Cleavage fragments of the third complement component (C3) enhance

Leukemia (2007) 21, 973–982 & 2007 Nature Publishing Group All rights reserved 0887-6924/07 $30.00 www.nature.com/leu

ORIGINAL ARTICLE Cleavage fragments of the third complement component (C3) enhance stromal derived factor-1 (SDF-1)-mediated platelet production during reactive postbleeding thrombocytosis M Wysoczynski1, M Kucia1, J Ratajczak1 and MZ Ratajczak1,2 Stem Cell Biology Program, James Graham Brown Cancer Center, University of Louisville, Louisville, KY, USA and 2Department of Physiopathology, Pomeranian Medical University, Szczecin, Poland 1

We hypothesized that the third complement component (C3) cleavage fragments (C3a and des-ArgC3a) are involved in stress/ inflammation-related thrombocytosis, and investigated their potential role in reactive thrombocytosis induced by bleeding. We found that platelet counts are lower in C3-deficient mice in response to excessive bleeding as compared to normal littermates and that C3a and des-ArgC3a enhance stromalderived factor-1 (SDF-1)-dependent megakaryocyte (Megs) migration, adhesion and platelet shedding. At the molecular level, C3a stimulates in Megs MAPKp42/44 phosphorylation, and enhances incorporation of CXCR4 into membrane lipid rafts increasing the responsiveness of Megs to SDF-1. We found that perturbation of lipid raft formation by statins decreases SDF-1/C3a-dependent platelet production in vitro and in an in vivo model statins ameliorated post-bleeding thrombocytosis. Thus, inhibition of lipid raft formation could find potential clinical application as a means of ameliorating some forms of thrombocytosis. Leukemia (2007) 21, 973–982. doi:10.1038/sj.leu.2404629; published online 1 March 2007 Keywords: megakaryopoiesis; platelets; complement; chemokines; CXCR4; lipid rafts

Introduction Megakaryopoiesis is regulated by several factors that affect the proliferation and differentiation of megakaryopoietic cells.1,2 The fact that mice with ‘knockout’ of thrombopoietin (TPO) and thrombopoietin receptor (c-mpl) show a 90% reduction in both the number of megakaryocytes (Megs) in bone marrow (BM) and circulating platelets indicates that TPO is a crucial regulator of megakaryopoiesis.3–5 Nevertheless, the fact that some Megs and platelets are still present in these animals even in the total absence of TPO suggests that some other factors can compensate for TPO deficiency and the cytokines that signal through gp130 such as interleukin-6 (IL-6), IL-11, leukemia inhibitory factor, cilliary neurotropic factor and oncostatin M have been suggested to be such factors.6–9 However, recent data show that IL-6 and IL-11 do not induce platelet production in thrombocytopenic, TPO-deficient (Tpo/) and TPO-receptor-deficient (Mpl/) mice.10 Thus, it is unlikely that gp130 signaling cytokines are significantly involved in platelet production in these animals.4–10 Correspondence: Dr MZ Ratajczak, Stem Cell Biology Program, James Graham Brown Cancer Center, University of Louisville, Louisville, KY 40202, USA. E-mail: [email protected] Received 4 December 2006; revised 16 January 2007; accepted 25 January 2007; published online 1 March 2007

The a-chemokine stromal-derived factor-1 (SDF-1) has been proposed as a new regulator of Megs maturation and platelet formation.11–16 A recently published study clearly demonstrated that SDF-1, together with fibroblast growth factor-4 (FGF-4), restored thrombopoiesis in Tpo/ and Mpl/ mice by directing trans-localization of megakaryocytic progenitors positive for CXCR4, the receptor for SDF-1, from the endostial to the vascular niche, thereby promoting survival, maturation to megakaryocytes and platelet production.17 Furthermore, there is growing evidence that the responsiveness of hematopoietic cells to SDF-1 is optimal when the CXCR4 receptor is included into membrane lipid rafts.18–20 The formation of lipid rafts in vivo may be perturbed by cholesterollowering drugs, for example, statins,18–20 which are effective in lowering LDL cholesterol and exert pleiotropic effects on mature platelets, for example, inhibit their activation.21 This latter effect is probably due to lowering of the cholesterol content in the platelet membranes leading to disruption of membrane lipid raft formation, as lipid rafts are necessary for proper platelet activation and signaling21 and, as we hypothesized, platelet production. However, under steady-state conditions, statins do not influence thrombopoiesis, no direct experimental studies have been performed on their effects on platelet production in reactive thrombocytosis. We recently reported that the complement (C) system and third complement protein (C3) cleavage fragments C3a and des-ArgC3a enhance the responsiveness of hematopoietic cells to an SDF-1 gradient.22 C3a binds to Gai-protein-coupled seven transmembrane receptor (C3aR) and we reported that C3aR is expressed on human and murine hematopoietic stem/progenitor cells (HSPC).22,23 In contrast, the receptor for des-ArgC3a has not been identified yet. As C is activated in several clinical conditions associated with high platelet counts (e.g., chronic inflammation, hemorrhage), we hypothesized that C3a and des-ArgC3a are involved in stress/inflammation-related thrombocytosis. Supporting this are the observations that C3-deficient mice, which have normal platelet counts in steady-state conditions,23 display delayed recovery of platelets after sublethal irradiation or hematopoietic transplants.23 We report that C3a and des-ArgC3a modulate the responsiveness of Megs to SDF-1 and postulate that the crosstalk between C3aR and CXCR4 receptors, we have newly identified, here plays an important and previously unrecognized role in stress/ inflammation-dependent Megs maturation and platelet formation.

Materials and methods

Human CD34 þ cells, megakaryoblasts and platelets Light-density BM mononuclear cells (BM MNC) were obtained from consenting healthy donors and enriched for CD34 þ cells

Complement as a new regulator of megakaryopoiesis M Wysoczynski et al

974 by immunoaffinity selection with MiniMACS paramagnetic beads (Miltenyi Biotec, Auburn, CA, USA) (purity 495%) and were expanded in a serum-free liquid system, and growth of CFU-Meg was stimulated with recombinant human (rh) TPO (100 ng/ml) (R&D Systems, Minneapolis, MN, USA) as described.16 After incubation for 11 days at 371C, about 85% of the expanded cells were positive for the megakaryocyticspecific marker CD41, and this population was further enriched to 495% purity by additional selection with immunomagnetic beads (Miltenyi Biotec) as described previously by us.16 Gelfiltered platelets (GFP) were prepared from four individuals as described previously.13,16 Marrow aspiration and blood donation from normal volunteers was carried out with the donors’ informed consent obtained through the Institutional Review Board.

CFU-Meg assay

CD34 þ BMMNC were resuspended in 1% methylcellulose in Iscove’s DMEM (Gibco BRL, Grand Island, NJ, USA) (104/ml) supplemented with 25% artificial serum as described.9,16 CFUMeg growth was stimulated with a suboptimal (10 ng/ml) and optimal dose of rhTPO (100 ng/ml). Cultures were incubated at 371C in a fully humidified atmosphere supplemented with 5% CO2. Under these conditions, after 11 days, approximately 100% of the colonies were glycoprotein aIIb/b3 positive.9,16

Cell lines The MO7E cell line used in these studies was purchased from ATCC (Rockville, MD, USA) and maintained in Iscove medium (Gibco BRL, Long Island, NY, USA) supplemented with 10% bovine calf serum (BCS) (Hyclone, Logan, UT, USA) and 5 ng/ml of GM-CSF (R&D Systems).

Bleeding procedure Mice were anesthetized by i.p. injection of Ketamine (100 mg/kg, Sigma, St Louis, MO, USA) and bled from the retro-orbital plexus as recommended by the Animal Resource Center, Case Western Reserve University (http://labanimals.case.edu/ templates_retro_orbital_bleeding.html).

ELISA on serum from murine BM Blood samples were collected into EDTA tubes and plasma was separated immediately by centrifugation at 2000 g for 15 min at 41C. For detection of des-ArgC3a in murine plasma, we employed the rat/mouse ELISA kit according to the producer’s protocol (Cedarline, ON, Canada). In brief, the microtiter plates were coated with a chicken anti- des-ArgC3a antibody and serial dilutions of mouse serum (1:100 starting dilution in blocking buffer) were added and incubation for 1 h at 371C was carried out. Subsequently, polyclonal chicken anti-mouse des-ArgC3a (biotinlated, 4 mg/ml in blocking buffer, ICN Pharmaceuticals Inc., OH, USA) was mixed with extravidin-peroxidase and the mixture was then added to the wells for further incubation of 1 h. des-ArgC3a-containing complexes were detected by measuring the optical density in an ELISA reader (KC junior, BIO-TEK Instruments Inc., VT, USA) at 450 nm with a reference wavelength set at 630 nm.

FACS analysis Expression of C3aR was evaluated by fluorescence-activated cell sorter (FACS) as described previously.22 C3aR was detected Leukemia

on human Megs with mouse anti-human C3aR monoclonal antibody (MoAb), clone no. 8H1 and a fluorescein isothiocyanate (FITC)-labeled goat anti-mouse antibody. Subsequently, the cells were stained with phycoerythrin (PE)-labeled anti-CD41 MoAb (Becton Dickinson, San Jose, CA, USA), washed, fixed in 1% paraformaldehyde and subjected to FACS analysis using a FACScan analyser (Becton Dickinson).

Isolation of mRNA and RT-PCR for detection of C3aR and C5L2 Total mRNA was isolated with the RNeasy Mini Kit (Quiagen Inc., Valencia, CA, USA) as described.20 We employed following primers – for detection of C3aR forward 50 -GCC GCC TGG AGA AAT GAA TGA TAG G-30 and reverse 50 -AGA AAG ACA GCC ACC ACC ACG-30 , and for detection of C5L2 forward 50 -CCT GGT GGT CTA CGG TTC AG-30 and reverse 50 -GGG CAG GAT TTG TGT CTG TT-30 . Amplified products (10 ml) were electrophoresed on a 2% agarose gel.

Real-time PCR analysis of MMPs and VEGF Detection of MMP2, MMP9, VEGF and b2-microglobulin mRNA levels was performed by real-time RT-PCR using an ABI PRISM 7000 Sequence Detection System (ABI). Each 25 ml reaction mixture contained 12.5 ml SYBR Green PCR Master Mix, 10 ng of cDNA template and primer for MMP9 forward 50 GGA CGA CGT GGG CTA CGT-30 , reverse 50 -AAT CTC ACC GAC AGG CAG CT-30 , for MMP2 forward 50 -TGG GAC AAG AAC CAG ATC ACA TA-30 , reverse 50 -TTT CGA GTC TCC ACG CAT CTC-30 , for VEGF forward 50 -TGA GCG GCT CAT CTA CTT CTA TGT-30 , reverse 50 -CAC CGG CTG GCC CTC TA-30 , for b2-microglobulin forward 50 -TGA CTT TGT CAC AGC CCA AGA TA-30 , reverse 50 -AAT GCG GCA TCT TCA AAC CT-30 . Relative quantitation of MMP2, MMP9 and VEGF mRNA expression was calculated with the comparative threshold cycle (Ct) method described elsewhere.24

Apoptosis assay Apoptosis of Megs was assessed by cytometric analysis after staining cells with FITC-Annexin V (apoptosis detection kit from R&D Systems) according to the manufacturer’s protocol as described.9,16

Megs proliferation by MTT assay The MTT assay was performed according to the manufacturer’s recommendations (Promega, Madison, WI, USA). Briefly, cells were seeded in 96-well plates at 5 104/well in 100 ml of RPMI medium containing 0.5% BSA plus 10 ng/ml or 100 ng/ml of TPO þ SDF-1 (300 ng/ml) (R&D Systems) þ /C3a or des-ArgC3a (1 mg/ml) (Calbiochem, San Diego, CA, USA). After 24 h, 20 ml of CellTiter 96 Aqueous One Solution reagent was added to each well and the plates were incubated for 3–4 h. Subsequently, plates were read at 490 nm using an automated plate-reader.9,16

Calcium flux assay Briefly, the cells were incubated for 30 min at 301C with 1–2 mM Fura-2/AM (Molecular Probes, Carlsbad, CA, USA). After incubation, the cells were washed once, resuspended in loading buffer without BCS, stimulated with C3a (1 mg/ml), des-ArgC3a (1 mg/ml) or SDF-1b (300 ng/ml) and analyzed within 1 h as described previously.22


Actin polymerization assay The cells were washed three times and resuspended in assay medium (Iscove’s DMEM with 0.5% BSA) at a concentration of 106 cells/ml. The ligands were added to the cell suspension and, at indicated time points, 100 ml cell solution was transferred to the fixation solution (Cytofix/Cytoperm; Pharmingen). Subsequently, cells were fixed and then centrifuged and resuspended in 100 ml permeabilization reagent (Perm/wash buffer; PharMingen). After 5 min of incubation in this solution, 1 U/ml Alexa 488 phalloidin (Molecular Probes) was added to visualize the F-actin. MFI was measured by FACScan (Becton Dickinson) and examined using a BX51 fluorescence microscope (Olympus America, Melville, NY, USA) equipped with a charge-coupled device camera (Olympus America). Separate pictures were merged using Image-Pro Plus software (Media Cybernetics, Silver Spring, MD, USA).

Chemotaxis assay Cells were resuspended in RPMI with 0.5% BSA and equilibrated for 10 min at 371C and chemotaxis assays were performed by employing Costar Transwell 24-well plates (Costar Corning, Cambridge, MA, USA) with 8 mm-pore filters as described.20 The results are presented as a chemotactic index (the ratio of the number of cells that migrated toward the medium containing C3a and/or SDF-1 to the number of cells that migrated toward the medium alone).

Adhesion assay Ninety-six-microtiter plates (Dynatech Labs, Chantilly, VA, USA) were covered with 4 mg/ml fibrinogen (Sigma) by incuba-

2000 1800 1600 1400 1200 1000 800 600 400 200 0

BLEEDING

Isolation of lipid rafts

MO7E cells (1 108) were lysed as described.20 Samples were overlain with 30 and 5% sucrose in MEB and were centrifuged at 100 000 g for 17 h. Fractions were gently removed from the top of the gradient, and n-octylglucoside was added to each fraction (60 mM final) to solubilize rafts. For detection, Western blot analysis was carried out using standard techniques with a CXCR4 antibody (Serotec, Oxford, UK) and cholera toxin subunit B conjugated with horseradish peroxidase (HRP) (Sigma, Milwaukee, WI, USA).20

wt C3 -/-

*

A

M

1

2

3

1

2

3

3500 3000

1

2

3

4

5

*

6 7 8 9 10 11 12 13 14 15 16 17 days after bleeding

*

*

wt C3 -/-

Megs

*

B

a

b

10000

2000

1000 FL2-H

2500 1500 1000

14.8

66.5

100 10

500 -1

4

9

14 19 hrs after bleeding

24

29

Figure 1 Thrombocytosis and complement activation during excessive bleeding. (a) Kinetics of platelet counts after bleeding in wt versus C3/ mice. (b) Changes in the level of des-ArgC3a in serum of wt mice and C3/ mice (control) after bleeding were determined using ELISA. The data represent the average of three independent experiments (using 6–8 mice for each experiment). Data are mean7s.d. *Po0.00001 wt versus C3/ mice.

c

100

100

80

80

60 40 20

8.68

1

0

platelets

% of Max

BLEEDING

4500 4000

ng/ml

Transwell chambers (8 mm pore size) were covered with an impermeable monolayer of human umbilical vein endothelial cells (HUVEC) which were isolated as described previously. HUVECs were activated overnight with IL-1b (10 U/ml) before assay. Megs that had been expanded from CD34 þ cells were added to the top transwell chambers, and aliquots of 650 ml serum-free medium containing SDF-1, C3a and des-ArgC3a, alone or in combination, were placed in the lower chambers. In the platelet formation assay performed for 48 h, the number of CD41 þ platelets present in the lower chambers was determined after staining with anti-CD41 MoAbs and using flow cytometry (FACSscan, Beckton Dickinson).

1. C3aR 2. C5L2 3. β-actin -1 0

b

*

*

975

Transmigration and platelet formation assay

% of Max

PLT (K/uL)

a

tion overnight at 41C, and CD41 þ cells (106/ml) were incubated for 30 min at 371C in serum-free Iscove DMEM in the absence or presence of SDF-1 (1 mg/ml), C3a (1 mg/ml), des-ArgC3a (1 mg/ml), SDF-1 þ C3a or SDF-1 þ des-ArgC3a. Cell suspensions (100 ml) were applied to the wells and incubated for 10 min at 371C. The number of adherent cells was determined using inverted microscopy.

1

10

10.1 100 1000 10000 FL1-H

0 10

10 10 FL2-H

40 20

99.1% 10

60

10

0 10

30.3% 10

10 10 FL2-H

10

Figure 2 Human Megs and platelets express C3aR. (A) RT-PCR analysis of mRNA expression for C3aR and C5L2 on human Megs and platelets. M, marker; C3aR, lanes 1; C5L2, lanes 2 and b-actin, lanes 3. (B) (a) double staining of human Megs with FITC-anti-C3aR and PEanti-CD41, (b) human platelets stained with PE-anti-CD41 and (c) with FITC-anti-C3aR. The data represent the average of two (RT-PCR) and three (FACS analysis) independent experiments. Representative results are shown. Leukemia


976

Phosphorylation of intracellular pathway proteins Western blot analysis was performed on extracts prepared from quiescent CD34 þ -derived Megs as described.16 Megs were then stimulated with 1 mg/ml of C3a or 1 mg/ml of des-ArgC3a for 1, 2, 5 and 10 min at 371C and then lysed for 10 min on ice in M-Per lysing buffer (Pierce, Rockford, IL, USA) containing protease and phosphatase inhibitors (Sigma). The extracted proteins were

1.6

Aa

separated and analyzed for phosphorylation, of 44/42 MAPK as described.16 Equal loading in the lanes was evaluated by stripping the blots and reprobing them with the appropriate monoclonal or polyclonal antibody p42/44 anti-MAPK antibody clone 9102 (New England Biolabs). The membranes were developed with an ECL reagent (Amersham Life Sci.), dried and exposed to film (HyperFilm, Amersham Life Sci.).

b 1.55 1.5 Ratio

Ratio

1.5

SDF-1

C3a

1.55

1.45 1.4

1.45 1.4

1.35 1.35

1.3 50

99 time (sec)

148

1

d 2.3

1.7

2.1

C3a

c 1.9

148

SB290157 pre-treated

1.1

1.7

SDF-1

Ratio

SDF-1

1.3

99 time (sec)

1.9

desArg

Ratio

1.5

50

C3a

1

1.5 1.3 1.1

0.9

0.9

0.7

0.7 1

B

50

99

148 197 time (sec)

246

295

240

180 160 140

99

148 time (sec)

197

246

SDF-1 desArgC3a

150 % of control

200

50

160

SDF-1 C3a SDF-1+C3a

220 % of control

1

SDF-1 + desArgC3a

140 130 120 110

120 100

100 0

20

40

C

60

80 100 time (sec)

CONTROL

120

140

160

SDF-1

180

0

C3a

20

40

60

80 100 120 time (sec)

140

160

180

desArgC3a

Figure 3 Human Megs express functional C3aR. (A) C3a but not des-ArgC3a induces calcium flux in human Megs. Normal human Megs Fura-2loaded were stimulated by (a) C3a (1 mg/ml), (b) SDF-1 (300 ng/ml), (c) des-ArgC3a (1 mg/ml) followed by SDF-1 (300 ng/ml) or (d) preincubated with SB290157 (C3aR-specific inhibitor) and then stimulated by C3a (1 mg/ml) and SDF-1 (300 ng/ml). All compounds were added at the indicated time points, and calcium flux was evaluated by spectrophotofluorimetry. Data presented are from a representative experiment, which was repeated three times with similar results. (B) FACS analysis of F-actin polymerization in human Megs. SDF-1, C3a and SDF-1 þ C3a but not des-ArgC3a induce F-actin polymerization. Data presented are from a representative experiment, which was repeated three times with similar results. Po0.0001. (C) Immunofluorescence analysis of F-actin polymerization in human Meg (green fluorescence, phalloidin; blue, DAPI). The experiment was repeated three times on cells from three independent donors with similar results. Representative staining is shown. Leukemia


*

8000

*

*

6000

*

4000 2000

b

700

C 3a +S B SD F1 hi gh

rg

sA

d

es SD A rg C F3a 1+ SD de FsA 1+ rg C 3a de

C 3a SD F1+ SD C F3a 1+ C 3a +S B

0 lo w

Given that excessive bleeding leads to reactive thrombocytosis,25,26 and C becomes activated in several stress-related situations,27 our aim was to determine whether C3 cleavage fragments might be implicated in post-bleeding thrombocytosis. To address this, we first investigated the kinetics of platelet counts after excessive bleeding (20% of blood vol) in wild-type (wt) and C3-deficient mice and, in parallel, measured C activation by detecting the level of C3 cleavage fragment (des-ArgC3a) in the peripheral blood of bled animals (Figure 1). Figure 1a shows that both wt and C3-deficient mice display a biphasic increase in peripheral blood counts at days 1–2 and 6–10. Although the first elevation in platelet counts may reflect the supply of platelets by Megs located in the BM endothelial niche, the second elevation is likely related to generation of new Megs from a pool of CFU-Meg that are attracted to this niche from the osteoblastic niche.17 Although C3-deficient mice have normal platelet counts in normal steady-state conditions,23 they generated fewer platelets after excessive bleeding as compared to wt littermates (Figure 1a). Next, using an ELISA assay (Figure 1b), we found that C3 is cleaved in peripheral blood during bleeding as seen by an increase in des-ArgC3a in the blood of normal mice. As expected, however, no des-ArgC3a was detectable in C3-deficient control animals.

a 10000

c-

Extensive bleeding in C3-deficient mice leads to a lower platelet count and in normal mice causes activation of C3

SD F1

Results

% of control

Arithmetic means and standard deviations were calculated on a Macintosh computer powerbase 180, using Instat 1.14 (Graphpad, San Diego, CA, USA) software. Statistical significance was defined as Po0.01. Data were analyzed using the Student’s t-test for unpaired samples.

977 Likewise, we also observed that C3a but not des-ArgC3a induced F-actin polymerization in human Megs (Figure 3b). However, the F-actin polymerization induced by C3a alone was shorter than after addition of SDF-1, and addition of C3a to SDF-1 significantly enhanced the SDF-1-mediated effect. The fact that des-ArgC3a did not induce F-actin polymerization (Figure 3b) and SB290157-inhibited C3a-induced F-actin polymerization (data not shown) indicates that C3aR is responsible for the polymerization of F-actin. Our FACS-based F-polymer-

* *

600 % of control

Statistical analysis

500 400 300

**

200 100 0 control

SDF-1

C3a

SDF-1+C3a desArgC3a SDF-1C3a desArg

C3aR is expressed on normal Megs and platelets

c % of control

Next, we evaluated whether normal human Megs and platelets express receptors that bind C3 soluble cleavage fragments. Using RT-PCR, we observed that highly purified normal human Megs obtained from expansion of BM-derived CD34 þ cells as well as peripheral blood platelets express mRNA for C3aR (Figure 2a). More importantly, we detected C3aR protein expression by FACS on megakaryocytic cells derived from ex vivo expanded BM-derived CD34 þ cells and purified peripheral blood platelets (Figure 2b). Normal Megs and platelets also express mRNA for the C5L2 receptor, an orphan receptor28 that may potentially bind des-ArgC3a (Figure 1a).

230 210 190 170 150 130 110 90 70 50

*

control

SDF-1

C3a

*

SDF-1+C3a desArgC3a SDF-1C3a desArg

C3aR is functional on normal human Megs The finding that C3aR is expressed on normal human Megs suggests a role for the C3 cleavage fragments in regulating normal human megakaryopoiesis. Highly purified human Megs were exposed to C3a or des-ArgC3a, and assayed for calcium flux, F-actin polymerization and chemotactic responses. As positive controls, we employed SDF-1 as a stimulant (Figure 3). Figure 3a shows that C3a, like SDF-1, stimulated calcium flux in megakaryocytic cells. The fact that des-ArgC3a did not stimulate calcium flux and that the C3aR inhibitor SB290157 inhibited C3aR but not SDF-1-mediated calcium flux supports the notion that this effect is specific to C3aR.

Figure 4 C3a and des-ArgC3a enhance SDF-1-dependent chemotaxis, adhesion and platelet production. (a) C3a and des-ArgC3a enhance the chemotactic response of human Megs to a low threshold dose of SDF-1 (10 ng/ml), in a C3aR-independent manner as shown after employing SB290157 (C3aR-specific inhibitor). As positive control, SDF-1, was used in high dose (300 ng/ml). The data represent the average of three independent experiments. Data are mean7s.d. *Po0.00001 versus SDF-1 low. (b) Both C3a (1 mg/ml) and des-ArgC3a (1 mg/ml) enhance SDF-1 (1 mg/ml)-dependent adhesion of Megs to fibrinogen. Lower part – representative pictures (inverted microscope). The data represent the average of three independent experiments. Data are mean7s.d. *Po0.00001 versus SDF-1 low. (c) C3a (1 mg/ml) and des-ArgC3a (1 mg/ml) increase SDF-1 (300 ng/ml)-dependent production of platelets. The data represent the average of four independent experiments. Data are mean7s.d. *Po0.00001 versus SDF-1 alone. Leukemia


978 ization data were subsequently confirmed at the single-cell level by direct F-actin staining (Figure 3c).

C3a and des-ArgC3a do not affect CFU-Meg proliferation and Megs survival Next, we investigated whether C3a and des-ArgC3a could influence the proliferation/survival of normal human Megs. Highly purified CD34 þ cells were cultured in a serum-free liquid cultures or serum-free methylcellulose cloning medium with various doses of C3a or des-ArgC3a (0.1–10 mg/ml) and stimulated to grow CFUMeg colonies with sub-optimal (10 ng/ml) and optimal (100 ng/ml) doses of TPO alone or TPO þ SDF-1. In another set of experiments, we incubated Megs for 24 h in serum-free medium in the presence or absence of C3a or des-ArgC3a (0.1–10 mg/ml) and estimated cell viability/Annexin V binding after this period of time. In all these experiments, however, no effect of C3a or des-ArgC3a on the proliferation/survival of Megs was observed (data not shown). Thus, C3a and des-ArgC3a do not affect the proliferation/survival of hematopoietic cells.

C3a and des-ArgC3a increase responsiveness of Megs to an SDF-1 gradient and enhance SDF-1-dependent migration, adhesion and platelet production Although C3a and des-ArgC3a alone did not chemoattract Meg in a transwell system, we observed that they did increase their

a c-

1

C3a 2 5

response to a low, ‘threshold’ dose of SDF-1 (10 ng/ml) (Figure 4a). In the presence of C3a or des-ArgC3a, the chemotactic response of Megs to a threshold dose of SDF-1 was comparable to that at a high dose of this chemokine (300 ng/ml). Interestingly, this priming effect of C3a on responsiveness of Megs to an SDF-1 gradient was not inhibited by the C3aR antagonist, SB290157, suggesting the involvement of a receptor other than C3aR that is probably activated both by C3a and des-ArgC3a. Next, as SDF-1 enhances adhesion of Megs to fibrinogen,16 we focused on the effect of C3 cleavage fragments on SDF-1dependent adhesion of Megs to fibrinogen (Figure 4b) and found that C3a and des-ArgC3a significantly enhanced SDF-1-dependent adhesion of Megs (Figure 4b). At the same time, adhesion of Megs was also slightly enhanced in the presence of C3a alone, but not des-ArgC3a. This C3a-induced adhesion of Megs correlated with the fact that C3a employed alone was able to enhance F-actin polymerization in these cells (Figure 3c). It is widely accepted that Megs produce platelets by extending proplatelet protrusions between endothelial cells,12 and it had been recently reported that SDF-1 enhances platelet formation by Megs that are in contact with endothelium.12 To address whether this process could be affected by C3a and des-ArgC3a, we evaluated platelet production in a transwell-migration assay.29 In this assay, transwell membranes were covered by confluent IL-1b activated HUVEC (Figure 4c), and megakaryocytic cells derived from day 12 expansion cultures were

desArgC3a

c-

10 min

1

2

5

10 min

phospho-MAPK42/44 total-MAPK42/44

b 1

2

3

FRACTION FROM TOP 4 5 6 7 8 9

10

11

CXCR4

CONTROL C3a desArgC3a

NO RAFT

RTFA

GM1

c

CXCR4

GM1

MERGE

CONTROL

C3a

desArgC3a

Figure 5 Western blot analysis of C3a- and des-ArgC3a-activated signaling pathways in Megs. (a) Both C3a (1 mg/ml) and des-ArgC3a (1 mg/ml) activate MAPK42/44. Experiments were performed three times with similar results. Representative blots are shown. (b) Analysis of the colocalization of CXCR4 in various fractions of cell membranes enriched in lipid rafts (fractions 3–5) and depleted of lipid rafts (fractions 9–11). Meg cell line MO7E cells were stimulated by C3a (1 mg/ml) or des-ArgC3a (1 mg/ml) or not stimulated (control). CXCR4 was detected in these membrane fractions by Western blot along with ganglioside M1 (GM1), a marker of lipid rafts. Experiments were performed three times with similar results. A representative blot is shown. (c) Lipid raft formation on Megs. Human Megs were cultured for 6 h in serum-free medium and than non-stimulated (control) or stimulated by C3a (1 mg/ml) or des-ArgC3a (1 mg/ml). The primary antibodies used for raft analysis are cholera toxin bsubunit conjugated with FITC and mouse monoclonal anti-hCXCR4 IgG. The stained cells are examined using a BX51 fluorescence microscope (Olympus America) equipped with a charge-coupled device camera (Olympus America). Separate pictures are merged using Image-Pro Plus software (Media Cybernetics Inc., Silver Spring, MD, USA). Lipid raft formation was analyzed on samples from three different G-CSF-mobilized patients. A representative study is shown. Colocalization of GM1 and CXCR4 is shown as yellow patchy staining. Leukemia


a

979 350

control 5mM MBCD

% of control

300

200 150

50

b

Having demonstrated that both C3a and des-ArgC3a enhances SDF-1-dependent migration and adhesion of megakaryocytic cells, we turned our attention to signaling pathways that could regulate these processes, for example, MAPK p42/44.9,16 First, we observed that, like SDF-1,16 C3a and des-ArgC3a induced phosphorylation of MAPK p42/44 in normal human Megs (Figure 5a). However, although maximal activation of MAPK p42/44 after stimulation by C3a occurred during first 1–2 min, des-ArgC3a activated MAPK p42/44 maximally B5 min latter. In our previous investigations on hematopoietic stem/progenitor cells, we reported that C3a and des-ArgC3a may enhance incorporation of CXCR4 into membrane lipid rafts,20,23 thereby allowing this receptor to form an optimal association with downstream signaling proteins. To determine whether a similar mechanism plays a role in megakaryocytic lineage cells, we evaluated the incorporation of CXCR4 into membrane lipid rafts in the human Meg line MO7E stimulated with C3a or des-ArgC3a (Figure 5b). As expected, Western blot analysis revealed that both molecules enhanced incorporation of CXCR4 into membrane lipid rafts. Finally, direct confocal analysis of normal ex vivo expanded human Megs confirmed that C3a and des-ArgC3a enhance incorporation of CXCR4 into membrane lipid rafts in these cells (Figure 5c).

Platelet formation is lipid raft dependent C3 cleavage fragments enhance incorporation of CXCR4 into membrane lipid rafts allowing optimal SDF-1 signaling by this receptor.18,20 Thus, we determined whether perturbation of lipid raft formation (by depletion of membrane cholesterol) affects SDF-1-dependent platelet formation.18,20 When human Megs were preincubated for 1 h in 5 mM MbCD, which depletes cholesterol from cell membranes,20,23 we found that perturbation of lipid raft formation by MbCD significantly inhibited platelet formation dependent on both SDF-1 þ C3a and SDF-1 þ des-ArgC3a (Figure 6a). Finally, we obtained proof of our hypothesis that lipid rafts play an important role in platelet formation when we used an

SDF-1

C3a

BLEEDING

control

1800 1600 1400

SDF-1+ desArgC3a SDF-1+ SDF-1 C3a high desArgC3a

*

**

*

contol statin

*

1200 1000 800 600 400 200 0

Molecular analysis of signaling pathways activated by C3a and des-ArgC3a in normal megakaryoblasts

*

*

250

100

PTK (K/uL)

placed in the upper chambers; C3a, des-ArgC3a and SDF-1 together or alone were placed in the lower chambers. After 48 h, the number of platelets present in the lower chambers was evaluated by FACS after staining with CD41 antibodies Furthermore, in control experiments, platelets were evaluated both morphologically by microscopy and functionally by evaluating the increase in expression of CD62P after stimulation with thrombin as described.12 We found that C3a and des-ArgC3a, if combined with SDF-1, significantly enhanced platelet formation by human Megs (Figure 4c). As migration of Megs in the BM environment is dependent on the secretion of MMP-916,29 and VEGF-mediated interaction with endothelium in the BM endothelial niche,12 we examined the influence of C3a and des-ArgC3a on MMP-9 and VEGF expression in these cells. We found that C3 cleavage fragments alone only slightly upregulated the expression of MMP-9 in normal megakaryocytic cells; however, when C3a or des-ArgC3a were used together with SDF-1, expression of MMP-9 was upregulated in a synergistic manner B 9-fold (data not shown). Furthermore, C3a and des-ArgC3a increased the production of VEGF by human Megs fivefold ( þ /1) and sixfold ( þ /2), respectively (data not shown).

5

10

15

20

25 30 35 40 days of feeding

45

50

55

60

Figure 6 Platelet secretion by Megs is lipid raft-dependent. (a) Platelet production in a trans-endothelial migration assay by normal human Megs (control) and Megs that were preincubated for 1 h with 5 mM of methyl-b-cyclodextran (MBCD). Platelets were identified and analyzed by FACS. Data are pooled from quadruplicate samples from three independent experiments. *Po0.0001. (b) Mice were fed with statins (Lipitor – atorvastatin calcium) or not (control) for 24 days. As shown Lipitor (750 mg/mouse/day) ameliorates post-bleeding thrombocytosis. Platelet counts in peripheral blood were assayed by employing an automatic analyzer (Hemavet). The data represent the average of three independent experiments (6–8 mice were pooled for each experiment). Data are mean7s.d. *Po0.00001 versus control.

in vivo model of excessive bleeding in mice which had received high doses of statins for 25 days before bleeding and found a significant decrease in platelet counts in response to acute bleeding as a result of the high statin diet (Figure 6b).

Discussion Mounting evidence suggests that complement plays a role in hematopoiesis, in particular under stressed situations where it acts to enhance the responsiveness of hematopoietic cells to an SDF-1 gradient.20,22,23 On the basis of previously published observations that SDF-1 is involved in Megs migration/maturation11–16 and platelet formation,29 and our data that C3-deficient mice as compared to wt littermates have a delayed recovery of platelet counts after sublethal irradiation or transplantation of bone marrow cells,23 we hypothesized that C3 cleavage fragments are implicated in stress-induced platelet production, as seen in reactive post-bleeding thrombocytosis. We found that human Megs express functional C3aR and confirmed its expression by platelets.30 Furthermore, we report here for the first time that C3 cleavage fragments appear to enhance SDF-1-dependent platelet production. It has also been postulated that platelets may play a role in activation and propagation of the C system.31 Thus, inflammation and thrombosis are two processes that are tightly linked together and these data point to the existence of a direct link between the complement system and platelets. Leukemia


980 Our findings also demonstrate that C3 cleavage fragments, as they do with other hematopoietic cells (e.g., CD34 þ cells), increase the responsiveness of Megs to an SDF-1 gradient.20,22,23 Megs similarly as CD34 þ cells in the presence of C3a or des-ArgC3a show more robust chemotaxis to SDF-1 and adhere better to fibrinogen. We also observed that as in other cell types, C3 cleavage fragments increase the incorporation of CXCR4 into membrane lipid rafts. CXCR4 þ cells respond much better to SDF-1 stimulation when incorporated into membrane lipid rafts,18,20 suggesting that the enhanced responsiveness of Megs to SDF-1 mediated by C3 cleavage fragments is membrane lipid raft-dependent. Lipid rafts have been reported to be essential to ADPmediated platelet activation21 and to play a role in the shedding of microparticles from platelets.32 Our current work suggests that the release of platelets from proplatelets is also lipid raftdependent. In fact, we have demonstrated that inhibition of lipid raft formation by depletion of cholesterol from cell membranes, using both methyl-b-cyclodextran (MBCD) and statins, inhibits platelet formation in vitro as well as in vivo, as shown in a model of thrombocytosis induced by excessive bleeding. This latter observation suggests that statins, drugs known to inhibit platelet activation,33 may also ameliorate stress-induced thrombocytosis. Supporting this contention is a recent report that a high dose of statins effectively reduced postoperative thrombocytosis in patients after coronary artery bypass graft-

stress

ing.26 Our data provide a molecular explanation for this effect and suggest that statins have potential use in preventing some forms of reactive thrombocytosis. Furthermore, as elevated platelet counts and levels of circulating platelet-derived microvesicles in blood are important risk factors in cancer,34–38 evidence suggesting that statins may lower stress-related platelet counts and inhibit shedding of prometastaic platelet-derived microparticles points to their potential use as new adjuvant drugs for the treatment of various malignancies.39 In fact, statins were recently successfully employed in an animal model to inhibit the metastasis of prostate cancer.40 We found that both C3a and des-ArgC3a activate MAPKp42/44 and enhance SDF-1-dependent production of MMP-9 and VEGF by human Megs, which are crucial for transendothelial migration of Megs and platelet release. Supporting this, recently published studies documented that the incubation of mature Megs with a synthetic MMP inhibitor, 5-phenyl-1,10-phenanthroline, resulted in the inhibition of platelet formation, suggesting that in fact the expression of MMPs is critical not only for megakaryocyte migration but also for platelet release.29 Similarly, an important step in the interaction of Megs with endothelial cells in the BM endothelial niche is secretion of VEGF by Megs,12 which in turn upregulates the expression of E-selectin on endothelium, a process crucial for the formation of proplatelets by these cells.12 Giving greater credence to this, we

Platelets

C3 cleavage

adhesion – interaction with endothelium

desArgC3a

C3a

Endothelium

SDF-1, VEGF

Meg – endothelial niche (Rafi S. J Exp Med. 1998) (Platelets shedding)

chemotaxis

Megakaryocyte chemotaxis

Meg – hematopoietic niche (Meg development) CFU - Meg

TPO

Megakaryoblast SDF-1

Osteoblasts/Stromal fibroblasts

Figure 7 C3a and des-ArgC3a modulate SDF-1-dependent attraction of Megs from bone niche to endothelial niche and platelet production. The concept of developmental allocation of Megs from the bone niche to the endothelial niche has been proposed by S. Rafii et al.17 This allocation is tightly regulated by an SDF-1 gradient and we envision that during stressed situations (e.g., chronic infection, hemorrhage), C3 cleavage fragments are important coregulators of this process. Accordingly, we propose that C3 is cleaved/activated in BM and as result of this C3 cleavage fragments (C3a and des-ArgC3a) modulate/enhance the responsiveness of maturing Megs to an SDF-1 gradient. As shown, Megs are chemoattracted by SDF-1 expressed by marrow endothelial cells and reallocate from the bone-marrow niche to the endothelial niche where proplatelets are formed and platelets shed from the Meg surface. At the molecular level, we envision that C3 cleavage fragments sensitize responsiveness of Megs to SDF-1 by enhancing its inclusion into membrane lipid rafts. CXCR4 if included into membrane lipid rafts regulates Megs migration more efficiently as well as adhesion to endothelium and secretion of MMP-9 and VEGF. All of these processes are crucial for proplatelet formation and platelet shedding and are enhanced by C3 cleavage fragments during reactive thrombocytosis. Leukemia


981 noticed that both C3 cleavage fragments increase expression of VEGF in human Megs. On the basis of a proposal by Rafii et al. regarding maturationrelated allocation of Megs progenitors from the bone niche into the endothelial niche,17 we postulate the following scenario for enhanced platelet production during excessive-bleeding (Figure 7), in which during stress situations (e.g., chronic infection, hemorrhage), C3 cleavage fragments are important coregulators of SDF-1–CXCR4 signaling. As shown in Figure 7, Megs are chemoattracted from the bone niche to the endothelial niche by SDF-1 expressed by marrow endothelial cells. In the endothelial niche, proplatelets are formed and subsequently shed platelets.17 We envision that C3 cleavage fragments sensitize the responsiveness of Megs to SDF-1 by enhancing its inclusion into membrane lipid rafts. Further, CXCR4 incorporated into membrane lipid rafts results in more efficient migration of Megs, adhesion to endothelium and secretion of MMP-9 and VEGF, all processes that are crucial for proplatelet formation and platelet shedding. In conclusion, both C3a and des-ArgC3a clearly have a role in modulating the responsiveness of Megs to an SDF-1 gradient. The crosstalk between C3aR- and CXCR4-G-protein-coupled receptors we identified here with respect to Megs plays an important and previously unrecognized role in stress and inflammation-dependent platelet formation. On the basis of this new information, we foresee that inhibitors of the CXCR4–SDF-1 axis have clinical potential as drugs to control certain forms of thrombocytosis (e.g., as described in postoperative thrombocytosis in patients after coronary bypass grafting).26 We are aware that statins show several pleiotropic effects on platelets (e.g., inhibit their activation). However, given that platelet formation is dependent on membrane lipid formation as we have shown here, statins, which perturb lipid raft formation, may ameliorate stress-related thrombocytosis and could be considered as potential drugs to prevent bleeding-related post-trauma or postoperative thrombocytosis.

Acknowledgements This work was supported by an NIH Grant R01 DK074720-01 to MZR.

References 1 Long MW, Hoffman R (eds). Hematology: Basic Principles and Practice. Churchill Livingstone: New York, 1999, p245. 2 Gewirtz AM. Megakaryopoiesis: state of the art. Thromb Haemost 1995; 74: 204–209. 3 Gurney AL, Carver-Moore K, de Sauvage FJ, Moore MW. Thrombocytopenia in c-mpl-deficient mice. Science 1994; 265: 1445–1447. 4 Gurney AL, de Sauvage FJ. Dissection of c-Mpl and thrombopoietin function: studies of knockout mice and receptor signal transduction. Stem Cells 1996; 14 (Suppl 1): 116–123. 5 Bunting S, Widmer R, Lipari T, Rangell L, Steinmetz H, CarverMoore K et al. Normal platelets and megakaryocytes are produced in vivo in the absence of thrombopoietin. Blood 1997; 90: 3423–3429. 6 Sasaki H, Hirabayashi Y, Ishibashi T, Inoue T, Matsuda M, Kai S et al. Effects of erythropoietin, IL-3, IL-6 and LIF on a murine megakaryoblastic cell line: growth enhancement and expression of receptor mRNAs. Leuk Res 1995; 19: 95–102. 7 Burstein SA, Mei RL, Henthorn J, Friese P, Turner K. Leukemia inhibitory factor and interleukin-11 promote maturation of murine and human megakaryocytes in vitro. J Cell Physiol 1992; 153: 305–312.

8 Neben TY, Loebelenz J, Hayes L, McCarthy K, Stoudemire J, Schaub R et al. Recombinant human interleukin-11 stimulates megakaryocytopoiesis and increases peripheral platelets in normal and splenectomized mice. Blood 1993; 81: 901–908. 9 Majka M, Ratajczak J, Villaire G, Kubiczek K, Marquez LA, Janowska-Wieczorek A et al. Thrombopoietin, but not cytokines binding to gp130 protein-coupled receptors, activates MAPKp42/ 44, AKT, and STAT proteins in normal human CD34+ cells, megakaryocytes, and platelets. Exp Hematol 2002; 30: 751–760. 10 Gainsford T, Nandurkar H, Metcalf D, Robb L, Begley CG, Alexander WS. The residual megakaryocyte and platelet production in c-mpl-deficient mice is not dependent on the actions of interleukin-6, interleukin-11, or leukemia inhibitory factor. Blood 2000; 95: 528–534. 11 Wang JF, Liu ZY, Groopman JE. The alpha-chemokine receptor CXCR4 is expressed on the megakaryocytic lineage from progenitor to platelets and modulates migration and adhesion. Blood 1998; 92: 756–764. 12 Hamada T, Mohle R, Hesselgesser J, Hoxie J, Nachman RL, Moore MA et al. Transendothelial migration of megakaryocytes in response to stromal cell-derived factor 1 (SDF-1) enhances platelet formation. J Exp Med 1998; 188: 539–548. 13 Kowalska MA, Ratajczak J, Hoxie J, Brass LF, Gewirtz A, Poncz M et al. Megakaryocyte precursors, megakaryocytes and platelets express the HIV co-receptor CXCR4 on their surface: determination of response to stromal-derived factor-1 by megakaryocytes and platelets. Br J Haematol 1999; 104: 220–229. 14 Riviere C, Subra F, Cohen-Solal K, Cordette-Lagarde V, Letestu R, Auclair C et al. Phenotypic and functional evidence for the expression of CXCR4 receptor during megakaryocytopoiesis. Blood 1999; 93: 1511–1523. 15 Hodohara K, Fujii N, Yamamoto N, Kaushansky K. Stromal cellderived factor-1 (SDF-1) acts together with thrombopoietin to enhance the development of megakaryocytic progenitor cells (CFU-MK). Blood 2000; 95: 769–775. 16 Majka M, Janowska-Wieczorek A, Ratajczak J, Kowalska MA, Vilaire G, Pan ZK et al. Stromal-derived factor 1 and thrombopoietin regulate distinct aspects of human megakaryopoiesis. Blood 2000; 96: 4142–4151. 17 Avecilla ST, Hattori K, Heissig B, Tejada R, Liao F, Shido K et al. Chemokine-mediated interaction of hematopoietic progenitors with the bone marrow vascular niche is required for thrombopoiesis. Nat Med 2004; 10: 64–71. 18 Nguyen DH, Taub D. CXCR4 function requires membrane cholesterol: implications for HIV infection. J Immunol 2002; 168: 4121–4126. 19 Gu Y, Filippi MD, Cancelas JA, Siefring JE, Williams EP, Jasti AC et al. Hematopoietic cell regulation by Rac1 and Rac2 guanosine triphosphatases. Science 2003; 302: 445–449. 20 Wysoczynski M, Reca R, Ratajczak J, Kucia M, Shirvaikar N, Honczarenko M et al. Incorporation of CXCR4 into membrane lipid rafts primes homing-related responses of hematopoietic stem/ progenitor cells to an SDF-1 gradient. Blood 2005; 105: 40–48. 21 Quinton TM, Kim S, Jin J, Kunapuli SP. Lipid rafts are required in Galpha(i) signaling downstream of the P2Y12 receptor during ADP-mediated platelet activation. J Thromb Haemost 2005; 3: 1036–1041. 22 Reca R, Mastellos D, Majka M, Marquez L, Ratajczak J, Franchini S et al. Functional receptor for C3a anaphylatoxin is expressed by normal hematopoietic stem/progenitor cells, and C3a enhances their homing-related responses to SDF-1. Blood 2003; 101: 3784–3793. 23 Ratajczak MZ, Reca R, Wysoczynski M, Kucia M, Baran JT, Allendorf DJ et al. Transplantation studies in C3-deficient animals reveal a novel role of the third complement component (C3) in engraftment of bone marrow cells. Leukemia 2004; 18: 1482– 1490. 24 Ratajczak MZ, Kucia M, Reca R, Majka M, Janowska-Wieczorek A, Ratajczak J. Stem cell plasticity revisited: CXCR4-positive cells expressing mRNA for early muscle, liver and neural cells ‘hide out’ in the bone marrow. Leukemia 2004; 18: 29–40. 25 Schafer AI. Thrombocytosis and thrombocythemia. Blood Rev 2001; 15: 159–166. 26 Christenson JT. Preoperative lipid-control with simvastatin reduces the risk of postoperative thrombocytosis and thrombotic complications following CABG. Eur J Cardiothorac Surg 1999; 15: 394–399. Leukemia


982 27 Goldfarb RD, Parrillo JE. Complement. Crit Care Med 2005; 33: 482–484. 28 Kalant D, Cain SA, Maslowska M, Sniderman AD, Cianflone K, Monk PN. The chemoattractant receptor-like protein C5L2 binds the C3a des-Arg77/acylation-stimulating protein. J Biol Chem 2003; 278: 11123–11129. 29 Lane WJ, Dias S, Hattori K, Heissig B, Choy M, Rabbany SY et al. Stromal-derived factor 1-induced megakaryocyte migration and platelet production is dependent on matrix metalloproteinases. Blood 2000; 96: 4152–4159. 30 Polley MJ, Nachman RL. Human platelet activation by C3a and C3a des-arg. J Exp Med 1983; 158: 603–615. 31 Del Conde I, Cruz MA, Zhang H, Lopez JA, Afshar-Kharghan V. Platelet activation leads to activation and propagation of the complement system. J Exp Med 2005; 201: 871–879. 32 Biro E, Akkerman JW, Hoek FJ, Gorter G, Pronk LM, Sturk A et al. The phospholipid composition and cholesterol content of plateletderived microparticles: a comparison with platelet membrane fractions. J Thromb Haemost 2005; 3: 2754–2763. 33 Colli S, Werba JP, Tremoli E. Statins in atherothrombosis. Semin Vasc Med 2004; 4: 407–415. 34 Shimada H, Oohira G, Okazumi S, Matsubara H, Nabeya Y, Hayashi H et al. Thrombocytosis associated with poor prognosis in

Leukemia

35 36

37

38

39 40

patients with esophageal carcinoma. J Am Coll Surg 2004; 198: 737–741. Aoe K, Hiraki A, Yamazaki K, Nakamura Y, Murakami T, Maeda T et al. Thrombocytosis as a useful prognostic indicator in patients with lung cancer. Respiration 2004; 71: 170–173. Janowska-Wieczorek A, Wysoczynski M, Kijowski J, MarquezCurtis L, Machalinski B, Ratajczak J et al. Microvesicles derived from activated platelets induce metastasis and angiogenesis in lung cancer. Int J Cancer 2005; 113: 752–760. Ratajczak J, Wysoczynski M, Hayek F, Janowska-Wieczorek A, Ratajczak MZ. Membrane-derived microvesicles (MV): important and underappreciated mediators of cell to cell communication. Leukemia 2006; 20: 1487–1495. Ratajczak J, Miekus K, Kucia M, Zhang J, Reca R, Dvorak P et al. Embryonic stem cell-derived microvesicles reprogram hematopoietic progenitors: evidence for horizontal transfer of mRNA and protein delivery. Leukemia 2006; 20: 847–856. Demierre MF, Higgins PD, Gruber SB, Hawk E, Lippman SM. Statins and cancer prevention. Nat Rev Cancer 2005; 5: 930–942. Zhuang L, Kim J, Adam RM, Solomon KR, Freeman MR. Cholesterol targeting alters lipid raft composition and cell survival in prostate cancer cells and xenografts. J Clin Invest 2005; 115: 959–968.