Published online before print December 14, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Montana State University, Department of Microbiology, Bozeman, Montana, USA
1 Correspondence: Montana State University, Department of Microbiology, 109 Lewis Hall, Bozeman, MT 59717, USA. E-mail: umbaj{at}montana.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, or GM-CSF. Together, these results may suggest that priming of suspended, circulating neutrophils is associated with a large-scale reorganization of the plasma membrane and associated membrane cortex in a process that is independent of cellular adhesion and gross morphologic polarization.
Key Words: inflammation • leukocyte • granulocyte • degranulation • cortical cytoskeleton
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The ordered exocytosis of granule subpopulations appears to correspond with progressive stages of neutrophil activation [5 ]. The first definable event in neutrophil activation is the transition of neutrophils from nonactivated to primed cellular states [9 ]. Priming is an intermediate state, which markedly enhances the functional response of cells upon secondary stimulation [9 , 10 ]. Neutrophil activation studies indicate that SV exocytosis coincides with priming in vitro, suggesting that SV loss facilitates functional activation [6 ]. Once activated, circulating neutrophils must interact with and firmly adhere to activated endothelial cells. Adherent neutrophils then leave the vasculature and migrate into infected tissues by the process of chemotaxis [11 ]. At the infected site, the exocytosis of primary and secondary granules initiates the microbicidal stage of neutrophil function, which has been investigated in significant detail in vitro, and data from such studies indicate that the neutrophil-primed state may be acquired rapidly upon the exposure of resting cells to proinflammatory products. One limitation, however, is that neutrophil priming is examined in isolated systems of purified agonist and chemically defined buffer solutions, instead of the complex, pathophysiologic microenvironment presented in vivo. Indeed, although in vitro studies imply the relative stability of primed neutrophils in suspension [10 , 12 , 13 ], in vivo studies examining the priming behavior of circulating neutrophils indicate the highly interactive and proadherent nature of this cellular state [14 15 16 17 ]. These studies indicate that priming facilitates conversion of circulating neutrophils to a proadherent state. In the initial stages of this process, primed but preadherent blood neutrophils would necessarily comprise a highly transient subpopulation of circulating neutrophils, which if isolated and examined in vitro, could provide a "snapshot" of the initial subcellular and organizational changes that occur during the beginning stages of in vivo priming.
During neutrophil-priming in vitro, SV exocytosis results in a marked increase in ß2 integrin expression, which contributes to a proadherent state [18 ]. SV are also enriched in flavocytochrome b [19 ], chemoattractant receptors [20 ], and other signaling proteins that may contribute to priming. We have previously compared two preparative methods of blood-neutrophil isolation, which resulted in the retention or selective exocytosis of SV in isolated populations [21 ]. Neutrophil isolation by a dextran-based method resulted in SV retention, and the use of a gelatin-based approach resulted in SV loss. Although the stimulus responsible for SV loss in gelatin neutrophils is unknown, the isolation conditions imposed by this procedure simulate a pathophysiologic microenvironment compositionally similar to what might be responsible for neutrophil activation within inflamed vascular regions in vivo. Because gelatin is made from porcine skin, it likely contains a complex mixture of bacterial products. In the gelatin procedure, concentrated blood cells suspended in diluted plasma are incubated at 37°C in the presence of gelatin. This suspension is thus enriched in leukocytes, platelets, red cells, and various inflammatory mediators. The gelatin method may therefore be useful in acquiring blood neutrophils, which occupy the primed, proadherent state, otherwise highly transient in vivo.
In analogy to neutrophil activation within the inflamed blood microenvironment in vivo, human neutrophils prepared by the gelatin method are exposed to a variety of priming agents while suspended in a plasma-poor but erythrocyte-, platelet-, and leukocyte-enriched mixture. We therefore propose a novel use for gelatin-prepared neutrophils as a model system for examining the initial (preadherent) stages of neutrophil activation in the blood microenvironment. The observed changes in gelatin-neutrophil biochemistry and cellular structure, relative to appropriate control populations, may thus provide a novel insight into mechanisms facilitating their phenotypic conversion from a nonadherent, inactive state to a primed, proadherent state in vivo.
In this study, we characterize the functional and organizational states of the plasma membrane in gelatin neutrophils relative to control neutrophils prepared by a less-activating, dextran-based method. We show that cellular adherence to a neutral plastic substrate and the N-formyl-Met-Leu-Phe (fMLF)-induced respiratory burst is enhanced substantially in gelatin relative to dextran neutrophils. The plasma membrane organization of gelatin and dextran neutrophils was determined by examining the cortical distribution of fodrin, actin, ezrin, and respective anchors CD45 and CD43 following subcellular fractionation, indirect immunofluorescence, and isolation of detergent-insoluble membrane skeletal fractions. Filamentous actin (F-actin), fodrin, and the fodrin anchor CD45 were confined primarily to the cortex or plasma membrane of primed gelatin neutrophils. It is surprising, however, that these proteins were largely cytosolic or intracellular in unprimed dextran neutrophils, possibly indicating an underdeveloped cortex in this population. In density fractionation experiments using gelatin neutrophils, the peak distributions of surface-fodrin, actin, ezrin, and respective anchors CD45 and CD43 were shifted significantly in peak activity relative to the plasma membrane marker alkaline phosphatase (AP) into gradient regions 4–6% heavier in sucrose density. Although CD43 and ezrin were surface-associated in dextran neutrophils, these proteins colocalized with AP in density fractionation experiments. Collectively, these results indicate that gelatin neutrophils occupy a primed, proadherent state and possess an assembled cortical membrane skeleton that is structurally segregated into lateral domains that are differentially enriched for AP and membrane skeletal proteins. In addition, the detailed comparison of gelatin- versus dextran-prepared neutrophils described in this study provides a rigorous context for comparison of numerous previous studies of the function and biochemistry of neutrophils prepared by the gelatin method with those obtained from dextran-prepared cells.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 spectrin and anti-ß2 spectrin (BD Transduction Laboratories, Rockville, MD, USA). Anti-gp91phox mAb 54.1 and the anti-p22phox mAb 44.1 (recognizing epitopes on the intracellular aspect of flavocytochrome b) are produced in this laboratory. The anti-gp91phox mAb 7D5 (recognizing an extracellular epitope of the flavocytochrome b) was the kind gift of Dr. Michio Nakamura (Institute of Tropical Medicine, Nagasaki, Japan); bovine rhodopsin mAb K16 [22
] was used in these studies as an irrelevant control. Secondary HRP- and FITC-conjugated goat antimouse IgG antibodies were purchased from BioRad Laboratories (Hercules, CA, USA) and Bethyl Laboratories, Inc. (Montgomery, TX, USA), respectively. Other antibodies were obtained from the investigators producing them and are so referenced.
Blood-neutrophil isolation
Neutrophils were isolated by a dextran- or gelatin-based methodology, as described previously [21
]. Briefly, isolation by the dextran-based procedure was performed using endotoxin-free reagents and materials following aseptic handling procedures. All solutions and buffers were made using distilled water for injection (WFI) grade (Hyclone, Logan UT, USA). Initial RBC sedimentation was by 1:1 incubation of blood and saline with 3% dextran T-500 at room temperature for 30 min. The leukocyte-enriched supernatant was then centrifuged at 740 g for 10 min, resuspended in 35 ml saline, and underlaid with 10 ml Ficoll-Hypaque medium (<0.01 ng/ml endotoxin) for gradient sedimentation, 20 min at 415 g. The resulting neutrophil-rich pellet was resuspended in ice-cold, distilled WFI-grade for 20 s to facilitate RBC lysis, immediately equilibrated to isotonicity by addition of an equivalent volume of 300 mM NaCl solution, and sedimented as above. Cells were finally resuspended in 25 mlDulbecco’s PBS supplemented with 0.9 mM MgCl2, 0.5 mM CaCl2, 0.1% BSA, and 0.1% dextrose (DPBS4+) and treated with 3 mM di-isopropylfluorophosphate (DFP) for 15 min on ice. Prior to DFP addition, an aliquot of cells was removed for cell count determination by a hemocytometer. Cells were then washed in the same buffer, resuspended in cavitation buffer [0.34 M sucrose, 10 mM HEPES, pH 7.4, 1 mM EDTA, 0.1 mM MgCl2, 1 mM ATP, protease inhibitor cocktail (P8340), and 100 µM PMSF (Calbiochem, La Jolla, CA, USA)], and N2-cavitated at 450 pounds per square inch for 15 min at 4°C. Postnuclear fractions were then isolated by centrifugation at 890 g for 10 min and loaded onto linear sucrose gradients.
For the gelatin-based procedure, 1 unit whole blood was initially centrifuged at 260 g for 20 min at room temperature. The leukocyte- and erythrocyte-enriched pellet fraction was then resuspended in 500 ml per unit blood of a prewarmed (37°C) saline solution supplemented with 2% gelatin, transferred into an autoclaved 500 ml polypropylene funnel, and incubated further for 30 min in a 37°C chamber to facilitate red cell sedimentation. The neutrophil-enriched supernatant was centrifuged at 360 g for 10 min, and pellets were resuspended in 75 ml per unit blood ice-cold ammonium chloride lysis buffer (155 mM NH4Cl, 9.4 mM NaHCO3, pH 7.4, 130 µM EDTA) and incubated for 5 min on ice with occasional mixing. Cells were then washed one to two times, counted, DFP-treated, disrupted by N2 cavitation, and prepared for isopycnic sucrose density sedimentation, exactly as described above.
Isopycnic sucrose gradient sedimentation
Reagents, cell treatment/preparation methods, sucrose gradients, and sedimentation conditions were as described previously [21
].
Functional assays
To evaluate the adherence capacity of neutrophil populations, it was essential to use a substrate that did not induce changes in the activation state of the plated cells upon cellular contact. As neutrophil adherence to plastic has provided a long-standing measure of cellular adherence capacity without stimulating adherence of resting cells [23
24
25
], we used this substrate for the current study in assays performed and quantified as described by Fehr and Jacob [25
]. Briefly, dextran- or gelatin-prepared neutrophils were suspended in autologous, heat-inactivated plasma at 4 x 106 cells/ml and then transferred to Petri dishes at a volume of 1 ml/plate, incubated for 40 min at 37°C. Plates were washed thoroughly afterward in saline, and the proportion of adherent cells (out of total cells plated) was evaluated by myeloperoxidase (MPO) assay.
Superoxide generation was determined by measuring superoxide dismutase (SOD)-inhibitable reduction of cytochrome c in a SpectraMax 250 microtiter plate reader (Molecular Devices, Sunnyvale, CA, USA). Neutrophil suspension (5 µL; 108 cells/ml) was added to microtiter plate wells containing 180 µl assay buffer (47 mM NaH2PO4, 18 mM K2HPO4, 1 mM EGTA, 2 mM NaN3, 200 µM cytochrome c). The stimulus PMA activates neutrophil-NADPH oxidase assembly directly (independent of receptor activation) and was used as a positive control. After 15 min incubation at 37°C, reactions were initiated by addition of stimulus alone (100 nM PMA or 100 nM fMLF) or stimulus with 5 µl bovine liver SOD (12,000 U/ml, Calbiochem). The possibility of adhesion-induced, spontaneous superoxide generation during analysis [26 ] was controlled by monitoring cells in assay buffer without additional stimulus or SOD and evaluated in parallel with test samples but was not observed under the assay conditions used. The rate of superoxide generation was monitored continuously as the absorbance change at wavelength 550 nm for 5 min after stimulus addition (T=0). As the rates of PMA-stimulated superoxide production were quantitatively indistinguishable within error limits in dextran- and gelatin-prepared neutrophils (see Table 1 ), data obtained from both populations were used to calculate the appropriate curve (see Fig. 1B ).
|
View this table: [in a new window] |
Table 1. Rates of Superoxide Generation in Response to fMLF
|
 View larger version (13K): [in a new window] |
Figure 1. Comparison of adherence and fMLF-stimulated superoxide generation in dextran- versus gelatin-prepared neutrophils. (A) Percent of gelatin- and dextran-prepared neutrophils adhering to plastic after settling out of suspension, measured, and quantified as indicated in Materials and Methods. (B) The SOD-inhibitable cytochrome c reduction elicited over the first 5 min after cellular stimulation with 100 nM PMA ( ) or 100 nM fMLF is indicated for gelatin ( )- or dextran ( )-prepared neutrophils. No superoxide generation was detected in the absence of stimulus as indicated (•). As PMA-stimulated cellular activity was quantitatively indistinguishable between dextran- and gelatin-prepared neutrophils, data from both populations are represented in this curve. (A, B) Each dataset represents mean ± SEM from four independent experiments.
|
AP activity was assayed in the presence and absence of 0.1% Triton X-100, as described previously [21 ], to detect total and surface AP, respectively. Latent AP is defined as the difference between total and surface AP and represents that amount of activity that is sequestered within sealed vesicles, such as intact secretory organelles or granules (i.e., intracellular or intravesicular). The latent activity of each fraction was calculated to be the total activity measured in the presence of detergent (or after freeze-thaw treatment) minus surface-accessible activity measured in the absence of detergent (or without freeze-thaw treatment). The sum of all activity in all fractions represents a measure of the total cellular activity. The percent of total activity is defined as the quotient of total activity of the fraction and the total cellular activity times 100 and is plotted to allow normalized comparison of activity distributions in the subcellular fractions.
ELISA detection assays
An ELISA procedure was used to assay CD45 and CD43 activities. Assays were performed according to the method described [21
] using the same controls. Latent activity (as defined and calculated above) was also measured for these activities. CD45 and CD43 mAb failed to bind antigen when samples were solubilized in Triton X-100. (Conversely, both mAb retained their binding capacity following saponin treatment in immunofluorescence assays.) Triton X-100 was therefore substituted with a freeze-thaw method of membrane disruption [27
]. The efficacy of membrane disruption by freeze-thaw was evaluated by quantitating latent AP activity in gradient fractions from the same dextran-prepared population detected using freeze-thaw versus detergent methods of membrane lysis and found 97 ± 3% efficient for n = 3 (data not shown). No latent activity was detected for CD43 in dextran- or gelatin-prepared neutrophils.
Indirect immunofluorescence
Freshly isolated neutrophils were treated with 3 mM DFP in DPBS, pH 7.4, for 15 min at 4°C. Cells were washed in the same buffer and fixed with 3% paraformaldehyde (in PBS, pH 7.4) for 15 min at 25°C and then washed several times in the same buffer. Sedimentation was performed as in flow cytometry experiments (see below). Fixed cells were then resuspended with PBS, pH 7.4, and 0.2% gelatin, with or without 0.01% saponin (blocking buffer), and incubated overnight at 4°C. Incubations were performed in 100 µl vol (106 neutrophils). Primary labeling was with 50 nM FITC-phalloidin, 8 ug fodrin mAb (specific for the ß2 chain), 6 ug control antibody (surface control mAb 7D5, cytoplasmic control mAb 54.1, irrelevant mAb K16), or 3 ug ezrin mAb in blocking buffer. Labeling with mAb 7D5 was performed on intact, nonpermeabilized neutrophils for 30 min at 37°C. All other labeling experiments used permeabilized neutrophils. After primary labeling, reactions were quenched with ice-cold blocking buffer, with or without detergent, as indicated, followed by two additional washings. Secondary labeling was with goat antimouse antibody anti-IgG FITC (diluted 1/400 in blocking buffer±detergent as indicated above) performed as above. Approximately 4000 cells were then distributed in 100 µl aliquots onto glass slides fitted within a Cytospin 2 centrifuge. Cells were then spun onto slides (500 g, 5 min) and viewed with Fluoromount-G (Southern Biotec, Birmingham, AL, USA) on a Zeiss Axioskop 2 Plus microscope equipped with a CCD camera to allow on-screen imaging via AxioVision Rel. 4.4 software.
Flow cytometry analysis
The proportion of cellular CD45 localized to the cell surface of neutrophils was examined using two different approaches. In the initial approach, surface membrane CD45 expression was examined before and after degranulation, and degranulated neutrophils were prepared by suspending 108 cells per ml in DPBS4+ supplemented with 80 µg catalase, 10,000 activity units SOD, and 2 µg/ml cytochalasin B in a 50-ml conical tube. Cytochalasin B (which prevents F-actin polymerization) disassembles the cytoskeleton to allow for the complete surface up-regulation of granule populations during fMLF stimulation. After 7 min (37°C), 1 µM fMLF was added, and incubation continued for an additional 3 min. Cellular activation was terminated by addition of ice-cold DPBS4+ with 2 mM EDTA, 100 µM PMSF, and cocktail (P8340), followed by sedimentation for 10 min at 740 g. Cells were then washed in DPBS4+. Pellets were resuspended in DPBS4+ with 3% goat serum at 107 cells per ml, and cells were distributed in 100 µl aliquots for immunolabeling. Buffer was then addedcontaining no primary antibody (control) or 2 µg each mAb for CD45, mAb K16 (specific for bovine rhodopsin), or mAb 44.1 (specific for the intracellular aspect of the cytochrome b558 component p22phox, used to verify the intactness of plasma membranes), followed by incubation cells for 30 min on ice. After several washes, a 1:200 dilution of secondary FITC-conjugated, goat antimouse IgG antibody was added followed by incubation as described above. After one to two washes, cells were resuspended in 500 µl PBS, pH 7.4, and examined by flow cytometry as described previously [28
]. All negative controls were quantitatively similar, so only data from mAb K16 are included (see Fig. 2A
and 2B
).
 View larger version (24K): [in a new window] |
Figure 2. Analysis of stimulus-induced CD45 redistribution in gelatin-primed neutrophils by flow cytometry. (A) Flow cytometry histogram (upper) and bar graph (lower) showing CD45 expression on the surface of gelatin-primed neutrophils, which were maintained on ice (Gelatin) or fully degranulated by stimulation with 1 µM fMLF in the presence of 2 µg/ml cytochalasin B (Gelatin + CB + fMLF). (B) Flow cytometry histogram (upper) and bar graph (lower) showing CD45 expression on dextran-prepared neutrophils maintained on ice (Dextran, ice) or incubated at 37°C in isotonic NaCl solution in the presence (Dextran + gelatin) or absence (Dextran – gelatin) of 2% gelatin. (A, B) The bar graphs below the flow cytometry histograms show the mean fluorescence intensity of cell-surface CD45, calculated as percent of maximum intensity obtained from fully degranulated populations (+CB/fMLF) and arbitrarily set at 100%. The average ± SEM of the data shown in each bar graph is averaged over three independent experiments. mAb K16 is specific for rhodopsin and included in experiments as irrelevant control (K16). Controls performed with intracellular staining, mAb 44.1, irrelevant antibody, mAb K16 (followed by secondary antibody), or secondary alone yielded similar curves.
|
Fig. 4B
).
 View larger version (19K): [in a new window] |
Figure 3. CD45 and AP distribution in sucrose density gradients made from gelatin-primed or dextran-unprimed neutrophils. Unprimed or primed neutrophils were prepared and density-fractionated as described in Materials and Methods. Percent of the total activity is plotted as a function of sucrose percentage (wt/wt) in gradient fractions. (A) CD45 and AP activities in gradient fractions prepared from dextran-unprimed neutrophils were measured before and after membrane disruption by detergent solubilization (AP) or freeze-thaw treatment (CD45). The activity measured before membrane disruption [Surface AP ( ); Surface CD45 ( )] was subtracted from the activity measured after membrane disruption (total activity, not shown), and the difference is plotted as the latent activity for each protein [Latent AP (•); Latent CD45 ()]. (B) CD45 and AP activities were measured in fractions prepared from gelatin-primed cells before and after membrane disruption and were shown to be quantitatively indistinguishable. Thus, latent activity was 0, and only the Surface AP (•) and Surface CD45 () in gradient fractions are plotted. Recovery of AP and CD45 activities from density gradients was greater than 90%. Error bars indicate SEM from analysis of three gradients each from dextran-unprimed or gelatin-primed neutrophils.
|
 View larger version (41K): [in a new window] |
Figure 4. Sucrose density gradient sedimentation patterns of cortical components and transmembrane-anchoring proteins in unprimed or primed neutrophils, which were prepared and density-fractionated as described in Materials and Methods. Data are plotted as described in Figure 3
. The subcellular fractionation pattern of each protein was analyzed in parallel within unprimed () and gelatin-primed () neutrophil populations by quantitative immunoblotting (A–C) or ELISA (D) for n = 3 experiments per neutrophil population. Representative immunoblots are shown below each figure (A–C). The distributions of (A) the fodrin ß-chain, (B) G-actin, (C) ezrin, and (D) CD43 are plotted. Surface AP (broken line) is shown to identify the distribution of the AP-enriched plasma membrane. Error bars indicate SEM. Recovery of activities from density gradients was  96% of the activity present in supernatants prior to fractionation.
|
for 20 min, or 12 ng/ml GM-CSF for 20 min. Control dextran-prepared neutrophils were maintained at 4°C for 30 min without stimulation. Neutrophils were then analyzed by isopycnic sucrose density fractionation and indirect immunofluorescence microscopy as described above.
SDS-PAGE and immunoblot analysis
SDS-PAGE and immunoblot analysis were conducted as described previously [21
]. mAb used with dilutions in parentheses included actin (1:14,000), ezrin (1:3000), or the 2 or ß2 chain of nonerythrocyte spectrin (also referred to as fodrin; 1:5000). Regarding fodrin staining, mAb specific for the 2 or ß2 chain yielded identical staining patterns, so only data from one mAb (anti-ß2 chain) are presented in this study. The primary and secondary antibody concentrations used were selected to yield optical density readings in the linear range of detection for each antibody.
Detergent solubility analysis
The Triton X-100 detergent solubility properties of the actin-based cytoskeleton of each neutrophil population were evaluated according to the method of Watts and Howard [29
]. Neutrophils isolated by the dextran- or gelatin-based procedures were suspended at 2 x 106 cells/ml, solubilized in Triton X-100 for 15 min at 25°C as described previously [29
], and processed further to isolate the different fractions of the actin cytoskeleton. Briefly, a Triton-insoluble, low-speed pellet fraction (representing the membrane skeleton) was isolated initially by sedimentation of cell lysates at 15,900 g for 2 min. Low-speed supernatants were then removed and further sedimented at 366,000 g for 5 min using a TL 100 ultracentrifuge (Beckman, Palo Alto, CA, USA) to isolate Triton-soluble F-actin from monomeric (G) actin. These latter two subpopulations are referred to as the detergent-soluble, high-speed (HS) pellet and HS supernatant fractions, respectively, and contribute to the cellular (noncortical) actin cytoskeleton. Each of the three cytoskeletal fractions was then examined for actin, fodrin, and ezrin content by immunoblot analysis as described above.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Significant differences were also observed when both populations were examined for the formyl peptide-induced respiratory burst, as measured by the SOD-inhibitable reduction of cytochrome c. Figure 1B
shows that both populations exhibited transient respiratory bursts lasting less that 5 min when stimulated with 100 nM fMLF. The rate of superoxide generation induced by 0.1 uM (PMA), the positive maximal stimulus control, was equivalent in gelatin- and dextran-prepared populations, exhibiting no decline in rate during the period of measurement and indicating that the maximal respiratory burst potential of both cell types was the same. Thus PMA-stimulated superoxide generation is displayed in Figure 1B
as the averaged activity from both populations. As shown in Table 1
, however, examination of initial rates induced by stimulation with fMLF differed fivefold, and gelatin neutrophils were able to produce superoxide during the first minute of stimulation as well the maximal positive PMA control [
3.0 nmol/106 cells/min in response to fMLF (n=4) or PMA (n=4)]. Dextran neutrophils exhibited a markedly weaker response to fMLF of 0.6 nmol/106 cells/min (n=4; Table 1
) or about one-sixth the maximal rate induced by PMA. Without exception, we observed no superoxide production in the absence of agonist addition in either neutrophil population (n=8). Collectively, these data indicate that gelatin- and dextran-prepared neutrophils occupy primed and unprimed cellular states, respectively, and from this point forward, we refer to each neutrophil population as "primed" (gelatin) or "unprimed" (dextran).
CD45
CD45 is an integral membrane protein found in leukocytes, which functions to anchor fodrin, the nonerythroid spectrin homologue, to the plasma membrane in an analogous manner to the Band 3/ankyrin complex, anchoring spectrin in the erythrocyte membrane [33
]. As CD45 has a complex distribution in unprimed neutrophils, which is largely intracellular [34
], we first verified the extent of its surface localization in primed and unprimed neutrophils by flow cytometry. The amount of surface CD45 in unprimed neutrophils was determined by comparing intact primed or unprimed neutrophils with the cells from each population that were fully degranulated by fMLF stimulation in the presence of cytochalasin B, an inhibitor of F-actin polymerization. Signal intensity of surface CD45 fluorescence inprimed and fully degranulated cells corresponded precisely, suggesting maximal surface expression of CD45 in primed neutrophils (n=3; Fig. 2A
). By contrast, unprimed neutrophils had a relatively low surface CD45 expression of 34 ± 5% (n=3; Fig. 2B
).
In this same set of experiments, we determined if gelatin exposure could induce redistribution of intracellular CD45 to the surface of unprimed neutrophils. Dextran-prepared, unprimed neutrophils were maintained on ice or incubated 30 min at 37°C in saline with or without 2% gelatin and then examined for surface CD45 expression by flow cytometry. As indicated in Figure 2B , 80 ± 3% and 32 ± 6% of cellular CD45 was detected on the cell surface of neutrophil populations incubated in the presence and absence of 2% gelatin, respectively. These data suggested that the conditions imposed by the gelatin procedure are sufficient to elicit a nearly maximal surface expression of CD45. The partial up-regulation of CD45 observed after 37°C incubation in saline without 2% gelatin is likely related to a temperature-dependent, surface up-regulation of SV described in previous studies [35 ].
The differences in the cellular distribution of CD45 in primed and unprimed neutrophils detected by flow cytometry analysis suggested marked differences the subcellular localization of CD45 activity of each neutrophil population. To characterize this difference further, primed and unprimed populations of neutrophils were disrupted by N2 cavitation, and postnuclear supernatants were fractionated by isopycnic sucrose density sedimentation [21
]. In these experiments, we found, in analogy to the flow cytometry results, that the distributions of the plasma membrane marker activity AP and CD45 were primarily intracellular in unprimed neutrophils but completely surface-associated in primed neutrophils. To determine the subcellular distribution of each protein in density gradients made from unprimed neutrophils, the latent versus plasma membrane-bound activities were determined. The latent activities shown in Figure 3A
, representing 70% of the total activity of each protein, were determined as the activity difference obtained when assayed in the presence versus the absence of detergent in each gradient fraction. As epitope binding by the CD45 mAb used in this study is inhibited severely in the presence of Triton X-100, the latent distribution of CD45 was instead determined by ELISA assay, before and after freeze-thaw treatment of gradient fractions. Figure 3A
compares the distribution of surface and latent AP to surface and latent CD45 in fractions obtained from unprimed cells. It is interesting that the latent AP activity shown in Figure 3A
has been shown previously as a marker for SV [4
], which is up-regulated rapidly to the cell surface of neutrophils upon cell stimulation [6
]. The plasma membrane and latent distributions of AP are resolved clearly with symmetrical distributions, having peaks at 29.4% and 33.6% sucrose, respectively. Figure 3A
also shows that the general distributions and peak activities of surface and latent CD45 correspond closely with those of the AP activity, suggesting that CD45 is similarly distributed in the subcellular fractions.
The same measurements were made following sucrose density fractionation of primed neutrophils, with no latent CD45 or AP activity detected (i.e., activities before and after membrane disruption were the same). The absence of latent AP activity in primed neutrophils is in agreement with several previous studies [2
, 6
]. The subcellular distributions shown in Figure 3B
thus indicate the complete surface accessibility of AP and CD45 in primed neutrophils. It is interesting, however, that the surface distributions of AP and CD45 in Figure 3B
vary with respect to each other. Whereas a peak density of 29% was observed for surface AP activity in primed and unprimed neutrophils (compare Fig. 3A
and 3B
), the surface CD45 activity in both populations differs markedly, with peak densities 34% and 29% sucrose, respectively. It is interesting that the magnitude of the shift in CD45 distribution observed in primed, relative to unprimed, neutrophils appears equivalent to that observed between surface and latent activities in the unprimed cells (see Fig. 3A
). The results depicted in Figure 3B
suggest that although AP and CD45 activities are derived from surface membranes, the populations of vesicles carrying these activities have different densities and that the isolated plasma membrane of primed cells is different than that of unprimed cells. For all density gradients examined, lactoferrin (a marker for specific granules) and MPO (a marker for primary granules) activities peaked at 42% sucrose and 48% sucrose, respectively, and neither activity was detected below 39% sucrose. We conclude that in unprimed neutrophils, CD45 and AP mirror one another in subcellular distribution, whereas in primed neutrophils, they reside in plasma membrane vesicles of different density.
Modulation of the fodrin and actin cortical skeleton
Fodrin, actin, ezrin, CD45, and CD43 play an essential role in organizing the cell surface of human leukocytes and in the assembly and function of signaling complexes. Any significant difference in the organization of these proteins between primed and unprimed cellular states could therefore suggest a contribution to the regulation neutrophil function. The activity profiles for each protein were examined in density gradients prepared from primed and unprimed neutrophils and presented in Figure 4A
4B
4C
4D
. Fodrin and actin of primed and unprimed neutrophils were found predominantly at or near the top of the sucrose gradients at 10–18% sucrose density, indicating a cytosolic localization (Fig. 4A
and 4B)
. By contrast, the distribution of fodrin and actin in gradient fractions from primed neutrophils appeared to codistribute within cytosolic and plasma membrane-enriched regions of density gradients. The amount of actin codistributing with the plasma membrane of each population differed by approximately twofold and was 20 ± 3% and 35 ± 4% of total cellular actin for unprimed and primed neutrophils, respectively (Table 2
). A more dramatic shift was observed for fodrin, as the relative amount cosedimenting with the plasma membrane was calculated as 7 ± 4% and 70 ± 6% for unprimed and primed neutrophils, respectively (Table 2)
. It is surprising that the fraction of fodrin and actin activity measured within the plasma membrane zone of primed neutrophils (indicated by traced surface AP activity) peaked at 33–36% sucrose, a 4–6% shift in sucrose density relative to surface AP and similar to the density shift observed for surface CD45 in primed neutrophils (Fig. 3B)
. Further examination of the actin-binding protein ezrin and the ezrin-anchor protein CD43 in the same density gradients revealed a similar shift in peak density in primed neutrophils, which was not observed in unprimed neutrophils (Fig. 4C
and 4D)
.
|
View this table: [in a new window] |
Table 2. Percent of Total Gradient Content of Actin and Fodrin in Plasma Membrane-Enriched Fractions of Sucrose Density Gradients from Dextran-Unprimed and Gelatin-Primed Neutrophils
|
 View larger version (53K): [in a new window] |
Figure 5. Subcellular localization of transmembrane and cortical proteins in unprimed and primed neutrophils by indirect immunofluorescence microscopy. Blood neutrophils were isolated by the dextran (Unprimed, left two columns) or gelatin (Primed, right two columns) methods, fixed, permeabilized with saponin (A, B), and then labeled with mAb specific for the indicated proteins as described in Materials and Methods. Each row, representing labeling of the indicated proteins, shows corresponding DIC (first and third columns) and fluorescence (second and fourth columns) images. (A) Representative micrographs show CD45, fodrin, and actin to be primarily intracellular or cytosolic in unprimed neutrophils but predominantly surface- and cortex-localized in primed cells. Ezrin and CD43 are cortex- and surface-localized, respectively, in both neutrophil populations. Cytosolic (or intracellular) and surface staining are highlighted by nuclear exclusion and ringed fluorescence patterns, respectively. (B) Representative control stains performed with the irrelevant mAb K16. (C) Unprimed, permeabilized neutrophils (left two images), labeled with mAb 54.1 (recognizing an intracellular epitope of flavocytochrome b), are shown as a positive control for cytoplasmic staining. Primed but unpermeabilized neutrophils (right two images), labeled with mAb 7D5 (recognizing an extracellular epitope of flavocytochrome b), are shown as a positive controls for surface staining. (A) Labeling experiments for each protein shown included the indicated control for a minimum of n = 11 experiments per protein per population. Original bars equal 5 µm.
|
|
View this table: [in a new window] |
Table 3. Analysis of Cytoskeletal Composition of In Vitro-Isolated Membrane Skeletons from Dextran-Unprimed and Gelatin-Primed Neutrophils
|
, or 12 ng/ml GM-CSF as described in Materials and Methods and examined relative to unstimulated populations maintained at 4°C. Figure 6
shows the subcellular distributions of CD45, fodrin, actin, ezrin, and CD43 in dextran-prepared, stimulated neutrophils. Each profile is the average of three individual distributions derived from each of the three stimulated conditions and is plotted along with the corresponding distributions from dextran-prepared, unstimulated cells (n=3). An examination of these average distributions shows that the peak positions, relative magnitudes, and variabilities (i.e., error bars) in the distributions are similar to those shown in Figures 3
and 4
for the gelatin-prepared cells. These results show that the variation among the distributions resulting from each of the stimulation conditions is no greater than the variation in distributions observed among individual gelatin preparations.
 View larger version (46K): [in a new window] |
Figure 6. Subcellular redistribution of transmembrane and cortical proteins of neutrophils by defined priming stimuli. (Left column) Dextran-prepared neutrophils were maintained on ice (unstimulated,  ) or exposed to LPS, TNF-, or GM-CSF (stimulated, •), as described in Materials and Methods, and then disrupted, fractionated, assayed, and plotted as described in Figures 3
and 4
. The relative activities in each fraction were averaged for the three stimuli and for the control and are plotted with SEM computed from the averages. The surface AP distributions () were invariant. (Right columns) Dextran-prepared neutrophils were exposed to 50 ng/ml LPS followed by fixation, permeabilization, and staining for immunofluorescence as described in Materials and Methods. The left set of micrographs represent unstimulated neutrophils, and the right set of micrographs represent LPS-stimulated cells. The primary antibodies used correspond to those used in the subcellular fractionation analysis and are identified by the corresponding labels on the right. Negative controls, using irrelevant isotype-matched antibodies, showed no detectable immunofluorescence and were indistinguishable from those shown in Figure 5B
. Positive controls were indistinguishable from those shown in Figure 5C
. Original bar equals 5 µm.
|
and GM-CSF (not shown). Together, these observations suggest that a redistribution of transmembrane and cortical proteins occurs after exposure to the priming stimuli and that the changes in membrane organization of the gelatin-prepared neutrophils described in this study may represent those generally induced by priming stimuli. |
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, or GM-CSF, a reorganization of transmembrane and cortical proteins similar to that observed for gelatin neutrophils was observed (Figs. 4
5
6)
. These results provide first evidence of large-scale compositional and organizational changes occurring in the neutrophil cortex during the initial stages of phenotypic conversion, in vivo. These changes involve the structural segregation of primed plasma membranes into AP-enriched and AP-depleted surface compartments as suggested in Figure 7
.
 View larger version (17K): [in a new window] |
Figure 7. Model of plasma membrane structure before (left) and after (right) priming. Prior to neutrophil priming, the plasma membrane-bound proteins, AP (a GPI-linked protein), CD43, and the peripheral CD43-associated ezrin protein, are distributed uniformly on the cell surface. Furthermore, predominantly intracellular distributions of F-actin and fodrin contribute to a minimal, underdeveloped cortical structure in unprimed neutrophils. Following neutrophil priming, cellular fodrin moves from an intracellular distribution to the cytoplasmic face of the plasma membrane. Cortical integration of cellular fodrin is accompanied by the translocation of intracellular CD45 to the cell surface. The resulting approximately threefold increase of CD45 expression in plasma membranes coincides with the dynamic reorganization of the plasma membrane structure in primed neutrophils. Two major domains can be distinguished clearly in postnuclear supernatants following equilibrium sedimentation on sucrose density gradients. These domains include an AP-enriched CD43/CD45/fodrin-poor domain at 29% sucrose and an AP-poor, CD45/fodrin-enriched domain at 34% sucrose. The latter domain is also enriched in F-actin, CD43, and ezrin, assembled into a functional complex in leukocyte plasma membranes. This structure is analogous to the glycophorin C-protein 4.1-F-actin complexes, which anchor erythrocyte plasma membranes to the cortical skeleton.
|
Surface AP was useful as a reference for determining the general distribution of plasma membrane in density gradients, as its activity profile in gradients made from the primed and unprimed neutrophils examined was indistinguishable. Although AP is linked to the outer membrane by a GPI anchor and therefore, considered a common component of Triton X-100-insoluble membrane rafts [42 ], the detergent-free sedimentation studies presented here are not designed to address raft organization. Instead, in this study, neutrophils were analyzed directly after blood isolation, with minimal in vitro manipulation of cells or cellular material to determine the extent of cortical change that occurs during a defined but transient stage of their activation in vivo. As such, we speculate that the changes in cortical organization reported in this study may reflect physiologic changes induced during the priming of circulating blood neutrophils in vivo.
The functional and membrane-organizational changes reported here in primed neutrophils are associated with changes in cell morphology. Although gelatin-prepared neutrophils do not undergo a morphological polarization, in terms of pseudopod development, they do acquire a uniformly ruffled, rounded morphology [36 ]. This contrasts with the microvilli, which decorate the surface of quiescent neutrophils in vivo [43 ]. The morphological changes that occur in gelatin neutrophils during priming are consistent with and are in fact suggestive of the large-scale changes in membrane cortical organization observed in this study. Our approach to examining submembrane architecture was based on mapping the subcellular distribution of components contributing to the two major anchoring complexes that exist in hematopoietic cell plasma membranes, including the CD43-ezrin-F-actin [44 , 45 ] complex and CD45-fodrin complex [33 ]. CD43 is the major transmembrane sialoglycoprotein species in leukocytes and analogous to erythrocyte glycophorin C. Both proteins anchor to F-actin through Protein 4.l-like linkages (Band 4.1 in erythrocytes, ezrin in leukocytes). Similarly, leukocyte CD45 is linked functionally and structurally with fodrin and analogous to the Band 3/ankyrin complex, which anchors erythrocyte spectrin. The cosedimentation of actin and fodrin with their respective anchoring complexes reported here in primed neutrophils is thus consistent with a physiologic pattern of cortical reorganization in this population.
In neutrophils, the lateral segregation of cortical actin and fodrin was first described in studies investigating regulation of the G-protein-coupled formyl peptide receptor (FPR) [46 47 48 49 50 51 52 ]. Such studies have elucidated the critical role of actin- and fodrin-enriched surface domains in mediating FPR desensitization in gelatin-primed, fMLF-stimulated human neutrophils. In this regard, the significant structural differences in the cortical organization of the primed versus nonprimed neutrophil populations observed in this study may be of particular importance to understanding the role of surface plasma membrane microdomains in G-protein-coupled receptor function.
The organizational patterns of fodrin and CD45 in primed and unprimed neutrophils reported in this study are analogous to a previous report in T cells showing the relative depletionand enhancement of these proteins in the surface structure of resting and activated T cells, respectively [53 ]. In that study, surface association of T cell fodrin and CD45 (induced by cross-linking surface receptor) also contributed to the lateral segregation of membrane structure. The activation-induced redistribution of cortical fodrin and fodrin-anchoring proteins from cytosol to cortex in functionally diverse leukocyte populations thus indicates the importance of cortical change in modulating cell function and the primary role of fodrin in facilitating such changes.
In summary, results from our study suggest that unprimed, nonadherent blood-neutrophil populations possess an underdeveloped/minimal cortex and that functional conversion to a primed cellular state is associated with the maturation of cortical structure as well as a lateral reorganization of cortical components and anchorage sites relative to surface AP. This structurally reorganized primed plasma membrane occurs in the absence of substrate adhesion, phagocytosis, and directed motility and thus highlights the intrinsic potential of neutrophils to undergo sustained, substantial changes in cortical remodeling in the absence of the classical polarizing stimuli.
Received August 10, 2006; accepted November 17, 2006.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
m ß 2) in human neutrophils J. Clin. Invest. 92,1467-1476[Medline]-actinin, and actin binding protein 280 in stabilizing Triton insoluble F-actin in basal and chemotactic factor activated neutrophils Cell Motil. Cytoskeleton 28,155-164[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  S. M. Uriarte, K. R. McLeish, and R. A. Ward Anti-proteinase 3 antibodies both stimulate and prime human neutrophils Nephrol. Dial. Transplant., April 1, 2009; 24(4): 1150 - 1157. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Starosta, K. Pazdrak, I. Boldogh, T. Svider, and A. Kurosky Lipoxin A4 Counterregulates GM-CSF Signaling in Eosinophilic Granulocytes J. Immunol., December 15, 2008; 181(12): 8688 - 8699. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Riesselman, H. M. Miettinen, J. M. Gripentrog, C. I. Lord, B. Mumey, E. A. Dratz, J. Stie, R. M. Taylor, and A. J. Jesaitis C-Terminal Tail Phosphorylation of N-Formyl Peptide Receptor: Differential Recognition of Two Neutrophil Chemoattractant Receptors by Monoclonal Antibodies NFPR1 and NFPR2 J. Immunol., August 15, 2007; 179(4): 2520 - 2531. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
