meaning that it can only accurately be used to assess pH levels from 5 to 7.5 8 , as it saturates at pH < 8 11 . As the neutrophil phagosomal pH can become much more alkaline . Compared to other intracellular dyes, S-1 has a relatively high pKa of 7.5 10 . In acidic conditions, the molecule is protonated and produces an emission signal between 560 and 600 nm when excited at 488 nm or above. When the molecule is deprotonated in more alkaline conditions, the emission wavelength is over 600 nm. A ratio of the fluorescence intensities at these two wavelengths indicates the emission shift, which is more is reliable than single fluorescence measurements, as it is unaffected by fluorophore concentration and cell structure. S-1 can be conjugated to antigenic material, such as zymosan 14 , although heat-killed (HK) Candida albicans is preferred, as the larger surface area gives a more consistent fluorescence reading.
We have also used a modification of this method to study temporal changes in pH (Figure 3) 5 . This method for measurement of cytosolic pH can be easily applied to other cell types, as described elsewhere 15,16 , and cells with more alkaline phagosomes 14 .

Protocol
Ethics statement: All animal work was conducted with the license and approval of the United Kingdom Home Office. Human participation in this research was approved by the Joint UCL/UCLH Committees on the Ethics of Human Research. All participants provided informed consent. Figure 1 presents snapshots of neutrophils from different origins to demonstrate varying phagosomal environments. To ease quantitative analysis, it is important to seed the wells with an appropriate number of cells: too many will cause the cells to layer over each other, making it difficult to view enclosed phagosomes accurately; too few will, of course, provide fewer results, particularly as not every neutrophil will phagocytose. Figure 2 is an image that is over-saturated; this can be assessed by splitting the image between its two channels (using the microscope software recommended in the Materials List or an equivalent) -red dots show where maximum fluorescence has been detected. This can be countered by reducing the intensity of the laser. Calibration curves using the various buffer systems are shown in Figure 3, adapted from Levine et al. 5 . The error bars show that there is some variation in fluorescence between readings. Figure 4 gives an example of how the data for phagosomal pH and area could be presented. This approach allows each individual measurement to be displayed with an over-laying boxplot. However, the data could also be displayed in a histogram bar chart. human peripheral blood neutrophils at the same point after phagocytosis. They appear slightly more alkaline than the mouse wildtype cells, but the phagosomes are still not as large and red as the Hcvn1 -/cells. (D) shows human neutrophils that have phagocytosed Candida in the presence of 5 µM diphenylene iodonium (DPI). All the phagosomes are very acidic, with a pH of 6 or less; the drug inhibits the NADPH oxidase, so there is no compensatory ion movement. The protons released from the acidic granules that fuse to the phagosome cause the acidic pH wildtype mouse bone marrow neutrophils, C: human peripheral blood neutrophils, and D: human neutrophils with 5 µM DPI n = 3/300. Individual measurements are shown as small squares, with an overlaying boxplot with median and interquartile range. A red bar represents the mean. As seen in the images in Figure 1, Hvcn1 -/cells have very alkaline phagosomes in comparison to wildtype mouse and human neutrophils. They also have a bigger phagosomal area ( Figure 4B, n = 3/300). Human neutrophil phagosomes are slightly more alkaline and larger than wildtype mouse neutrophils, while human neutrophils incubated with DPI have very acidic and small phagosomes. Please click here to view a larger version of this figure.

Discussion
Once the appropriate reagents, microscope settings, and calibration experiments are set up, this method is relatively simple to perform. The critical steps include: labeling the Candida with S-1 to ensure that there is no overloading of the dye, calibrating, and analyzing the image.
S-1 is a reagent suited to more alkaline pH environments, which is particularly important for neutrophils 21 but limits its use in certain cell types.
For more acidic environments, such as macrophage phagosomes, SNARF-4, or S-4, is more suitable because of its lower pKa 22 . Moreover, for more accurate cytoplasmic readings, it is better to use S-4, as the standard curve for S-1 shows that fluorescence ratios begin to plateau below pH 6 (Figure 3). Other dyes, such as 2',7'-Bis-(2-Carboxyethyl)-5-(and-6)-Carboxyfluorescein (BCECF) or pHrodo Red may also be more suitable in a context that is expected to be acidic. Yet the cytoplasm staining is still necessary for correct identification of the phagosomes containing Candida.
An important feature of a phagosomal pH indicator is that it is not irreversibly altered by the reactive phagosomal environment. S-1 seems to be resistant to the neutrophil milieu. This is shown by Levine et al. 5 (see Supplementary Video 4 of reference 5 ) which demonstrate the phagocytosis and subsequent release of an S-1-labeled Candida particle by an Hvcn1 -/neutrophil. When phagocytosed, the particle turns from yellow/orange to red (neutral to alkaline pH), but when the particle is released by the neutrophil, it returns to its original color.
It is important to mention some of the limitations associated with using S-1. The fact that this dye has two emission spectra allowing ratiometric measurement is an advantage, but specialist equipment is needed to acquire images; the microscope used for the experiments must be able to record two images simultaneously or with an insignificant time delay. The authors assume that the researcher attempting this protocol has experience using confocal microscopy, or has access to a trained professional. We cannot list all the specific microscope parameters as they will differ for each microscope and need to be optimized by the researcher. The acetoxymethyl ester conjugated to S-1 that allows the dye to diffuse into the cell cytoplasm is degraded by non-specific esterases in the cell cytoplasm to form the fluorescent molecule. Esterases, such as alkaline phosphatase, are present in human serum and fetal bovine serum, which are used to supplement cell culture media. Accordingly, the medium in which the cells are loaded with S-1-AM (section 4.5) must not contain serum. This may prove challenging if using cells that require a more nutrient-rich medium to sustain them than the balanced salt solution used throughout this protocol. Similarly, other fluorescent medium components, such as phenol red, may interfere with S-1 measurements.
The error bars in Figure 3 indicate that there is some variation in the ratio measurements at each pH. A suitable number of repeats of each experiment (at least n = 3) and as many individual measurements in each single experiment are needed to overcome the inter-vacuolar variation. It is thus advisable to measure the pH of at least 100 phagosomes for each condition and as many phagosomal areas that appear to contain only one Candida. The phagosomes to be measured for quantitation should be those that have completely engulfed a Candida particle (i.e., those completely surrounded by cytoplasm). To mitigate against unintentional biases in the selection of cells/phagosomes for quantitation, all analyses should be performed while blind to the experimental conditions.
Here, we describe the isolation of neutrophils by dextran sedimentation of whole blood followed by centrifugation of the plasma layer through a density gradient. We use this technique as it quickly and efficiently produces a pure (>95%) population of neutrophils, although there are other methods available, such as whole blood centrifugation through other density gradient formulas or negative selection of neutrophils using specialist kits with antibodies or magnetic beads. However, the latter can be prohibitively expensive for most groups who isolate neutrophils routinely. In addition, we use the anticoagulant heparin in the blood-collecting tube, whereas other researchers may be more accustomed to using ethylenediaminetetraacetic acid (EDTA) or acid citrate sodium (ACD). As there are many different methods to choose from, it is up to the personal preference of the researcher.
Furthermore, when isolating and manipulating neutrophils they should be handled with some care to avoid excessive activation. Precautionary steps include: only using plastic ware, no glass; filter-sterilize all buffers to remove any contaminating endotoxin; when spinning neutrophils make sure the centrifuge is well balanced to avoid excessive vibrations; limit as much as possible the time neutrophils remain in a pellet after centrifugation; do not maintain neutrophils in solution of more than 5 x 10 6 /mL; and perform the experiment as soon as possible after isolation.
This method can be adapted to measure changes in the pH and phagosomal area over time by using a heated stage set to 37 °C on the microscope and taking snapshots once every 30-60 s from the same position, as described for the calibration steps. It could also theoretically be adapted for higher-throughput experiments, such as in 96-well plates, and for flow cytometry experiments, where S-1 can be used as a pH indicator 23 . However, in these settings, the emphasis on individual cell activity is replaced by a more global effect on the cell population.
This method aims to provide a relatively simple experimental setup upon which individual researchers can adapt to suit their area of interest.
Researchers may want to explore neutrophil phagosomal pH and area whilst also measuring movement of other ions, for example, intracellular calcium concentration. There are several fluorescent Ca 2+ indicators readily available for confocal microscopy, such as Indo-1, which also has dual emission spectra at 400 and 475 nm 24 . These emission wavelengths do not overlap with S-1 emission spectra, but the excitation wavelength is at the ultraviolet (UV) end of the spectrum, which can be damaging to cells, and a UV laser is not commonplace on all microscopes. A comprehensive review of the different indicators to measure intracellular calcium flux is covered by Takahashi et al. 25 and Hillson et al. 26 .
In conclusion, there are other methods available to measure phagosomal pH using different fluorescent dyes as Nunes et al. have demonstrated 13 as well as other groups 27,28 . Other researchers have also used S-1 to measure cytosolic pH 29 or phagosomal pH