Published online before print November 1, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Division of Hematology, Department of Medicine, University Medical Center Groningen, Groningen, the Netherlands;
Laboratory of Cell Biology and Electron Microscopy, University of Groningen, Groningen, the Netherlands;
Department of Immunology, University Medical Center Utrecht, Utrecht, the Netherlands; and
Sanquin Blood Bank North East Netherlands, Groningen, the Netherlands
1 Correspondence: Division of Hematology, Department of Medicine, University Medical Center Groningen, Hanzeplein 1, 9713 GZ Groningen, the Netherlands. E-mail: e.vellenga{at}int.umcg.nl
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: MDS • signal transduction • NADPH
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The myelodysplastic syndromes (MDS) are a group of clonal hematological disorders, which are characterized by an impaired differentiation of the hematopoietic stem cell compartment, affecting the erythroid, megakaryocytic, and myeloid lineages [4 , 5 ]. With regard to the latter, it is known that the impaired hematopoiesis might not only result in peripheral neutropenia but that the remaining neutrophils do not function properly, resulting in a high incidence of recurrent bacterial infections [6 ]. For instance, we and others have shown previously that GM-CSF priming of the fMLP-mediated ROS production was impaired severely in most MDS patients, potentially contributing to the increased susceptibility to infections [7 8 9 ].
The molecular mechanisms underlying the process of priming have so far remained elusive. ROS production is mediated by the NADPH oxidase, which consists of membrane-bound flavocytochrome b558 and cytosolic factors p47phox, p40phox, p67phox, and Rac [10 ]. Upon stimulation of neutrophils, the cytosolic components are activated and translocate toward the plasma membrane, where subsequent assembly of the NADPH oxidase results in ROS production. Recent studies have suggested that NADPH oxidase activity is dependent on the presence of lipid rafts [11 , 12 ], which are cholesterol-enriched, detergent-resistant membrane microdomains within the plasma membrane [13 ].
Several signal transduction pathways have been implicated in NADPH oxidase activation. For example, lipid products of PI-3K have been shown to associate with the cytosolic oxidase proteins p47phox and p40phox [14 ]. In addition, ERK1/2, a key member of the MAPK signaling route, is involved in neutrophil NADPH oxidase activation [15 16 17 18 ]. As for ROS production, the activation of ERK1/2 and PI-3K by fMLP can be primed by pretreatment of neutrophils with GM-CSF [7 , 19 , 20 ]. Recent evidence suggests that these signal transduction pathways are also dependent on the presence of lipid rafts. ERK1/2 activation was inhibited by disruption of lipid rafts [21 , 22 ], and PI-3K activation, localization, and signal transduction were dependent on lipid raft integrity in several cell lines [23 24 25 ]. It was reported recently that in contrast to ERK1/2 activation, phosphorylation of STAT5 by the growth hormone (GH) was not affected by lipid raft disruption [26 ]. We have shown previously that whereas activation of the ERK1/2 and PI-3K-protein kinase B (PKB) pathways was impaired in MDS neutrophils, activation of STAT5 was normal [7 , 27 ]. In combination with the impaired ROS production observed in GM-CSF-primed neutrophils from these patients, these results suggest a possible role for lipid rafts in the defective neutrophil functioning in MDS. In this study, we show that disruption of lipid rafts results in abrogation of the neutrophil respiratory burst and a reduced activation of ERK1/2 and PKB. We demonstrate further that the expression of the lipid raft marker GM1 increases upon the plasma membrane of healthy neutrophils through de novo synthesis and that lower GM1 expression on MDS neutrophils correlates with a reduced ROS production in these cells.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Patients
MDS was classified as refractory anemia (RA) or RA with ringed sideroblasts, according to French American British Cooperative Group criteria [28
]. Informed consent was obtained from all patients. The Human Subject Review Board of the University Medical Center Groningen (the Netherlands) approved the protocol.
Isolation of neutrophils
Peripheral blood, anticoagulated with 0.32% sodium citrate, was obtained from healthy volunteers and MDS patients. Neutrophils were isolated as described previously [29
]. In short, mononuclear cells were removed by centrifugation over Fycoll-Paque (Amersham, Uppsala, Sweden), and erythrocytes were lysed with ice-cold NH4Cl solution. Neutrophils were allowed to recover for 30 min at 37°C in RPMI 1640 (BioWhittaker, Verviers, Belgium) supplemented with 0.5% human serum albumin (HSA; CLB). Before stimulation, cells were resuspended in incubation buffer (20 mM HEPES, 132 mM NaCl, 6 mM KCl, 1 mM MgSO4, 1.2 mM KH2PO4, 5 mM glucose, 1 mM CaCl2, 0.5% HSA). In all cases, the cell population isolated consisted of >95% neutrophils as determined by May-Grünwald Giemsa staining.
Cholesterol measurement
Neutrophils were treated with CD for 30 min and washed twice with ice-cold PBS. Lipids were extracted from resuspended neutrophils as described by Bligh and Dyer [30
]. Free cholesterol levels in the cells were determined fluorimetrically by convergence of cholesterol in the samples by chol-oxidase (Boehringer, Alkmaar, the Netherlands) to cholestone and H2O2 and subsequent convergence of parahydroxyphenylacetic acid (Sigma Chemical Co.), H2O2 (Boehringer), and cholate to a measurable fluorimetric complex [31
].
Respiratory burst measurement
Production of ROS was measured as described previously [7
]. Briefly, neutrophils (2.5x106 cells/ml) were incubated with DHR123 for 15 min and stimulated with 1 µM fMLP for 30 min. For priming experiments, cells were pretreated with 5 ng/ml GM-CSF for 15 min prior to fMLP stimulation, and stimulation of the neutrophils with fMLP was terminated by washing the cells with ice-cold PBS containing 1% HSA and placing them on ice. Oxidation of DHR123 (Molecular Probes) to the fluorescent Rhodamine 123 was measured by FACS analysis within 30 min.
Lipid raft preparation
Lipid rafts were isolated as described previously [32
]. In short, neutrophils were washed twice with PBS and once with 25 mM Tris, pH 7, 150 mM EDTA, 1 mM DTT, 150 mM NaCl, and one tablet protease inhibitor Complete (TNE). Cells were lysed in 300 µl TNE containing 0.75% Triton X-100 and sonicated three times for 10 s. After incubation on ice for 30 min, samples were mixed with 600 µl 60% Optiprep (Sigma Chemical Co.) and transferred to SW55 ultracentrifuge tubes (Beckman, Munchen, Germany). Fractions of 600 µl Optiprep (Sigma Chemical Co.; 35%, 30%, 25%, 20%, and 0% in TNE) were layered on top. After 18 h of ultracentrifugation at 40,000 g, six fractions were removed from the top of the gradient. Fractions were numbered 1–6 from bottom to top. Lipid rafts were present in Fractions 3–5.
Western blotting
After boiling samples for 10 min, proteins were separated on 10% SDS-PAGE and transferred electrophoretically to nitrocellulose membranes (Protran, Schleicher & Schuell, Dassel, Germany). Membranes were probed with antibodies according to the manufacturer’s protocols. Proteins were detected by ECL. Quantification of phosphorylation levels was performed by densitometry of the films, using ImageMaster1D Elite (Pharmacia, Woerden, the Netherlands).
GM1 measurement
GM1 expression was measured essentially as described by Sheriff et al. [33
], with some alterations. In short, cells were stimulated with 1 µM fMLP, with or without prior pretreatment with 5 ng/mL GM-CSF, and stimulation was stopped by adding ice-cold PBS centrifugation. Cells were resuspended in PBS containing 5 µg/mL cholera toxin subunit B-Alexa Fluor 555 and stained for 30 min at 4°C. After staining, cells were washed twice, and FACS analysis was performed within 30 min.
Electron microscopy
Neutrophils were processed for immunolabeling according to standard procedures [34
]. In brief, cells were fixed in 4% paraformaldehyde plus 0.5% glutaraldehyde in 0.1 M PBS for 30 min after stimulation and subsequently rinsed with PBS. Cells were imbedded in 5% gelatine, infiltrated with 2.3 M sucrose solution, and placed on copper holders for utrathin cryosectioning, where cryosections were stained with 25 µg/ml cholera toxin subunit B-biotin for 1 h, which was visualized with 5 nm gold particles conjugated to streptavidin. A minimum of two cryosections was examined for every samples with a CM100 electron microscope (Philips, Eindhoven, the Netherlands). A total of six cells was studied, and three pictures were taken of every cell investigated. Cytoplasmic area per cell was measured using AnalySIS (SIS, Munster, Germany) and divided by the number of gold particles per cell to obtain a quantitative measure of gold labeling.
Statistical analysis
The effect of CD or Nystatin treatment on cholesterol content or ROS production in healthy controls was calculated using the Wilcoxon signed ranks test for paired samples, as were the differences in ROS production and GM1 expression between MDS patients and healthy donors. Data were expressed as mean ± SEM. P values 
0.05 were considered significant.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Next, we incubated neutrophils with CD for 30 min prior to measuring the respiratory burst by FACS analysis. As shown in Figure 1A (left panel), stimulation of neutrophils with 1 µM fMLP resulted in production of ROS, which could be abrogated completely by CD pretreatment [mean fluorescence intensity (MFI) of 16±2 vs. 4±0.3, t=20 min, n=6, P=0.03]. To ascertain that CD was not toxic to neutrophils, cells were washed after CD pretreatment and allowed to recover at 37°C. Subsequent stimulation of neutrophils with fMLP resulted in restoration of ROS production.
 View larger version (24K): [in a new window] |
Figure 1. Effect of lipid raft disruption on ROS production and signal transduction pathways. (A) Unprimed neutrophils and neutrophils that were primed with 5 ng/ml GM-CSF for 15 min were stimulated with 1 µM fMLP for the indicated times. (B) Neutrophils were stimulated with 100 ng/ml PMA for the indicated times. The production of H2O2 after stimulation was measured by FACS analysis and expressed as the MFI of the cells, which were untreated (•) or pretreated with 10 mM CD for 30 min ( ). Alternatively, neutrophils were treated with CD for 30 min, washed twice in PBS, and stimulated in incubation buffer after a 30-min recovery period ( ). Statistical differences between CD-treated and untreated groups (**) and CD-treated and washed groups (*) were were calculated using the Wilcoxon nonparametric test for paired samples (P<0.05). (C) Unprimed neutrophils and neutrophils that were primed with 5 ng/ml GM-CSF (GM) for 15 min were stimulated with 1 µM fMLP (f) for 30 s. Where indicated, cells were pretreated with 10 mM CD for 30 min. Alternatively, neutrophils were treated with CD for 30 min, washed twice in PBS, and stimulated in incubation buffer after a 30-min recovery period (CD+wash). Phosphorylated PKB/Akt (p-PKB) and ERK1/2 (p-ERK1/p-ERK2; top and middle panels) were detected by Western blot analysis. (D) Cells were incubated with GM-CSF for 15 min, with or without pretreatment with CD, and phosphorylated STAT5 was detected by Western blot analysis. Equal loading in the samples was confirmed by immunodetection of ERK1/2 (C, bottom panel, and D, lower panel). Three individual experiments were performed, and a representative example is shown.
|
The effect of CD on ROS production induced by the extremely potent stimulus PMA was also investigated (Fig. 1B) . CD treatment abrogated PMA-induced ROS production (MFI1754±78 vs. 123±16, t=20 min, n=6, P=0.03), and ROS production could be restored partially by washing away the CD (MFI 123±16 with CD vs. 1060±152 after wash, t=20 min, n=6, P=0.03). These experiments show that not only receptor-dependent activation of ROS production is dependent on lipid rafts but also ROS production stimulated by a receptor-independent process.
The effect of CD treatment on signal transduction pathways
Our previous studies showed that only receptor-dependent activation (i.e., GM-CSF priming) of ROS production was impaired in MDS neutrophils, and PMA-triggered ROS activity was normal. CD treatment could theoretically perturb receptor-dependent ROS production by interfering with the signal transduction pathways leading to ROS production or with the NADPH oxidase directly. To investigate the first hypothesis, we examined the effect of CD on two signaling pathways known to be involved in the neutrophil respiratory burst: the MEK-ERK and PI-3K-PKB pathways. Neutrophils were stimulated with fMLP, with or without preincubation with GM-CSF. As described before [7
], priming of neutrophils with GM-CSF resulted in enhanced phosphorylation of ERK1/2 and PKB in response to fMLP stimulation (Fig. 1C)
. Pretreatment of healthy neutrophils with 10 mM CD resulted in near-complete abrogation of ERK1/2 and PKB activation. In addition, washing cells after CD treatment partially restored the fMLP-triggered activation of ERK1/2 and PKB in unprimed and GM-CSF-primed neutrophils. These results indicate that lipid rafts are involved in the fMLP-induced signal transduction pathways involved in ROS production.
In contrast to stimulation with fMLP, GM-CSF stimulation of cells results in the phosphorylation of STAT5. Pretreatment of cells with CD did not affect STAT5 activation (Fig. 1D) , indicating that depletion of cholesterol does not result in disruption of all signaling pathways.
The effect of CD treatment on flavocytochrome b558 localization in healthy neutrophils
Another mechanism for CD to attenuate receptor-dependent ROS production could be by affecting the NADPH oxidase directly. We investigated the presence of the NADPH oxidase components in lipid rafts from neutrophils by isolating low-density Triton X-insoluble fractions. The successful isolation of lipid rafts was confirmed by the presence in Fraction 4 of the Src-kinase Lyn, which is a known marker for rafts [36
]. In unstimulated neutrophils, the flavocytochrome b558 subunits gp91phox and p22phox were present in the lipid raft fraction (Fig. 2
), whereas p47phox was not (not shown). Treatment of neutrophils with CD resulted in disruption of the rafts, as indicated by a decreased presence of Lyn in Fraction 4 (n=2). Furthermore, raft association of gp91phox and p22phox was decreased upon CD treatment, indicating that inappropriate localization of the NADPH components might be a contributing factor in the disrupted ROS production upon CD treatment.
 View larger version (33K): [in a new window] |
Figure 2. CD treatment reduces association of Lyn, gp91phox, and p22phox with the lipid raft fractions. Isolated neutrophils were untreated or treated with 10 mM CD for 30 min. Cells were lysed in 0.75% Triton X-100 and subjected to ultracentrifugation in Optiprep gradients as described in Materials and Methods. Fractions were collected and numbered 1–6 from bottom to top. Each fraction (100 µl) was run on Western blot, and the presence of gp91phox, Lyn, p22phox, and Rac was shown by immunodetection. Floating lipid rafts are present in Fraction 4, as determined by the presence of the lipid raft marker Lyn in this fraction. A representative experiment is shown.
|
 View larger version (34K): [in a new window] |
Figure 3. Decreased presence of Lyn, gp91phox and p22phox in lipid raft fractions from MDS patients. (A) Isolated neutrophils from MDS patients and healthy controls were lysed in 0.75% Triton X-100 and subjected to ultracentrifugation in Optiprep gradients as described in Materials and Methods. Fractions were collected and numbered 1–6 from bottom to top. Each fraction (100 µl) was run on Western blot, and the presence of gp91phox, Lyn, p22phox, and Rac was shown by immunodetection. Four individual experiments were performed, and two representative examples are shown. (B) Protein levels of Lyn, gp91phox, and p22phox were quantified by densitometry of the films and divided by the densitometry values of Rac present in the high-density fractions (not shown) to control for equal loading. The means of the normalized values were calculated for the healthy donors and the MDS patients (n=4).
|
 View larger version (16K): [in a new window] |
Figure 4. Decreased GM1 expression on the plasma membrane on neutrophils from MDS patients. (A) Unprimed and GM-CSF-primed neutrophils were stimulated with 1 µM fMLP for the indicated time. Stimulation was stopped by adding ice-cold PBS, centrifugation, and resuspension of cells in PBS containing 5 µg/mL cholera toxin subunit B-Alexa Fluor 555. After staining, cells were washed twice, and FACS analysis was performed. One representative example of six individual experiments is shown. (B) Neutrophils from six MDS patients and healthy controls were left unstimulated (solid bars) or were primed with 5 ng/ml GM-CSF for 15 min prior to stimulation with 1 µM fMLP (shaded bars). GM1 staining was measured by FACS analysis. The means of the six individual experiments are shown. (C) Neutrophils were left unstimulated (solid bars) or were primed with 5 ng/ml GM-CSF for 15 min prior to stimulation with 1 µM fMLP for 30 min (shaded bars). The production of H2O2 after stimulation was measured by FACS analysis, and the mean of six individual experiments was expressed as the MFI of the cells. Significant differences in ROS and GM1 expression on GM-CSF-primed neutrophils from MDS patients and healthy controls were calculated using the Wilcoxon nonparametric test for paired samples and are indicated with an asterisk (P=0.028).
|
In one patient, we found a normal GM-CSF priming of fMLP-triggered ROS production when compared with the healthy control. In this patient, the fMLP-induced expression of GM1 at the plasma membrane of GM-CSF-primed neutrophils was also normal (data not shown).
Decreased intracellular GM1 expression in neutrophils from MDS patients
The increased expression of GM1 at the plasma membrane upon stimulation of neutrophils might be a result of fusion of GM1-containing granule membranes with the plasma membrane or by de novo synthesis of this ganglioside. To address this question, we studied the intracellular expression of GM1 in neutrophils. Ultrathin cryosections of neutrophils were stained with cholera toxin subunit B, followed by straptavidin gold, after which the gold label was visualized by electron microscopy. As shown in Figure 5A
, stimulation of GM-CSF-primed healthy neutrophils with fMLP resulted in an increase in gold spheres present in healthy neutrophils, representing an increase in the total amount of GM1 present in these cells. These results suggest that the presence of the lipid raft marker at the plasma membrane increases upon stimulation through de novo synthesis, rather than fusion of GM1-containing granule membranes with the plasma membrane.
 View larger version (39K): [in a new window] |
Figure 5. GM1 expression is lower in MDS neutrophils stimulated with GM-CSF and fMLP. (A) Unstimulated neutrophils and neutrophils that were primed with 5 ng/ml GM-CSF for 15 min prior to stimulation with 1 µM fMLP for 1 min were fixed in 4% paraformaldehyde + 0.1% glutaraldehyde. Cryosections were prepared as described in Materials and Methods and stained with 25 µg/ml cholera toxin subunit B-biotin followed by streptavidin-labeled 5 nm gold particles. Cryosections were analyzed with a CM100 electron microscope. One representative example of three individual experiments is shown. (B) Unstimulated neutrophils and neutrophils that were primed with 5 ng/ml GM-CSF for 15 min prior to stimulation with 1 µM fMLP were compared between two MDS patients and two healthy controls. At least two cryosections were analyzed per sample with a CM100 electron microscope. A total of six cells was studied per sample, and three pictures were taken of every cell investigated. Cytoplasmatic area per cell was measured using AnalySIS (SIS) and divided by the number of gold particles per cell to obtain a quantitative measure of gold labeling. A representative example of two individual experiments is shown. A significant difference in gold labeling in cells from the MDS patient versus the healthy control was calculated using the Wilcoxon nonparametric test for paired samples and is indicated with an asterisk (P<0.05).
|
Taken together, these results indicate that GM1 expression, a measure of lipid raft formation, correlates with the amount of ROS that is generated in response to fMLP in GM-CSF-primed neutrophils. In MDS patients, a reduced lipid raft formation at the plasma membrane of GM-CSF-primed neutrophils correlates with a reduced priming of fMLP-triggered ROS production in this subset of cells.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the present study, we measured ROS production using an assay that measures intracellular H2O2 production. As CD extracts cholesterol only from the plasma membrane, it is likely that its effect on intracellular ROS production is mainly through interference with efficient signaling. Lipid rafts have been postulated to function as signaling platforms by facilitating the receptor-induced activation of several signal transduction molecules [21 , 23 , 24 , 45 , 46 ]. In this study, we show that fMLP-induced activation of PKB and ERK1/2 in unprimed or GM-CSF-primed neutrophils was dependent on cholesterol-rich lipid rafts. In contrast, GM-CSF-induced STAT5 activation was not affected by disruption of lipid rafts. Our previous results indicate that in MDS, PI-3K-PKB and ERK1/2 signaling in GM-CSF-primed neutrophils is impaired in response to fMLP, whereas STAT5 activation by GM-CSF stimulation is normal [7 ]. These results suggest that lipid raft formation in MDS is affected. Recently, it has been shown that the fMLP receptor moves into lipid rafts upon ligand binding and that this receptor clustering might be necessary for maximal signaling [47 ]. It is conceivable that the GM-CSF receptor does not enter lipid rafts itself, explaining the lack of effect of CD treatment on STAT5 activation, but rather, induces an increased clustering of fMLP receptor-containing rafts, resulting in enhanced fMLP signaling. It has been suggested recently that stimulation of platelet-derived growth factor receptors in lipid rafts causes tyrosine phosphorylation of epidermal growth factor receptors present in the same membrane fractions, suggesting that cross-talk between receptors might be facilitated by movement into the same microdomains [48 ].
To investigate lipid rafts in stimulated neutrophils, we studied the expression of GM1, which has been shown to be a measure of lipid raft accumulation [49 , 50 ]. We showed that in MDS patients, GM1 expression on the plasma membrane of fMLP-stimulated, GM-CSF-primed cells was significantly lower than in healthy controls. Furthermore, this decreased GM1 expression correlated with an impaired ROS production found in this subset of cells. These results indicate that lipid raft formation is impaired upon stimulation of GM-CSF-primed MDS neutrophils. By electron microscopy, we demonstrated that GM1 was newly synthesized upon stimulation of neutrophils with GM-CSF and fMLP and that this de novo synthesis was impaired in MDS patients.
In conclusion, our studies show that lipid rafts are involved in signal transduction pathways that play a role in ROS production in fMLP-stimulated neutrophils. We demonstrate that in MDS patients, accumulation of lipid rafts upon stimulation of GM-CSF-primed cells with fMLP is affected. This in turn might lead to the impaired activation of the PI-3K-PKB and ERK1/2 pathways, resulting in a decreased ROS production in GM-CSF-primed neutrophils.
Received May 9, 2006; revised September 20, 2006; accepted September 21, 2006.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and GM-CSF J. Leukoc. Biol. 64,537-545[Abstract] is signaled by enhanced phosphatidylinositol 3,4,5-triphosphate but not inositol 1,4,5-triphosphate accumulation FEBS Lett. 439,147-151[CrossRef][Medline]-induced phosphorylation of extracellular signal-regulated kinase result in decreased migration of neutrophils from patients with Myelodysplasia J. Leukoc. Biol. 77,257-266
RIIIb and cytochrome b558 Exp. Cell Res. 308,156-165[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  G. M. Fuhler, A. L. Drayer, S. G. M. Olthof, J. J. Schuringa, P. J. Coffer, and E. Vellenga Reduced activation of protein kinase B, Rac, and F-actin polymerization contributes to an impairment of stromal cell derived factor-1 induced migration of CD34+ cells from patients with myelodysplasia Blood, January 1, 2008; 111(1): 359 - 368. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
