Intraphagosomal Free Ca2+Changes during Phagocytosis

1. Introduction

Phagocytosis, the process whereby an extracellular object is internalised by a cell (phagocyte), is a complex process, involving the extension of small pseudopodia which attached to the extracellular target, initially immobilising the particle within a ‘phagocytic cup’; and then, after a large cytosolic Ca²⁺ signal, a rapid extension of the pseudopodia fully encloses the target and draws it into the cell interior [1,2,3]. This results in the formation of an intracellular vesicle, the phagosome [4,5] whose membrane originates from the plasma membrane. The relationship between cytosolic Ca²⁺ signalling and phagocytosis is complex, the cytosolic Ca²⁺ signal which occurs during phagocytosis by neutrophils regulating both the rate of extension of pseudopodia around the target [1,2,3] and activation of the bactericidal oxidase system in the phagosomal membrane [6,7]. Ca²⁺ signalling within these cells involves both Ca²⁺ influx from the extracellular environment and the release from stored Ca²⁺ from within the cell [8]. Hots spots of Ca²⁺ in the cytosol near the closed phagosome have also been reported in some phagocytic cells such as mouse embryonic fibroblasts and dHL60 cells [9] and RAW 264.7 cells [10]. It is suggested that these Ca²⁺ microdomains (hot spots) arise from ‘leaky’ interactions of STIM-1 on the endoplasmic reticulum (ER), the Ca2+ storage organelle with channels on the phagosomal membrane [9]. These Stim-1 mediated Ca²⁺ hotspots are necessary for efficient phagocytosis and oxidase actvation [11,12]. However, neutrophils, which are highly efficient phagocytes, have very little or no endoplasmic reticulum [13]. This suggests that the source of the peri-phagosome Ca²⁺ rise in these cells may be the phagosome itself. However, very little is known about the intra-phagosomal Ca²⁺ concentration and how it changes during phagocytosis. Pioneering work by Dahlgren’s group more than 20years ago [14] using fura2 as a Ca²⁺ indicator, showed that intra-phagosomal Ca²⁺ rapidly decreased during phagocytosis. However, the kinetics and relationship to the process of phagocytosis were not investigated. The Ca²⁺ probe used, fura 2, was susceptible to ‘bleaching effects’ [15] making measurements within the phagosome difficult [14]. However advances in imaging and Ca²⁺ sensing fluor design has enabled these early observation to be extended and to further investigate intra-phagosomal Ca²⁺ changes. In this paper, we have used a Ca²⁺-sensor, fluo4, which has the more robust fluorescein-like component, covalently coupled to a phagocytic target. When phagocytosed by neutrophils, this indicator dynamically reported the intra-phagosomal Ca²⁺concentration and enabled a correlate of changes in phagosomal Ca²⁺ with binding, uptake and phagosomal closure. Using this approach, we report here that there was an initially high Ca²⁺ concentration within the phagosome which was reduced to the cytosolic concentration within 200s of phagosomal closure. This is consistent with the incorporated into the phagosomal membrane, of opened plasma membrane Ca²⁺ channel, so that the open phagocytic cup acts as a conduit for local Ca²⁺ elevation, but that once closed, phagosomal Ca²⁺ is ‘drained’ into the cytosol.

2. Results

2.1. Fluo4-Zymosan Targets Report Ca²⁺ Concentration

In order to monitor intra-phagosomal Ca²⁺, the Ca²⁺ sensitive fluorescent probe, fluo4, was coupled to C3bi-opsonised zymosan particles as a phagocytic target. The sensitivity of the fluo4-zymosan particles to Ca²⁺ concentration was established in vitro by measuring the fluo4 intensity at various Ca²⁺ concentrations (Figure 1). There were two micro-anatomical zones of the zymosan particle [16], the outer zone and the central core, the central core having a brighter fluorescence signal as a result of fluo4 plus a strong auto-fluorescent signal. . The fluorescence intensity of the outer zone acted as a true the Ca²⁺ indicator, with little autofluorescence. The central fluorescent signal was, however, useful for identifying the location of the target during phagocytosis. Each zone also had a different apparent kd for Ca²⁺ (Figure 1), fluo4 in the core zone having an apparent of approximately kd 300nM, similar to that reported for free fluo4 ie 345nM at pH7.2, [17]; whereas fluo4 in the outer zone had an apparently higher kd of approx. 1mM Ca²⁺(Figure 1). However, in both zones of the zymosan particle, fluo4 had maximum intensity in 1.3mM Ca²⁺ (ie the extracellular Ca²⁺ concentration) and near zero intensity at 100 nM , the cytosol free Ca²⁺ concentration, (Figure 1). The difference in apparent Kds, fortuitously, also gave the opportunity for a double estimation of intra-phagosomal Ca²⁺ to be made from the same zymosan particle within the same phagosome. It should be noted that fluo4 at both locations was almost totally saturated with Ca²⁺ above 10mM, so the effective measurement range was thus below 10mM Ca²⁺.

2.2. Phagosomal fluo4 Signal is Decreased

Having established that the fluo4-zymosan particles reported Ca²⁺ concentration, the effect of phagocytosis was examined. The fluorescence of zymosan particles within phagosomes was significantly reduced after internalisation compared with non-phagocytosed particles in the same microscopic field (Figure 2a,c). It is unlikely that this was an optical effect caused by the interference from the cytoplasm surrounding the phagocytic target because, (i) during the process of internalisation, no partial or zonal reduction in the fluorescent signal from zymosan was observed as would be expected from optical obstruction by the encroaching pseudopodia and (ii) although there is an effect of light scattering through cytoplasm which attenuates the detected fluorescent signals from intracellular fluors [18], this would be very small. The light scattering effect of cytoplasmic projections covering the phagosome depends on the number of cytosolic granules per unit volume (density), the granule refractive index and their size (19,20). The engulfing pseudopodia were thin and excluded light-scattering granules, such that cytosolic fluorescence at these locations often has increased excitation efficiency and appear brighter (18). In a worst-case scenario, taking the published values for bulk cytosolic granules (density, and their refractive index and their size), Mie scattering calculations give an estimate of the attenuation coefficient for 488nm laser light through neutrophil cytoplasm of about 700 mm^-1 . As the approximate thickness of the cytoplasm covering the zymosan particle within the phagosome is less than 0.5 mm, less than 30% of the laser light entering the pseudopodia would be scattered and thus unavailable for excitation of fluo4, ie the intensity would be a minimum of 70% of its maximum. This is insufficient to account for the reduction in intensity of internalised fluo4-zymosan, which was less than about 5% of the non-internalised particles (Figure 2c). Thus, such a decrease in intensity of phagosomal signal (>95%) cannot be accounted for by an optical effect caused by the attenuation of light scattering effects of the pseudopodal cytoplasm. On the contrary, the efficiency of excitation and emission of fluors in this pseudopodia and near the forming phagosome is increase [18]. The elimination of the possibility that an optical effect decreased the fluo4 signal was confirmed by the use of a non-Ca²⁺ sensitive fluor (fluorescein) conjugated to zymosan, unlike fluo4-zymosan, the intensity of fluorescein was not significantly reduce when internalised (Figure 2b,c) .

Other possible causes of the reduced emission of phagosomal fluo4 were also excluded (Figure 3b). Changes in the phagosomal pH caused by activation of the oxidase/proton pump system [21] has been shown to cause a small reduction (approx. 10%) in fluorescence of non-Ca²⁺ sensing but pH-sensitive, probe, FITC-zymosan at the time when fluoe4-zymosan fluorescence was almost zero. At longer times, when the intra-phagosomal pH falls significantly, a similar reduction of fluorescence of intra-phagosomal fluo4-zymosan was also observed. This however was only evident at longer times after internalisation (5-10 mins) when the intra-phagosomal pH decreases significantly, stabilising at about pH5 after 15-20 mins [20]. Interestingly, the intensity of the fluo4-zymosan core was often seen to increase at low pH (Figure 3b). Thus these effects of pH cannot account for the decrease in fluo4 fluorescence observed within 0-30s after internalisation. It was thus concluded that pH changes could not explain the decrease of intra-phagosomal fluo4 signal.

2.3. Phagosomal fluo4 Signal Decrease not due to Oxidants

In addition to pH changes, the intra-phagosome environment is exposure to oxidants, notably short-lived superoxide which forms semi-stable hydrogen peroxide. These are generated by activation of the phagosomal oxidase at the phagosomal membrane [22]. The oxidative effect of H₂O₂ produced is enhanced by myeloperoxidase, a lysosomal enzyme in azurophilic granules of the neutrophil which are released into the phagosome by fusion with the phagosome. We have previously used zymosan-conjugated dichlorodihydrofluorescein (DCDHF) as an indicator of intraphagosomal oxidants, [23] and have shown that the oxidation within the phagosome begins after phagosomal closure and continues with similar kinetics to those of the fluo4 intensity decreased reported here. However, no immediate in vitro effect of H₂O₂ and myeloperoxidase was observed and fluo4 zymosan intensity was not significantly reduced before approx. 100s of contact (Figure 3a). Even by 100s, the intensity was reduced only be approx. 15%. It was thus concluded that intra-phagosomal oxidation could not account for the decrease in fluo4-zymosan intensity observed during phagocytosis It should also be noted that fluorescein (a chemically related molecule) was also resistant to the bleaching effect of the intra-phagosomal environment as it remained at the same brightness whether outside or inside the phagosome (Figure 2b).

2.4. Phagosomal fluo4 Zymosan REMAINED Responsive to Ca²⁺ Changes

In order to test whether fluo4 attached to zymosan remained responsive to Ca²⁺ after contact with the intra-phagosomal environment, three approaches were adopted. The first approach was simply to liberate the zymosan particle from within the phagosome by lysis of the cells with Triton X-100 in order to allow contact of the released particles with the extracellular high Ca²⁺ (1.3mM). This procedure significantly increased the fluorescent signal, consistent with a retained Ca²⁺ sensitivity (Figure 3c). As Triton X-100 destroyed the cell morphology, this approach could not be undertaken whilst single cell imaging. Instead, cell populations were mixed with zymosan particles, centrifuged to bring the particles and cells in close contact, then after 5 mins at 37^oC, resuspended and 3-cycles of sedimentation at 1g used to separate free zymosan particles from cell associated zymosan. The resultant cell pellet was sampled for imaging to confirm that all zymosan particles were within phagosomes and their fluorescence intensities measured. Triton X-100 was then added to a portion of the cell pellet and sampled for imaging to confirm that no zymosan particles were within phagosomes and their fluorescence intensity measured. There was a highly significant increase in the fluorescence intensity of the zymosan after Triton X-100 treatment (Figure 3c). This data was consistent with retention of Ca²⁺-sensitivity by the phagosomal fluo4-zymosan particles, so that when released into the high Ca²⁺ extracellular environment their fluorescence increased. The second approach was to use saponin (digitonin), a plant detergent-like molecule which selectively permeabilizes the cell membranes [24] and collapses the transmembrane Ca²⁺ gradient, so that intracellular fluorescent Ca²⁺ indicators become saturated with Ca²⁺ [25] but leaves the cell morphology recognisable [24]. Saponin-permeabilisation significantly increased the fluorescent signal from fluo4-zymosan within phagosomes (Figure 3c), confirming that after phagocytosis, fluo4-zymosan retained its Ca²⁺ sensitivity. However, in both the previous approaches, the intra-phagosomal fluo4 zymosan particles was exposed to high concentrations of Ca²⁺. A third approach was therefore also adopted to exclude the possibility that fluo4 remained only weakly sensitive to Ca²⁺ and responded to only high Ca²⁺. The third approach was to artificially elevate cytosolic Ca²⁺, using a Ca²⁺ ionophore, ionomycin. Elevation of cytosolic Ca²⁺ by ionomycin caused a significant rise fluo-4 fluorescence from particles within phagosomes (Figure 3c,d). This was consistent with a rise in intra-phagosomal Ca²⁺ being detected by the functional Ca²⁺ probe whilst within the phagosome. The kinetics of the intra-phagosomal Ca²⁺ rise also suggested that the Ca²⁺ channels on the phagosomal membrane were open. However it cannot be excluded that ionomycin partitioned across all intracellular membranes and that Ca²⁺ entered the phagosme via this route. Whatever the mechanism for the rise in intra-phagosomal Ca^2+, the effect showed that the intra-phagosomal fluo4 remained sensitive to physiological changes in Ca²⁺ concentration. It was thus concluded that the decrease in fluo4 intensity during phagocytosis was the result of a decrease in intra-phagosomal Ca²⁺conentraion from an initially saturating concentration of Ca²⁺ (probably the extracellular concentration of 1.3mM), to the cytosolic Ca²⁺ concentration (approx. 100nM),

2.5. Kinetics of INTRA-phagosomal Ca²⁺ Decrease

By monitoring the intensity of fluo4-zymosan particles during internalisation by the phagocyte, the relationship of the decrease in intra-phagosomal Ca²⁺ with the stages of phagocytosis were established. On engaging the zymosan particle, the cells form a “phagocytic cup” at the base of the zymosan, the “cup” then extends until the particle is enclosed within the phagosome [1,26]. Although the time of contact of the particle with the cell could be determined, the exact time of complete phagosome closure was more difficult to establish but could be estimated within a few frames (about 0.2-0.5 s) from phase contrast imaging. At the time of phagosome closure, the intensity of fluo4-zymosan was unchanged and equal to free zymosan particles in high Ca²⁺ (1.3mM). However, within 5-10 seconds of closure, the intensity of the internalised fluo4-zymosan particles abruptly decreased, reaching its low equilibrium level within 20 seconds (Figure 4 and movie 1). Presumably, the high (millimolar) phagosomal Ca²⁺ at the time of phagosome closure leaked out of the phagosome into the cytosol over this time. From knowledge of the phagosome water volume, it can be estimated that the phagosome initially contained about 2x10⁷ Ca²⁺ions, and thus that it would be lost at an average rate of about 10⁶ ions/sec over 20 seconds to reduce this to the cytosolic concentration to c.100nM. After the time when phagosomal Ca²⁺ had been reduced to below the fluo4 saturation level (10mM), there was an approximately linear decline in fluo4 intensity equivalent to a decrease in intra-phagosomal Ca²⁺ from 10mM to 0.1mM over 5s, suggesting a slower loss of Ca²⁺ (as expected if the outward leak were driven by the Ca²⁺ concentration gradient across the phagosomal membrane) over this time of about 0.2x10⁴ Ca²⁺ ions/s. The plasma membrane Ca²⁺channels found on the neutrophil phagosomes is TRPM2 (27) carry currents of about 1pA (28) or about 10⁷ ions/sec suggesting that only a few channels need to be open on the phagosome membrane (eg 5 channels each open for 0.2s would drain the phagosomal Ca²⁺ at that rate). It is thus possible that that plasma membrane Ca²⁺ channels opened by the phagocytic stimulus, which gives rise to the well-document phagocytic cytosolic Ca²⁺ signal [1,2,23,26], remain open in the phagosomal membrane and “drain” the enclosed phagosome of its Ca²⁺.

2.6. Location of Open Ca²⁺ Channels

In order to test the conclusion that Ca²⁺ channels remain open on the phagosomal membrane, the translocation of the YFP-tagged C2 domain of PKC-γ (YFP–C2-γ) from the cytosol to the plasma membrane was used as a real time spatial marker of Ca²⁺ influx [29,30] into the phagocytic cell line RAW 264.7 macrophages. When presented with mouse C3bi-opsonised-zymosan (without fluorescence tags), the translocation of YFP–C2-γ to the plasma membrane was observed after phagocytic cup formation, reflecting the large cytosolic Ca²⁺ signal expected. Initally, there was localised translation to the phagocytic cup, indicating that Ca²⁺ channels were open in this region (Figure 5a-1 sec). Within the next second (labelled ‘2 sec’ on Figure 5a) translocation of YFP–C2-γ) was observed at all plasma membrane location (Figure 5a-2 sec). This rapid opening of Ca²⁺ channels remote from the phagocytic site explains the difficulty in observing localised cytosolic Ca²⁺ rises during phagocytosis as the global open Ca²⁺ channels overwhelmed the effect of a localised influx of Ca²⁺ (23). Clearly, Ca²⁺ channels remain open on the forming phagosome after phagosome closure. In the example shown, the cell also contained a previously formed phagosome (Figure 5b marked ‘x’ ) which initially had no YFP–C2-γ on its surface indicating the absence of a Ca²⁺ flux from the phagosome. However, when the cytosolic Ca²⁺ signal was triggered by the second phagocytotic event, it was seen that YFP–C2-γ also translocated to the membrane of the previous phagosome (especially obvious at 20 sec, in Figure 5c). This can be explained if either (i) the Ca²⁺ channels on the previously formed phagosome had remained open, so that the elevated cytosolic Ca²⁺ began to refill the previously drained phagosome which begun leaking out (detected by YFP–C2-γ translocation) or (ii) the channels were previously closed by were re-opened as result of diffusible Ca²⁺ opening stimuli (eg IP₃). Thus the use of YFP–C2-γ has demonstrated that Ca²⁺channels on the forming phagosome were open during phagosome formation and were open or operational after phagosome closure.

3. Discussion

The results presented in this paper show that the open phagosome entraps extracellular Ca²⁺ which ‘empties’ into the cytosol once the phagosome has closed. The concentration of phagosomal Ca²⁺ within the phagocyte equilibrates with the cytosolic Ca²⁺ within 20 seconds of phagosome closure. This sequence is shown as a cartoon (Figure 6). Thus the phagosome cannot act as a long term store of Ca²⁺ for later signalling events. Also, it cannot act as a reservoir for loading lysosomes with Ca²⁺ for other signalling events [31]. The rapid decline of intra-phagosomal Ca²⁺ was reported over 20 years ago by Dahlgren’s group, who monitored Ca²⁺ in a similar way to that described here but using fura2 as the Ca²⁺ reporter. Although fura2 was bleached during the experiments (both 340 and 380nm decreasing), the group found that the intra-phagosomal Ca²⁺ level dropped rapidly. Little notice has been given to this early observation which is confirmed (and extended ) here. However, Anke et al [27] have more recently shown that in phagosomes from phagocytes taken from Trpm2^−/− mice ie lacking the TRPM2 Ca²⁺ channel, intra-phagosomal Ca²⁺ rose slowly (over a 10 min time period) compared to the intra-phagosomal Ca²⁺ expressing TRPM2. They suggested that the phagosomes devoid of TRPM2, lacked a leakage channel and so Ca² accumulated within the phagosome. This conclusion is similar to that which was drawn from the data shown here.

From the work presented here it seems unlikely that Ca²⁺ events could be driven by phagosomal Ca²⁺ close to the time of completion of phagocytosis. However, if the Ca²⁺ channels closed slowly (over minutes rather than seconds), and the phagosomal membrane contained functional plasma membrane Ca²⁺ pumps, it is expected that the phagosome would re-load with Ca²⁺ over a longer time periods. The equilibrium generated across the plasma membrane would thus also be generated across the phagosomal membrane. Under these circumstances, phagosomal Ca²⁺could be important for driving peri-phagosomal Ca²⁺ events. The earliest report of such a Ca²⁺ cloud localised to near the phagosome [32] is usually dismissed as almost certainly an optical artefact caused by the thinner cytoplasm around the pseudopodia which increased the excitation and efficiency in that area [18]. However, the work presented here opens the possibility that in phagocytes with endoplasmic reticulum, the phagosome itself may be a source of Ca²⁺ which adds to or generates peri-phagosomal Ca²⁺ events such as Ca²⁺ hotspots (9,10).

4. Materials and Methods

4.1. Cell Preparation

Human neutrophils, isolated from the blood of healthy volunteers whole had given informed consent as described previously [10] were suspended in Krebs medium (NaCl 120 mM, KCl, 4.9 mM KH2PO4, 1.2 mM MgSO4, 1.2 mM CaCl2, 1.3 mM, HEPES 25 mM and bovine serum albumin, 0.1% adjusted to pH 7.4 with NaOH).

4.2. Properties of the Phagocytic Target

Zymosan particles (killed yeast cells) of approximately ellipsoid shape, having radii in the x, y and z planes (or semi-axes) of 1:1.5 and 1mm were opsonised with iC3b by incubation with human or mouse serum as previously described [11]. The particles have a cell wall enclosing an outer space and a dense central core [12]. The water space in the zymosan-containing phagosome is approximately 50% of the total volume [12], and so would each contain 6.3mm³of extracellular medium (1.3mM Ca²⁺). A cadaverine-derivative of the fluorescent Ca²⁺ indicator fluo4 (Fluo4 cadaverine (F36201 fluo-4 cadaverine, pentapotassium salt, Molecular Probes; Thermo Scientific) was linked to amines in the zymosan particles by carbodiimde linkage using (1-ethyl-3-[3-dimethylaminopropyl]carbodiimide hydrochloride, Thermo Scientific).The reagents were used at 1mg/ml and incubated with a suspension of zymosan particles at room temperature for 4 hours before separation of the reactants from the solid particles by 3 rounds of centrifugation and resuspension in fresh medium to remove the reactants. A similar procedure was used to produce non-Ca²⁺ sensing fluorescein-linked zymosan particles by reacting the particles with fluorescein isothiocyanate (FITC, Sigma) as described previously (34).

4.3. Imaging and Intra-Phagosomal Ca²⁺ Monitoring

The fluo4-zymosan particles were sedimented onto a glass coverslip mounted onto a thermostatically control stage (37 ± 0.1 °C) in Krebs medium. The coverslip was then washed with medium to remove particles which had not made contact with the coverslip and were loosely adhered. Neutrophils were then allowed to adhere to the same glass coverslip whilst observing microscopically and selecting a microscopic field in which neutrophils and particles were close and expected to undergo phagocytosis. The field was then imaged using a resonant laser scanning head of the LeicaSP5 confocal inverted microscope (Leica Microsystems, Heidelberg) and 63x objective. Fluorescence (excited by 488nm laser scanning) and transmitted light (phase contrast images) were acquired simultaneously. Image analysis and presentation was achieved using ImageJ software (https://imagej.nih.gov/ij/)

4.4. Estimation of Intraphagosomal Ca²⁺ Concentration

The concentration of intra-phagosomal Ca²⁺ was estimated from the fluorescence intensity of the zymosan particles during the experiment (F). The intensity of zymosan particles outside the cell (Zo) in 1.3mM Ca²⁺ was taken as the Ca²⁺ saturation maximum (Fmax) and the minimum fluorescence (Fmin) was estimated as being the autofluorescence signal from unlabelled zymosan particles, as in the in vitro calibration curves (see Figure 1). Using the kd values from the calibration curves the intra-phagosomal Ca²⁺concentration was estimated using a variant of the standard equation for single wavelength fluors [25], was used to estimate:- : Ca²⁺= Kd (α−1)/(β−α), where α =F/Fmin and β =Fmax/Fmin.

4.5. Raw 264.7 Cell Transfection

RAW 264.7 cells were electroporated to introduce the C2-γ-YFP plasmid (3 µg plasmid DNA per 2×10⁶ cells) using the Cell Line Nucleofector (Lonza) as described previously (33). Cells were incubated at 37°C in 5% CO₂ for 3–4 h to enable expression of the newly introduced DNA, before imaging on a Leica SP5 confocal microscope. Fluorescent protein expression in transfected RAW 264.7 cells was detected by ∼1 h post transfection, but expression was optimal at ∼4 h post transfection. The plasmid encoding for C2-γ-YFP was a kind gift from Theodorus W. Gadella (Swammerdam Institute for Life Sciences, University of Amsterdam, The Netherlands)