Main

The newborn is particularly susceptible to overwhelming bacterial and fungal sepsis(1). This is in part related to deficiencies in the first line host defenses provided by phagocytic cells(2). Defective newborn PMN chemotaxis may be demonstrated in vitro and in vivo(3–5), and may be observed in clinical disease states such as pneumonia, meningitis, and necrotizing enterocolitis in which there is a paucity of PMNs at the site of infection(6). The molecular basis for reduced newborn PMN chemotaxis is poorly understood.

The polymerization of monomeric G-actin to form filamentous F-actin is an essential early step in PMN locomotion(7). Studies of this phenomenon in human newborn PMNs have shown that it is reduced compared with that of adult PMNs(8–11). The reasons for this reduction in actin polymerization are unclear, but it may be related to reduced signaling consequent to reduced expression of numerous cell surface receptors, including Mac-1 and LFA-1(12–15). Contact between the contractile actin filament meshwork and the adherent cell membrane provides traction for cell locomotion(16). Directed movement of the cell in a chemotactic gradient depends on cytoskeletal rearrangements and the assumption of polarized morphology by the cell with concentration of F-actin in the anterior lamellipodium and posterior uropod(17).

PKC, a phospholipid-dependent serine-threonine kinase, of which there are at least 12 isoforms, regulates biologic processes in many eukaryotic cells(18,19). Differential function and location, of individual isoforms in specific cellular locations, probably relate to individual substrate specificity(20). The most abundant isoforms in PMN are α, β, and ζ(19). Observations regarding PKC activation and actin polymerization suggest that these events may be linked. Chemoattractants, which stimulate actin polymerization, such as the bacterial peptide, FMLP, act via membrane receptors, and the subsequent cascade of enzyme activation results in phospholipid hydrolysis and the generation of DAGs which activate PKC. Phorbol esters and synthetic DAG are known to directly stimulate PKC(21), and treatment of intact PMN with these substances causes PKC to translocate from the cytosol to the membrane(22). They also cause actin polymerization, albeit to a lesser degree and with a slower kinetic than chemoattractants such as formyl peptides(23). Despite these observations, inhibition of PKC has not been shown to alter actin polymerization in adult PMNs(23,24).

Protein phosphorylation is central to signal transduction. Cellular phosphatases constitute a counterregulatory system to the kinases. FMLP has been shown to rapidly and transiently activate the MAPKK and MAPK in granulocytes. Activation is maximal at 1 min and is decreasing at 10 min(25). This kinetic closely parallels that of the actin polymerization response to FMLP. In the same study, activation of MAPKK and MAPK by PMA was slow and sustained and was maximal at 5 min; again the kinetic closely parallels that of the actin polymerization response to the same stimulant. Activation of MAPKK and MAPK was antagonized by phosphatases 1 and 2A. In our study we have investigated the role of both phosphorylation and dephosphorylation, in the regulation of chemotaxis, by using kinase and phosphatase-specific inhibitors.

Insight into control of chemotaxis by phosphorylation may help in the understanding of the newborn chemotactic defect, whereas characterization of PMN function in this defective model may provide insights into PMN function in general.

METHODS

FMLP, PMA, cytochalasin B, calyculin A, okadaic acid, and staurosporine were obtained from Sigma Chemical Co., Poole, Dorset, UK. BIM was obtained from Calbiochem Nottingham, UK. All these agents were dissolved initially in DMSO and diluted for use in HBSS (GIBCO Life Technologies, Paisley, Scotland). Aliquots were stored at -20°C. The concentration of DMSO in experiments did not exceed 0.1%. Zymosan A, lysophosphatidyl choline, hematoxylin, and DPX mountant were obtained from Sigma Chemical Co. BODIPY FL phallacidin was from Molecular Probes Europe BV, Leiden, Holland. PBS tablets were from Oxoid, Unipath Ltd, Basingstoke, UK. EDTA and centrifuge tubes were from Sarstedt, Wexford, Ireland, and Falcon 2054 tubes for flow cytometry were from Falcon, Becton Dickinson Laboratory Supplies, Franklin Lakes, NJ. Cell culture plates were from Bibby Sterilin, Stone, Staffordshire, UK, and 3-µm pore size filters, type "SS," were from Millipore, Bedford, MA. Filter paper disks (antibiotic test) were from Whatman International Ltd, Maidstone, Kent, UK. MonoPoly resolving medium was from ICN Flow Laboratories, Aurora, OH, and Histopaque was from Sigma Chemical Co. Hydroxyethylstarch 6% solution in 0.9% NaCl ("Haesteril") was from Fresenius AG, Cheshire, UK.

PMN isolation. Venous blood was obtained from healthy adult volunteers. Newborn blood was obtained from the placental side of umbilical cords after full-term normal delivery of healthy infants. All samples were collected in EDTA (0.01%) tubes.

PMNs were isolated first by sedimentation: whole blood was mixed 1:1 with 6% hydroxythyl starch and allowed to settle for 30 min at room temperature. The leukocyte-rich supernatant was then layered over a double density gradient in 15-mL conical bottom plastic tubes. The gradient consisted of a lower layer of 3 mL of MonoPoly resolving medium, and an upper layer of 2 mL of Histopaque. Tubes were centrifuged for 30 min at 400 × g at room temperature, after which PMN and mononuclear cells formed two distinct layers(26). Viability and purity were checked by fluorescent microscopy of cells stained with ethidium bromide/acridine orange; no cell suspension with less than 80% viability and purity was used.

For chemotaxis experiments, cells were washed twice at room temperature in PBS, and suspended for use in PBS with added glucose (0.1% wt/vol), and CaCl2 and MgCl2 (both 0.1 mM). For actin polymerization studies, cells were washed twice at 4°C in HBSS with calcium and magnesium, then suspended and kept in HBSS over ice until use. A standard cell suspension of 106 cells/mL was used for all experiments.

Chemotaxis assay. Chemotaxis was measured by a modified micropore filter assay(27). The chemoattractant was ZAS (10%) prepared as described previously(3) from serum from a single healthy adult donor, diluted 1:10 with PBS, and frozen in 1-mL aliquots at -70°C until use. Fifteen-millimeter wells on 12-well cell culture plates were used, into which were placed 250 µL of chemoattractant. Aliquofs of enzyme inhibitors for phosphorylation experiments were added as appropriate to the required concentration. A 10-mm filter paper disk was then placed and soaked in the solution, over which was placed a 10-mm diameter, 3-µm pore size Millipore filter. A small quantity of cell suspension in a plastic container was then inverted over the Millipore filter, and the apparatus was incubated in a humid atmosphere for 1 h at 37°C. We used 180 µL of cell suspension in a test tube cap, the diameter of which was 7 mm. Aliquots of enzyme inhibitors as appropriate were added to the cell suspension. All experiments were carried out in duplicate. After incubation, the Millipore filters were rinsed in PBS, fixed in propan-2-ol, stained with hematoxylin, and mounted on glass slides with DPX mountant. They were then viewed by light microscopy (Leitz Dialux 22). The distance migrated downward toward the chemoattractant, measured on the micrometer scale of the microscope, was that between the cells resting on the upper surface of the Millipore filter and the leading front, defined as the farthest point where two cells were in focus at ×40 magnification(28). This was measured at three random points on each filter, the result was expressed as a mean of all values on duplicate filters.

Actin polymerization. Actin polymerization was measured by flow cytometry (FACScan, Becton Dickinson, San Jose, CA) according to a previously described method(7,29). Experiments on samples from each donor were performed in triplicate. A cell suspension(250 µL) was placed in a flow cytometry plastic tube and warmed in a water bath at 37°C for 5 min. For resting values, cells were fixed and permeabilized at this point by addition of formaldehyde to a final concentration of 3.7% and lysophosphatidyl choline (100 µg/mL), gently mixed, and incubated for 5 min at 37°C. Cells were then stained with BODIPY FL phallacidin (1.65 × 10-7 M), and incubated at 37°C for 10 min. F-actin was expressed as mean fluorescence intensity. For analysis of data, F-actin levels, for each sample, were expressed relative to their respective resting values, as RFI.

Enzyme inhibitors. Staurosporine and BIM were both used at 10-6 M concentration, which is well above the IC50 for both inhibitors(30,31), for studies of actin polymerization. BIM is a potent and highly selective inhibitor of PKC(30). Staurosporine inhibits PKC and numerous other cellular kinases(30,31). We found that the phosphatase inhibitors calyculin A and okadaic acid altered chemotaxis at 0.01 and 1 µM, respectively, therefore these inhibitors were used at or above these concentrations for actin polymerization studies.

For studies of actin polymerization, cells were first warmed for 5 min at 37°C before treatment with enzyme inhibitors to avoid interference with the temperature induced rise in F-actin, then co-incubated for 5 min with inhibitors before stimulation. Treatment for this duration has been shown to have measurable effects on phosphorylation(23,24,32). FMLP was used as stimulant in these experiments, because of cell clumping induced by ZAS at incubation times greater than 5 min. In actin polymerization experiments, which involved preincubation of PMNs with enzyme inhibitors, the data obtained was more consistent and reproducible, when FMLP was used as stimulant.

Statistical analysis. Chemotaxis data were expressed as mean distance, in micrometers, migrated by neutrophils in 1 h, plus or minus SD, using three measurements from each of duplicate experiments. Actin polymerization data were expressed as mean relative fluorescence intensity plus or minus SD from triplicate experiments. Median values were analyzed by the Mann Whitney U nonparametric test, using an IBM XT personal computer with the INSTAT 2 software package (Graphpad Software, San Diego, CA). ANOVA was used to determine the effects of inhibitors on chemotaxis of adult and newborn PMNs. p < 0.05 was considered significant.

RESULTS

Chemotaxis and actin polymerization in PMNs. PMN chemotaxis in response to ZAS (10%) was significantly reduced in healthy newborn cord blood PMNs (median 18 µm, range 9-21 µm) relative to adult PMNs(median 23 µm, range 17-34 µm) (p < 0.001)(Fig. 1). Cytochalasin B, used as a positive control, inhibited chemotaxis in adult PMNs (median 5 µm, range 3-7 µm, n = 5).

Figure 1
figure 1

PMN chemotaxis in newborns and adults. Results are expressed as median distance (µm) ± SD migrated by PMNs in 1 h into a 3-µm filter toward ZAS (10%) and in the presence of PKC inhibitors. (Left to right) ZAS (10%) only, staurosporine(10-8 M), (10-6 M), BIM (10-6 M), (10-5 M). *p < 0.05, adult PMNs: ZAS only vs ZAS + staurosporine.

Full size image

Resting F-actin levels were similar in healthy newborn and adult PMNs(median 62.6 versus 63.4). Stimulation of adult PMNs with ZAS resulted in F-actin levels peaking rapidly at 15 s to approximately twice resting levels and gradually declining over 5-10 min to resting levels(Fig. 2). Peak F-actin levels were significantly reduced in newborn relative to adult PMNs after ZAS (10%) stimulation (median RFI 1.74 versus 2.10, p < 0.05).

Figure 2
figure 2

PMN actin polymerization in adults and newborns in response to ZAS (10%). The values represent median relative F-actin levels, i.e. stimulated relative to baseline, expressed as RFI, means from triplicate experiments, measured at baseline (0), and 15, 60, and 300 s of stimulation. *p < 0.05. Newborn n = 6, adult n = 8.

Full size image

The actin polymerization response to FMLP was dose-dependent, and the peak F-actin level at 15 s was significantly reduced in the newborns at 5 and 10 nM FMLP (p < 0.05). The time kinetics and magnitude of the PMN actin polymerization response to ZAS (10%) was similar to that seen in response to FMLP (10 nM) (Figs. 2 and 3).

Figure 3
figure 3

The effect of phosphatase inhibitors on PMN actin polymerization in response to FMLP (10 nM). Values represent median relative F-actin levels, i.e. stimulated relative to resting, expressed as RFI, means from triplicate experiments, measured at baseline, and at 15, 60, and 300 s post stimulation. Coincubation with okadaic acid (1µM), or calyculin A (0.02 µM) did not significantly alter the response to FMLP. Adult n = 8, newborn n = 4.

Full size image

There was a small early rise in RFI to approximately 1.1 with control DMSO(0.1% vol/vol). This also occurred with PBS and it appears to be a nonspecific effect of shaking the cell suspension.

When PMA was used as a stimulant, F-actin levels peaked at 8 min post addition, at approximately twice resting levels (Fig. 4). Actin polymerization in adult and newborn PMN was similar after stimulation with PMA (10-8 M) (Fig 4).

Figure 4
figure 4

Actin polymerization in adult(n = 10) vs newborn (n = 9) PMNs stimulated with PMA (10 nM). The values represent median relative F-actin levels, i.e. stimulated relative to resting, expressed as RFI, means from triplicate experiments, measured at baseline (0), and 1, 2, 3, 5, 8, and 10 min of stimulation.

Full size image

Influence of PKC inhibitors on PMN chemotaxis and actin polymerization. Co-incubation with the PKC inhibitors staurosporine or BIM, did not significantly alter chemotaxis of newborn PMNs(Fig. 1). The same was true for adult PMNs with the exception of staurosporine at the higher concentration of 10-6 M, where median chemotaxis was reduced.

BIM did not cause any change in baseline relative F-actin levels before stimulation and had no effect on subsequent actin polymerization in either adult or newborn PMNs. Median RFI at 15 s of stimulation with FMLP(10-8 M) alone was 2.52 in adult PMNs, and was 1.56 in newborn PMNs. When PMNs were preincubated with BIM (10-6 M), median RFIs were 2.62 and 1.51, respectively (p > 0.05).

Influence of phosphatase inhibitors on PMN chemotaxis and actin polymerization. Coincubation with calyculin A, an inhibitor of phosphatases 1 and 2A, caused significant parallel dose dependent reductions in chemotaxis in adult (p < 0.001) and newborn PMN (p< 0.001) (Fig. 5). Coincubation with the less potent okadaic acid also caused similar reductions in chemotaxis in adult(p < 0.001) and newborn (p < 0.005) PMNs(Fig. 5). Many of the cells treated with calyculin A and okadaic acid did not migrate at all but assumed a rounded shape and remained on the surface of the filter. That this was due to nonspecific toxicity was excluded by examining trypan blue exclusion after incubating the cells with calyculin A or okadaic acid. There was no decrease in viability relative to controls (data not shown). Preincubation with calyculin A or okadaic acid(10-6 M) did not alter actin polymerization of either cell population (Fig. 3).

Figure 5
figure 5

The effect of phosphatase inhibitors on PMN chemotaxis. Distance migrated was reduced by coincubation with calyculin A (adult PMNs p < 0.001, newborn PMNs p < 0.001), or okadaic acid (adult PMNs p < 0.001, newborn PMNs p < 0.005). Chemotaxis is expressed as median distance(µm) ± SD, means from duplicate experiments, migrated by PMNs into a 3-µm pore size filter in 1 h toward ZAS (10%). (Left to right) ZAS only (control), okadaic acid, 1 and 2 µM, and calyculin A, 0.01, 0.02, 0.04, and 0.2 µM.

Full size image

DISCUSSION

Subtle differences have been reported in the literature relating to actin polymerization in newborn PMNs. An early study showed poor generation of intracellular Ca2+ and failure to generate F-actin(8). Another group showed elevated resting levels of F-actin in the newborn, with generation of similar amounts to adult PMNs upon stimulation; however, because of the presence of elevated basal levels, the relative rise was less in the newborn(9). A bovine study found newborn PMNs to be rapidly responsive but with a less sustained rise in F-actin in response to platelet-activating factor and a somewhat reduced response to C5a(10). A more recent human study showed reduced actin polymerization in response to FMLP and platelet-activating factor(11). All these data indicate that actin polymerization is reduced in the newborn PMN, but the nature and extent of the deficit remain undefined.

Our data show that peak F-actin is reduced in newborn PMNs in response to the chemotactic stimuli, FMLP and ZAS, which act via surface receptors, but not to PMA which acts directly on PKC at the cytoplasmic level. This suggests that there may be a defect in signal transduction in the newborn PMN, and the importance of phosphorylation and of dephosphorylation by phosphatase activity in this process was therefore investigated.

The sequence of biochemical events from PMN surface receptor stimulation to actin polymerization is complex and involves the hydrolysis of membrane GTP binding ("G") proteins which in turn activate phospholipase Cγ), with subsequent cleavage of membrane inositol phospholipids, principally inositol 4,5-biphosphate to form inositol triphosphate and DAG(33). Inositol triphosphate facilitates Ca2+ channel opening and Ca2+ influx, and there is an early peak in intracellular [Ca2+]. DAG activates PKC, which can then regulate several intracellular processes through many pathways(33).

Despite reported links between cell activation, actin polymerization, and PKC activation, coincubation with PKC inhibitors did not consistently influence newborn or adult PMN chemotaxis in vitro in this study. There was a reduction in the distance migrated by adult PMNs toward a chemotactic stimulus in the presence of the higher concentration of staurosporine (Fig. 1), but this did not occur with the potent and highly specific PKC inhibitor BIM. The relatively nonspecific PKC inhibitor H7 has previously been shown not to influence adult PMN chemotaxis in an agarose gel assay system(34), whereas phorbol esters and DAG analogs, both of which stimulate PKC, inhibit in vitro chemotaxis(35). Inhibition of DAG production by treatment of PMN with propranolol(34) also has no effect on chemotaxis.

Our findings showing a lack of influence of PKC inhibitors on actin polymerization are similar to those in previous studies in adult PMNs which demonstrate an early rise in F-actin in response to treatment with H7 and staurosporine; this does not occur with more specific PKC inhibitors such as BIM, and there is no effect on subsequent actin polymerization in response to FMLP or PMA(24,36). PKC inhibitors do not influence PMA induced actin polymerization; this is despite the abolition of PKC induced phosphorylation in PMN treated with staurosporine (10 nM) for 5 min(23). PMN actin polymerization in response to PMA, therefore, does not appear to be mediated via PKC.

The phosphatase inhibitors okadaic acid and calyculin A both produced similar dose dependent reductions in chemotaxis in adult and newborn PMN, with abolition of chemotaxis at high doses of calyculin A. This occurred despite a lack of any effect of these agents on actin polymerization. Downey et al.(23) showed greatly increased phosphorylation in PMN treated with okadaic acid for 5 min, yet no alteration in actin polymerization. Okadaic acid and calyculin A reduced FMLP-induced actin polymerization at 30 min, but had no effect on the early rise in F-actin(32), which we have studied. We have likewise shown that the early rapid phase of actin polymerization in response to chemoattractant is unaffected by phosphatase inhibitors. The data from our study suggest that the actions of phosphatases 1 and 2A facilitate PMN chemotaxis, because we have shown that okadaic acid and calyculin A inhibit chemotaxis in a dose-dependent fashion and that this is not related to reduced cell viability. Our data show, however, that actin polymerization proceeded normally in the presence of phosphatase inhibitors, suggesting that the site of action of these enzymes is at a later stage than, or via a different signaling pathway to, actin polymerization.

PKC and the important PKC substrate, MARCKS, colocalize in a punctate distribution at sites of focal cell contacts in the membrane(37). MARCKS is an actin filament cross-linking protein which, when phosphorylated, dissociates from the cell membrane(37). When dephosphorylated by phosphatases, it reassociates with the cell membrane(37). Chemoattractant receptors become concentrated at the front of an advancing cell(35,38), thus promoting internal gradients. Selective activation of PKC in cell contacts at the front of a PMN is possibly important in "steering" the cell into a gradient. The study of in vitro chemotaxis on a synthetic substrate in a unidirectional chemotactic gradient, such as in our study, may not detect a subtle role of PKC in vivo. It is possible that phosphorylation of MARCKS and related substrates, by allowing dissociation of actin filaments, followed by dephosphorylation and reattachment, may facilitate the detachment-reattachment sequence in leukocyte motility proposed by Stossel(17). MARCKS has an amphipathic domain which can insert in plasma membranes; one of its functions in the dephosphorylated state is to anchor the cell membrane to the actin cytoskeleton and therefore to provide traction for forward locomotion(16). Inhibition of phosphatases would uncouple these structures. Such a role could explain the inhibition of chemotaxis by phosphatase inhibitors. It has been shown that inhibition of phosphatase 2B (calcineurin) prevents the recycling of integrin receptors to the front of an advancing cell, which is required for directed migration(39).

It is clear from our data that the cellular phosphatases play an essential role in PMN chemotaxis. The substrate(s) involved have not been identified, but one possible site of action is at the level of attachment between the cytoskeleton and the cell membrane. This is compatible with a lack of effect on the formation of actin filaments by actin polymerization. We showed no differences between adult and newborn PMNs regarding the effects of phosphatase inhibitors on chemotaxis and actin polymerization. It is therefore unlikely that the site of action of cellular phosphatases is the site of the defect causing reduced newborn PMN chemotaxis.