Ca²⁺ and calpain control membrane expansion during the rapid cell spreading of neutrophils

Summary

Following adherence of neutrophils to the endothelium, neutrophils undergo a major morphological change that is a necessary prelude to their extravasation. We show here that this shape change is triggered by an elevation of cytosolic inositol (1,4,5)-trisphosphate (IP₃), to provoke physiological Ca²⁺ influx through a store-operated mechanism. This transition from a spherical to ‘flattened’ neutrophil morphology is rapid (∼100 seconds) and is accompanied by an apparent rapid expansion of the area of the plasma membrane. However, no new membrane is added into the plasma membrane. Pharmacological inhibition of calpain-activation, which is triggered by Ca²⁺ influx during neutrophil spreading, prevents normal cell flattening. In calpain-suppressed cells, an aberrant form of cell spreading can occur where an uncoordinated and localised expansion of the plasma membrane is evident. These data show that rapid neutrophil spreading is triggered by Ca²⁺ influx, which causes activation of calpain and release of furled plasma membrane to allow its apparent ‘expansion’.

Introduction

An essential feature of the way in which neutrophils behave in the body is their capacity to change morphology and undertake significant and rapid changes in cell shape. During chemotaxis, neutrophils develop a leading front edge and a trailing uropod, and during phagocytosis, they form localised pseudopodia and phagosomes. Not surprisingly, the mechanism and intracellular signalling events controlling chemotaxis (Kay et al., 2008) and phagocytosis (Swanson, 2008) continue to be widely studied. However, the mechanism by which neutrophils change shape rapidly has been less well studied, even though the qualitatively different and slower cell spreading behaviour in non-immune cells is increasingly understood (Zhang et al., 2008; Huang et al., 2009; Cuvelier et al., 2007). Despite this lack of attention, neutrophil spreading is the key event in determining the rate of neutrophil extravasation both physiologically and pathologically. After rolling along the endothelium, the neutrophil flattens onto the endothelial surface (‘spreading’) at sites of inflammation in response to cues from the endothelial surface. This dramatic cell shape change minimises the shear force of blood flow on the neutrophil that would otherwise tend to cause cell rolling. The flattening morphology transition is also a necessary prelude to neutrophil extravasation, either between the endothelial cells or through an individual cell, and subsequent chemotaxis in the extravascular space (Ley et al., 2007).

Pioneering work by Maxfield showed that the spreading of neutrophils and macrophages onto surfaces is preceded by large global cytosolic free Ca²⁺ signals (Kruskal et al., 1986: Kruskal and Maxfield, 1987). These cytosolic free Ca²⁺ signals were caused by adhesive signals to the cell from the substrata, and it has subsequently been shown that immobilisation of β2 integrin by surfaces (Jaconi et al., 1991; Pettit and Hallett, 1996) or cross-linking by anti-integrin antibodies (Ng-Sikorski et al., 1991: Petersen et al., 1993) triggers similar Ca²⁺ signals. There is, however, a major problem in designing experiments to investigate the causal link between the Ca²⁺ signal and the cell spreading event. The stimulus for the Ca²⁺ signal (i.e. a surface) cannot be studied in isolation from the spreading response as both the delivery of the signal and the ability to respond depend on the same event, namely contact with the same surface. In our studies, we have overcome this problem by studying neutrophils in conditions that permit contact with a surface in the absence of signalling and flattening, and subsequently imposing changes in their cytosolic chemistry using (non-contact) photolytic manipulation. Using this approach, we have found that the elevation of cytosolic free Ca²⁺ triggered by photolytic generation of cytosolic inositol (1,4,5)-trisphosphate (IP₃) is sufficient to induce a rapid cell shape change. The membrane expansion accompanying this cell shape change occurred without the insertion of new membrane, and thus probably results from unfurling of wrinkled plasma membrane (Dewitt and Hallett, 2007; Hallett and Dewitt, 2007). This is also suggested from biophysical measurement (Herant et al., 2005) and the manipulation of neutrophil membrane tension (Houk et al., 2012). The cell spreading and membrane expansion depended on activation of the Ca²⁺-activated protease, calpain. Calpain activation, triggered by Ca²⁺ influx, was therefore the key step for the rapid neutrophil flattening morphology transition.

Results

Morphology changes during neutrophil flattening

Rapid spreading onto glass can be triggered in adherent neutrophils by elevating cytosolic free Ca²⁺ using a chemotactic stimulus (Fig. 1). The cytosolic free Ca²⁺ concentration and cell morphology of the non-stimulated neutrophils is stable for many minutes before stimulation but after the peptide-induced cytosolic free Ca²⁺ signal, the cells flatten onto the surface over the subsequent 80 seconds. Although there is no obvious increase in cell–substrate contact area immediately on addition of the peptide or at the peak of the cytosolic free Ca²⁺ signal (Fig. 1A), a dramatic increase in the area in contact with the glass substrate is evident immediately afterwards (Fig. 1). Upon reconstructing the cell surface of the cells from confocal optical sections, it can be seen that both the cell adhesion footprint increases and cell thickness decreases to a point limited by the thickness of the nuclear lobes within the cell. The individual lobes of the multi-lobed nucleus within the non-spread cell are held with the longer axis upright and are restrained from sedimenting by the resistance of cell periphery (Fig. 1B,C). However, after the Ca²⁺ signal, the restraint to sedimentation of the nuclear lobes is relaxed and the lobes lie on their long axis and adopt the characteristic polymorphonuclear display (Fig. 1B,C). A striking change during flattening is in that the apparent cell surface area increased by up to 200% in these experiments. As a spherical surface is the minimum that encloses a given volume, it is not surprising that the transition to a non-spherical cell morphology is accompanied by an increase in the apparent cell surface area.

Fig. 1.

Rapid cell spreading in neutrophils is induced by a Ca²⁺ signal. (A) The upper series of images show the phase contrast images and the lower series the fluo4 intensity (as a readout of Ca²⁺ concentration) of an individual human neutrophil undergoing cell spreading in response to f-mlp (1 µM). The graph shows the time course of cytosolic free Ca²⁺ concentration with times at which the images were taken indicated and the time of addition of f-mlp shown. (B,C) The cell surface (stained with DiI) and the nucleus (stained with acridine orange) have been reconstructed from confocal optical slices of an adherent but not ‘spread’ and a spread cell and are shown in orthogonal planes (B,C) as isosurfaces (Imaris). Each of the examples shown are representative of at least 20 experiments using neutrophils taken from at least four different donors.

Uncaging IP₃ triggers neutrophil flattening

Studies using soluble stimuli (such as those shown in Fig. 1) are of limited value for studying the mechanism of the morphology change. First, only cells which are adherent to some extent can be studied (otherwise the addition of stimulus would simply wash away the cell). Second, chemotactic stimuli do not provide a simple cytosolic signal. Occupancy of chemotactic receptors generates a number of chemical changes within the cytosol and the cell membrane, each of which could contribute to the effect. Third, soluble stimuli do not reflect the physiological situation, where ICAM-1 molecules on the endothelial cells engage β2 integrin on the neutrophil to trigger the morphology change. For these reasons, studies of neutrophils spreading onto stimulating surfaces are often performed. However, there is also an important logical problem in designing experiments to investigate the link between the Ca²⁺ signal and the cell spreading. As the surface onto which the neutrophil sediments both delivers the stimulus Jaconi et al., 1991; Pettit and Hallett, 1996) and is required for the spreading response to be evident, ‘stimulus’ and ‘response’ become logically entangled. In this study, we have overcome these problems by imposing a defined chemical change within the cytosol on neutrophils that are in contact with an underlying (non-stimulatory) surface by using photolytic uncaging of Ca²⁺-elevating agents.

Uncaging cytosolic caged IP₃ will generate biologically active IP₃ within the cell on command. On exposure to the uncaging illumination, the characteristics of the Ca²⁺ signal were variable between individual neutrophils, but three phases could often be distinguished (Fig. 2B), an initial lag (0–5 seconds), followed by a small rise or shoulder (100–200 nM), before a single large and transient Ca²⁺ signal (peaking at 800 nM). As uncaging fluorescein by this method occurred without delay (Fig. 2A), the lag period between UV illumination and the onset of the Ca²⁺ signal was attributed to the time required to generate sufficient IP₃ to breach the threshold cytosolic concentration of ∼20 µM that is required to release of Ca²⁺ from intracellular storage sites (Davies-Cox et al., 2001) and trigger Ca²⁺ influx across the plasma membrane. For neutrophils on glass, this IP₃ uncaging strategy induced a rapid transformation to the flattened ‘spread’ morphology in all cells studied in which a Ca²⁺ influx signal was triggered (n>50, Fig. 2C; supplementary material Movie 1). Exposure of non-loaded neutrophils to uncaging wavelengths or exposure of caged-IP₃-loaded cells to non-uncaging wavelengths failed to increase the spreading rate. Furthermore, the effect cannot be attributed to the photolytic products other than IP₃ because uncaging other caged compounds (including nitr5, see later) failed to trigger cell spreading. IP₃-induced spreading was also triggered in neutrophils on an integrin-engaging surface (fibronectin-coated glass), showing that the flattening response was not unique to the glass surface (Fig. 3B). However, fibronectin was not routinely used as it was necessary to select cells in contact with the surface before as the surface itself had triggered the Ca²⁺ signals (Jaconi et al., 1991; Pettit and Hallett, 1996). The Ca²⁺ signal triggered by IP₃ uncaging in neutrophils on this surface was also more complex than on glass, with multiple peaks, reminiscent of the Ca²⁺ signals, which occur during integrin-mediate phagocytosis (Dewitt and Hallett, 2002; Dewitt et al., 2003). The complexity of the Ca²⁺ signal has been attributed to cycles of Ca²⁺-induced integrin untethering and mobilisation followed by further integrin-induced Ca²⁺ signalling (Dewitt and Hallett, 2002).

Fig. 2.

Uncaging cytosolic IP₃. (A) The effect of uncaging caged fluorescein is shown. There is no delay between the onset of UV illumination and the increase of fluorescent signal. (B) The effect of uncaging caged cytosolic IP₃ on cytosolic free Ca²⁺ is shown for a platelet (pl) and neutrophil (n) in the same microscopic field. The upper traces show the intensity of fluo4 and the lower trace the uncaging illumination (at 360 nm). (C) The upper series of images shows the phase contrast images and the lower series the fluo4 intensity (Ca²⁺) of an individual human neutrophil undergoing cell spreading in response to uncaging of caged IP₃. In the series of images the time at which uncaging occurred is shown by the ‘lightning bolt’. (D) Graph of the cell–substrate contact area (footprint, upper line); cytosolic free Ca²⁺ change (middle line) and uncaging illumination (UV, lower line). Each of the examples shown in this figure are representative of at least 50 experiments using neutrophils taken from at least four different donors. Examples of cell spreading induced by uncaging IP₃ are also shown in supplementary material Movie 1.

Fig. 3.

Integrin engagement and inhibition of Ca²⁺ influx affects the IP₃-induced Ca²⁺ signal. The rate of cell spreading and cytosolic free Ca²⁺ signalling induced by uncaging IP₃ in neutrophils on (A) glass, (B) fibronectin-coated glass substrates and (C) in the presence of Ni²⁺ (1 mM) which blocks Ca²⁺ influx are shown as a series of phase contrast images on which the fluo4 intensity has been superimposed (top section) and quantified in the graphs which show the cytosolic free Ca²⁺ signal (top line), the UV illumination intensity (middle line), and the cell footprint (bottom line) as described in Fig. 1.

The crucial role of Ca²⁺ influx across the plasma membrane

In all cases in which IP₃ was uncaged (n>50), the onset of cell spreading occurred after the Ca²⁺ influx signal and never before (i.e. during the IP₃-induced Ca²⁺ release event). However, given that there was a delay in the spreading response, this data did not distinguish whether the IP₃-induced store Ca²⁺ release (i.e. the shoulder level of Ca²⁺) or the larger Ca²⁺ influx signal was responsible for the morphology change. In order to investigate this, cytosolic free Ca²⁺ was elevated to the same level achieved by IP₃-induced store release (∼100 nM above resting cytosolic free Ca²⁺) by uncaging ‘caged Ca²⁺’ (the nitrobenzhydrol tetracarboxylate Ca²⁺ chelator nitr5). Uncaging caged Ca²⁺ elevated the cytosolic free Ca²⁺ concentration without stimulating Ca²⁺ influx. Although the concentration of cytosolic free Ca²⁺ in this experiment approximately matched that of the IP₃-induced Ca²⁺ release, neutrophil flattening was not triggered (Fig. 4A–C). We found that nitr5 itself or its photolytic release products did not prevent neutrophils from spreading because f-mlp (formyl-methionyl-leucyl-phenylalanine) was able to induce normal cell spreading in nitr5-loaded cells after photolysis of nitr5 (Fig. 4A–C; supplementary material Movie 2). This confirmed (1) that photolysis itself was not a stimulus for neutrophil spreading, and (2) that the Ca²⁺-release signal induced by IP₃ was below the critical concentration or in the wrong cellular location to trigger spreading. This result is consistent with an earlier study in which it was necessary to elevate global cytosolic free Ca²⁺ to an unphysiologically high level (∼30 µM) to induce neutrophil flattening (Pettit and Hallett, 1998). That the crucial Ca²⁺ component was Ca²⁺ influx was confirmed by using Ni²⁺ ions to block Ca²⁺ influx without inhibiting IP₃-induced Ca²⁺ store release (see Hillson and Hallett, 2007); this inhibited IP₃-induced neutrophil spreading (Fig. 3C). We therefore concluded that IP₃ itself was not the crucial component for cell spreading, but rather that it was physiological Ca²⁺ influx through Ca²⁺ channels opened on the plasma membrane as a result of IP₃-mediated Ca²⁺ store release.

Fig. 4.

Factors affecting neutrophil spreading. The effect of uncaging Ca²⁺ (nitr5) to a level of the initial (release) phase of the caged IP₃-induced Ca²⁺ signal. (A) The matching of initial (release) phase of the caged IP₃-induced Ca²⁺ signal (upper graph ‘cIP₃’) and nitr5 uncaging followed by addition of f-mlp (lower graph). (B) Graphs are showing the effect of f-mlp addition from A in more detail, with the accompanying effect on the cell ‘footprint’ and (C) images of the cell and its cytosolic free Ca²⁺ from which the data was taken. (D–F) The effect of the pre-treatment of neutrophils with cytochalasin B (5 µg/ml) before uncaging IP₃: (D) a series of phase contrast and fluo4 images, (E) quantification of cytosolic free Ca²⁺, footprint and cell roundness [taken as (cell perimeter/π.max cell dimension)], and (F) changes in the reconstructed isosurface of the cell and the position of the nucleus, with the arrowheads indicating local sites of cell deformation. An example of the ‘wriggling response’ to IP₃ uncaging in cytochalasin-B-treated neutrophils is shown in supplementary material Movie 2.

Actin-polymerisation-independent surface area changes

Although a rise in cytosolic free Ca²⁺ is not required for actin polymerisation in neutrophils (al-Mohanna and Hallett, 1990), actin polymerisation clearly drives cell spreading because the IP₃-induced Ca²⁺ influx signal failed to trigger spreading in neutrophils treated with the actin polymerisation inhibitor cytochalasin B (Fig. 4D,E). However, under these conditions, the IP₃-induced Ca²⁺ signal triggered a noticeable periodic loss of sphericity of the cells as they appeared to ‘wriggle’ (Fig. 4D,E; supplementary material Movie 2). This loss of sphericity could not be attributed to the nuclear lobes ‘pushing against’ the plasma membrane as 3D imaging of the cells showed that the lobes of the multi-lobular nucleus remained in a tight formation and that bulges in the cell surface did not correlate with places that could be directly ‘pushed’ by nuclear lobes (Fig. 4F). As the spherical shape is the minimum surface area for a given volume, loss of sphericity must be accompanied by an increase in surface area. Cell wriggling therefore was a manifestation of an actin-polymerisation-independent cell surface area increase.

Cell surface area expansion without new membrane insertion

The plasma membrane has no facility to stretch beyond ∼5% of its length as lateral forces pull apart the phospholipids in the bilayer. The mechanism by which the surface area of the neutrophil apparently expands must therefore involve either the addition of new membrane, possibly by fusion of vesicles from within the cell, or the unfurling of its wrinkled membrane (Dewitt and Hallett, 2007; Hallett and Dewitt, 2007). The latter possibility was proposed over 30 years ago (Erickson and Trinkaus, 1976), but, until now, it has proved difficult to provide evidence for the mechanism of membrane expansion. The two possibilities can be distinguished by the use of FM1-43, which binds to membrane and increases the quantum yield of its fluorescence. The addition of new internal membranes to the plasma membrane as the surface area expands would thus give an increase in the total cellular FM1-43 fluorescence with the intensity of FM1-43 in the ‘expanding’ plasma membrane remaining constant. In contrast, the unfurling of wrinkled membrane would result in the total cellular FM1-43 fluorescence remaining constant, while FM1-43 intensity in the ‘expanding’ plasma membrane would decrease. The latter was found to be the case, with total cell FM1-43 fluorescence remaining constant during cell spreading and the fluorescence intensity of plasma membrane decreasing (Fig. 5A,B). This is despite the abundance of intracellular membranes within the cell that are capable of being staining by FM1-43, which are revealed by permeabilising the plasma membrane (Fig. 5B). In cells with asymmetric spreading, the FM1-43 signal intensity in the sub-region of the membrane, which did not expand, remained constant (Fig. 5B,C). Although these measurements were performed with the confocal pinhole opened wide to increase the thickness of optical sections, and so minimise any artefactual decrease in fluorescence signal owing to the movement of the spreading cell below the confocal plane, this possibility was excluded by the use of a second membrane marker, DiI. As the DiI intensity would be subject to the same optical or cell movement artefacts as FM1-43, the ratio of FM1-43 intensity to DiI intensity would be insensitive to these potential artefacts. However, unlike FM1-43, DiI was loaded into the membrane before observation and was absent from the extracellular medium during the experiment. Hence, DiI would decrease in intensity as a result of either the addition of new (unstained) intracellular membrane or unfurling of wrinkled membrane. The ratio of FM1-43:DiI would therefore have greater sensitivity than the FM1-43 signal alone. This ratio (FM1-43:DiI) failed to increase on cell spreading, and, in fact, often decreased slightly (Fig. 5C), and was the same in regions of the cells that had flattened and those that had not. Although this approach might not detect a small amount of membrane insertion (given fluorescent signal increases <10%), such an insertion would not contribute significantly to the increase in surface area. These data were therefore consistent with the unwrinkling mechanism for membrane expansion being dominant rather than the insertion of new membrane.

Fig. 5.

The mechanism of membrane expansion during cell spreading. Time series of images of neutrophils in the presence of FM1-43 with the accompanying phase contrast images for (A) an unstimulated neutrophil, which remains approximately spherical and has a constant FM1-43 signal (quantified in the graph as ‘footprint’ and FM intensity respectively), (B) a neutrophil undergoing rapid spreading, with the final fluorescent image resulting from the permeabilisation of the plasma membrane with digitonin (DGT; 150 µM), and (C) a rapidly spreading neutrophil pre-stained with DiI (red) in the presence of FM1-43 (green) as in the previous two panels. In C, the ratio of the FM1-43 and DiI signals is shown together with the change in cell footprint.

The crucial role of calpain activation

It has been proposed that plasma membrane wrinkles are held in place by calpain-sensitive cytoskeletal elements (Hallett and Dewitt, 2007; Dewitt and Hallett, 2007), and that localised high Ca²⁺ restricted to wrinkled membrane (Brasen et al., 2010) might activate sub-plasma membrane calpain-1, whose K_d is ∼30 µM (Goll et al., 2003). Calpain activity, monitored using a ratiometric fluorescence approach, to avoid the effect of optical path-length changes during cell spreading on the measurement (Dewitt et al., 2009; see Materials and Methods), increased during cell spreading (Fig. 6A). However, measuring activity using this approach gives limited information because (1) it is not possible to synchronously monitor cytosolic free Ca²⁺ (fura2 or fluo4 requires excitation in the 340–390 nm or 470–490 nm regions thus preventing simultaneous calpain measurement), and (2) it is not possible to time the onset of calpain activation and so correlate this to cell spreading (the signal observed was the result of the accumulation of the product of calpain activation, which accompanied neutrophil spreading). However, the method showed that calpain was activated and permitted effect of the use of pharmacological inhibitors of calpain to be established. Two classes of calpain inhibitors were used, mono-halide mercaptoacrylates, such as PD 150606 (Adams et al., 2012), which acts on a regulatory site required for Ca²⁺ triggering of the proteolytic activity, and calpeptin, which inhibits the proteolytic domain of the enzyme. The spreading morphological changes were reduced by these inhibitors at concentrations which inhibited the calpain activity increase associated with cell spreading. Electric cell-substrate impedance sensing (ECIS) showed that calpain inhibition significantly reduced the rate of neutrophils spreading on either the electrode itself or fibronectin-coated electrodes (Fig. 6B). The inhibitors also had a dramatic and clear inhibitory effect on IP₃-induced neutrophil spreading: on IP₃ uncaging, the evoked Ca²⁺ signal remained at levels that would stimulate rapid cell spreading (Fig. 6C), but the majority of cells failed to spread on either glass or integrin-engaging fibronectin (Fig. 6C,D). However, on glass, a small number of calpain-suppressed cells exhibited a slow and aberrant type of cell spreading. Unlike in cells with active calpain, where cell spreading was synchronised and uniform over the majority of the cell periphery (Figs 1,3), these calpain-suppressed cells flattened by producing numerous small bleb-like structures, which subsequently coalesced into a larger area and gave rise to an asymmetrical spreading response (Fig. 7A; supplementary material Movie 3). The FM1-43 signal in these mini-blebs decreased (Fig. 7B), which is consistent with them resulting from localised plasma membrane unwrinkling (Fig. 7D). We propose that, in the absence of full calpain activity, the release of membrane wrinkles is slowed sufficiently for the cellular events that normally accompany rapid neutrophil spreading to be temporally resolved into two clear stages of membrane expansion: first, the release of surface wrinkles, which is, second, permissive for the subsequent actin-dependent morphology change.

Fig. 6.

The role of calpain activity in neutrophil spreading. (A) Calpain activity monitored by the ratio of the signals from fluorogenic substrate Suc-LY-AMC and fluorescein in two neutrophils undergoing spreading is shown as a series of ratio images from which the graph was constructed. (B) The effect of a calpain inhibitor on the impedance, quantified by ECIS as a measure of spreading of neutrophils. The graphs shown are the means (black lines) and ±s.d. (grey lines) for five electrodes (the total cell number was ∼125). (C) Sample data comparing the effect of two calpain inhibitors (calpeptin, CPTN) and PD151746 (PDa) with a solvent control (DMSO). In all sections, uncaging of IP₃ was at 30 seconds, and the fluo4 (Ca²⁺) and phase contrast images are shown together with the quantification of cytosolic free Ca²⁺ and the contact area of the cell (footprint). (D) The percentage of neutrophils that did not spread but were adherent (grey area of bars) and those that spread (black area of bars) is shown for three calpain inhibitors [PD150606 (PD1), PD151746(PD2) and calpeptin (CPT)] after IP₃ uncaging. Cells were cultured on glass or fibronectin (indicated by ‘F’).

Fig. 7.

Aberrant spreading mechanism in calpain-inhibited cells. (A) A series of phase contrast images that are typical of neutrophils that spread after calpain inhibition. Unlike uninhibited cells, clear ‘bleb-like’ structures form before spreading. The complete data set for this is shown in supplementary material Movie 3. (B) A close up view of a ‘bleb-like’ structure in a neutrophil pre-stained with DiI (red) in the presence of FM1-43 (green) showing that FM1-43 intensity decreases, consistent with unfurling of existing wrinkled membrane, rather than the addition of new membrane. (C) The kinetics of cell spreading under calpain inhibition, where ‘steps’ in the increase in cell area are observed that correlate with the formation of ‘bleb-like’ structures (indicated by the asterisks). (D,E) Our hypothesis for the underlying mechanism at a macroscale (D), where the wrinkled membrane locally and patchily unfurls; and at a micro-scale (E), where the molecules holding the wrinkles in place are cleaved by activated calpain.

Discussion

In this study, we have shown that elevated cytosolic IP₃ is an initiator of the Ca²⁺ influx that stimulates the morphological flattening of neutrophils. The Ca²⁺ influx is sufficient to activate cytosolic calpain, an event that is crucial for permitting spreading to proceed and for the apparent expansion in plasma membrane area. Because the apparent expansion is not accompanied by the addition of new membrane, we propose that the wrinkles on the surface of the neutrophils represent a ‘membrane reservoir’ and that calpain mediates the unfurling of these wrinkles.

Activation of phospholipase C (PLC), which generates IP₃, probably plays a crucial role in physiological neutrophil spreading. Inhibition of PLC activity by compound U73122 (Bleasdale et al., 1990) has a profound inhibitory effect on neutrophil spreading (Smith et al., 1996). Similarly, in lymphocytes, the IP₃-generating enzyme (PLCγ2) was found to be crucial for cell spreading (Weber et al., 2008). However, the authors of this latter study concluded that, in these lymphocytes, IP₃ itself was not involved because cell spreading occurred in mutant cells lacking the three types of IP₃ receptor (Weber et al., 2008). Clearly, our study shows that photolytically generated IP₃, produced to a level sufficient to trigger a Ca²⁺ signal, induces cell spreading in human neutrophils. However, we must be careful not to extrapolate from this that IP₃ is involved in neutrophil spreading physiologically. Indeed, at the integrin-engaging surface in myeloid cells, there is a strong and localised production of phosphotidylinositol (3,4,5)-trisphosphate (PIP₃) (Dewitt et al., 2006), which might be involved in Ca²⁺ influx by a non-IP₃ route. The data in this present paper shows that the Ca²⁺ influx itself, however it is generated physiologically, is a signal for rapid neutrophil spreading.

In other cell types, there is growing evidence for a role of calpain-2 in focal adhesion complexes and slow motility (Franco et al., 2004; Franco and Huttenlocher 2005). Neutrophils, however, mainly express calpain-1 and do not form focal adhesion complexes, and there are two reports that calpain inhibitors increase neutrophil motility (Lokuta et al., 2003; Katsube et al., 2008). Although it is not clear whether this latter effect was a result of decreased cell adhesion, there is a significant body of evidence showing that calpain inhibitors are effective in preventing neutrophil extravasation during experimental inflammation in animal models (Noble et al., 1998; Ikeda et al., 2002; Tissier et al., 2004; Cuzzocrea et al., 2000; Marzocco et al., 2004; Yoshifuji et al., 2005) and that calpain inhibition prevents full extension of pseudopodia during phagocytosis by neutrophils (Dewitt and Hallett, 2002) and spreading by lymphocytes (Stewart et al., 1998) and platelets (Croce et al., 1999). In our studies, calpain-inhibited cells were essentially quiescent, or sluggish even after IP₃ uncaging (e.g. see Fig. 6). The ‘blebbing prelude’ to slow cell spreading, which we observed in calpain-suppressed neutrophils (Fig. 7), might occur physiologically in some cells (Fackler and Grosse, 2008; Charras and Paluch, 2008), especially in Dictyostelium, where recent reports have shown that blebbing even accompanies chemotaxis (Yoshida and Soldati, 2006; Langridge and Kay, 2006).

Our study therefore demonstrates that rapid cell spreading is triggered by the influx of Ca²⁺ through physiologically opened Ca²⁺ channels. This depends on calpain activation to release tethered undulations (wrinkles) in the plasma membrane, which, in turn, permits the apparent plasma membrane expansion. It is possible that the activation of calpain in neutrophils leads to the cleavage of an important set of cytoskeletal plasma membrane linkage molecules, including talin (Sampath et al., 1998) and ezrin (Potter et al., 1998), which are responsible for maintaining the membrane its is wrinkled state and thus acts as the permissive step in rapid neutrophil spreading (Dewitt and Hallett, 2007; Hallett and Dewitt, 2007).

Methods and Materials

Cell preparation and probe loading

Human neutrophils were isolated from the blood of healthy volunteers as described previously (Pettit and Hallett, 1995) as approved by SMREC ethics board (Cardiff University School of Medicine Research Ethics Committee, Cardiff, UK) with the informed consent of participants. The neutrophils were then suspended in Krebs medium (120 mM NaCl, 4.9 mM KCl 1.2 mM KH₂PO₄, 1.2 mM MgSO₄, 1.3 mM CaCl₂, 25 mM HEPES and 0.1% BSA, adjusted to pH 7.4 with NaOH). The neutrophils were loaded with the fluorescent Ca²⁺ probe fluo4 acetoxylmethyl ester (AM) (Molecular Probes; 1–5 µM) as described in detail elsewhere (Hillson et al., 2007). This produced cytosolic concentrations of the probes of 50–100 µM. When loading neutrophils with both Ca²⁺ probe and caged-IP₃ or caged Ca²⁺ (nitr5), the incubation with the caged IP₃ propionyloxymethyl ester (PM) (Alexis Biochem) or nitr5-AM (Biochem) was started before adding the fluo4-AM compound. Loading with caged compounds was performed in the dark to prevent uncontrolled uncaging by ambient light. The times for loading of the probes were: fluo4-AM, 40 minutes, IP₃ (caged)-PM, 1 hour; and nitr5-AM (1 hour). Extreme care was taken during the isolation and loading procedures to avoid mechanical activation of the cells, which can cause them to adhere rapidly and spontaneously to uncoated glass. Under the isolation conditions used here, neutrophils adhere to glass, but remain in the apparently ‘spherical’ (i.e. non-spread) morphology.

Ca²⁺ measurement and manipulation

Cytosolic free Ca²⁺ was measured as described in detail elsewhere (Hillson et al., 2007) using the resonant scanning head of the Leica RS confocal microscope, which had a frequency of ∼40 MHz. Neutrophils were allowed to adhere to the glass coverslip mounted onto a thermostatically controlled stage (37±0.1°C) in Krebs medium and soluble stimuli such as f-mlp (Sigma) were added to them during a data acquisition run. During image acquisition, uncaging was achieved by positioning of an additional dichroic mirror and band-pass filter to illuminate the microscopic field with UV (330–300 nm) light (430 DCLPO2 dichroic; Omega Optics) from a 100 W mercury arc lamp. It was shown, using caged fluorescein as a surrogate read-out, that this procedure generated photolytic product linearly with no initial time delay. After loading with caged IP₃ from its permeant ester, human neutrophils retain the ability to undergo adhesion and spontaneous spreading. In addition, other neutrophil functions, such as chemotaxis and phagocytosis were unimpaired. However, cell–cell adhesion (aggregation) in suspension was often observed to increase. This effect was attributed to the ester loading method, presumably esterase-cleavage products, rather than caged IP₃, as loading with other esters (e.g. those of fluo4, fluo3, fura2) can also induce this. The Ca²⁺ and morphological responses reported here were also mimicked in individual neutrophils that had been microinjected with non-esterified forms of fluo4 and caged IP₃ (Dewitt et al., 2003). Given that no esterase products are generated under these latter conditions, none of the phenomena reported can be attributed to these side products. From knowledge of the relative uncaging indices for caged IP₃ and nitr5 (εφ) (Ellis-Davies, 2007), photolysis was estimated to released IP₃ at a rate of 4 µM/second.

Quantification of Ca²⁺ and morphological changes

Adjustments to gain and offset were applied to whole time-series data sets to allow genuine comparisons within a time sequence. Ratio images (F/F_o) were taken to negate differences in Ca²⁺ probe loading, but this never revealed Ca²⁺ signals that were not evident in the raw fluorescent intensity data set. Cytosolic free Ca²⁺ was estimated using the following equation, which the K_d for fluo4 taken as 516 nM (Hillson and Hallett, 2007), Ca²⁺ = K_d[F−F_min]/[F_max−F]. F_min and F_max were found for Ca²⁺-free and Ca²⁺-saturating probes as described (Hillson et al., 2007). The cell morphology change during neutrophil spreading was dramatic and easily quantified in populations by counting individual cells. For time relationships, the cell contact area (footprint) was measured frame by frame using Image Master (PTI) or ImageJ software. The kinetics of cell spreading was also quantified by ‘electric cell-substrate impedance sensing’ (ECIS, Applied Biophysics), which is a non-invasive means of monitoring the kinetics of spreading of very small populations of neutrophils (about five cells per electrode) onto a 250-µm-diameter gold electrode coated with fibronectin. As neutrophils spread, they reduce the effective area of the electrode and the impedance signal rises (Wegener et al., 2000). All measurements were performed at 37°C and the final outcome was confirmed by microscopic inspection of the electrodes. For 3D reconstruction, optical xy planes were taken at z planes through cells whose plasma membrane had been stained with DiI (Invitrogen) or PKH (Sigma), and with either the nuclear membrane stained with PKH or DNA content stained with Acridine Orange (Sigma). Optical sectioning was performed in real time as the cell morphology changed, using a fast resonant scanning system (Leica RS 2) and the data later reconstructed using Imaris software (Bitplane). All data shown are representative of a number of different experimental runs from at least three different blood donors as the source of neutrophils.

Calpain measurement and manipulation

Calpain activity was assessed in neutrophils by dual labelling with fluorescein from its diacetate ester (fluorescein diacetate, Sigma) and the cell-permeant fluorogenic calpain substrate, Suc-Leu-Tyr-7-amido-4-methyl coumarin (Suc-LY-AMC). The calpain-derived product, AMC was measured by excitation at 360 nm and fluorescein by excitation at 470 nm using a rapid changing monochromator (DeltaRAM, PTI, UK) and emission >510 nm was detected using an intensified CCD camera (PTI, UK). Ratio images (360:470 nm) were acquired in real time using ImageMaster software (PTI, UK). Calpain activity was inhibited by pre-incubation (15 minutes) with either calpeptin, which inhibits the proteolytic domain, or PD150606 or PD 151746 (50 µM), which inhibit the Ca²⁺ activation domain.

Measurement of membrane expansion

The membrane was stained by the addition of FM1-43 (Molecular Probes) to neutrophils in suspension (8 µM) and imaged confocally (Leica RS SP2; excitation 488 nm, emission 530–560 nm) with the pinhole set open (2 AU). The dye was present throughout the experiment. In experiments with DiI, the cell membranes were pre-stained (as described above) before the addition of FM1-43.

Imaging and quantification

Confocal imaging was achieved using a resonant scanning head for the Leica RS confocal laser microscope with a 63× oil immersion objective (NA 1.32) (HCX- PL- APO). The confocal pinhole was set to 1 AU for all experiments except where FM1-43 was used, where it was adjusted to 2 AU in order to increase the thickness of the confocal slice sufficiently to monitor total cell fluorescence. Neutrophils were allowed to adhere onto the glass coverslip mounted on a thermostatically controlled stage (37±0.1°C) in Krebs medium and soluble stimuli (f-mlp) were delivered or uncaging achieved during image acquisition. Adjustments to gain and offset were applied to whole time-series data sets to allow genuine comparisons within a time sequence. ‘Raw fluorescent’ images with no post acquisition contrast enhancement or other manipulation were used for quantification. The intensity and area occupied by the spreading cell was measured using either Leica analysis software, ImageMaster analysis software (PTI) or Image J (NIH). The number of individual neutrophils analysed exceeded 50 in all experiments. All data shown were representative of at least three different experimental runs from different blood donors as the source of neutrophils.