Coagulation Cascade Activation Causes CC Chemokine Receptor-2 ...

Coagulation Cascade Activation Causes CC Chemokine Receptor-2 Gene Expression and Mononuclear Cell Activation in Hemodialysis Patients Giovanni Pertosa,* Simona Simone,* Michela Soccio,* Daniela Marrone,* Loreto Gesualdo,† Francesco Paolo Schena,* and Giuseppe Grandaliano* *Division of Nephrology, Department of Emergency and Transplantation, University of Bari, Policlinico, Bari; and † Division of Nephrology, Department of Biomedical Sciences, University of Foggia, Policlinico, Foggia, Italy Priming of the coagulation cascade during hemodialysis (HD) leads to the release of activated factor X (FXa). The binding of FXa to its specific receptors, effector protease receptor-1 (EPR-1) and protease-activated receptor-2 (PAR-2), may induce the activation of peripheral blood mononuclear cells (PBMC) and promote a chronic inflammatory state that is responsible for several HD-related morbidities. In the attempt to elucidate the mechanisms underlying the coagulation-associated inflammation in HD, 10 HD patients were randomized to be treated subsequently with a cellulose acetate membrane (CA) and Ethylen-vinyl-alcohol (EVAL), a synthetic membrane that has been shown to reduce FXa generation. At the end of each experimental period, surface FXa and thrombin receptors (EPR-1 and PAR-1, -2, and -4) and CCR2 (monocyte chemoattractant protein-1 receptor) gene expression in isolated PBMC were examined. the ability of dialytic membranes to activate proteintyrosine kinases and the stress-activated kinase JNK and to modulate the generation of terminal complement complex (TCC) was also investigated. EPR-1 and PAR-2 and -4 mRNA expression, barely detectable in normal PBMC, were significantly upregulated in HD patients, particularly in those who were treated with CA. A striking increase of tyrosine-phosphorylated proteins and JNK activation was observed at the end of HD only in CA-treated patients. Simultaneously, an increased gene expression for both splicing isoforms of CCR2, A and B, only in PBMC from CA-treated patients was demonstrated. The increased CCR-2 mRNA abundance was followed by a significant increase in its protein synthesis. The high expression of CCR2 was associated with an increased generation of plasma TCC and a significant drop in leukocyte and monocyte count. By contrast, EVAL treatment slightly lowered TCC generation and normalized leukocyte count. In vitro FXa induced CCR2 A and B expression and JNK activation in freshly isolated PBMC. FXa-induced CCR2 mRNA expression was completely abolished by JNK and tyrosine kinase inhibition. In conclusion, these data suggest that subclinical clotting activation may cause an increased CCR2 gene and protein expression on uremic PBMC, contributing to HD-related chronic microinflammation. The use of the less coagulation-activating membrane, EVAL, may reduce PBMC activation through the modulation of the stress-activated kinase JNK. J Am Soc Nephrol 16: 2477–2486, 2005. doi: 10.1681/ASN.2004070621

B

lood interaction with cellulosic membranes during hemodialysis (HD) may cause the activation of several biologic systems, including complement and coagulation cascades (1). Despite clinically effective anticoagulation, priming of the clotting system commonly occurs during HD, leading to the release of several proteases, including thrombin and activated factor X (FXa) (2). FXa has been shown to induce cell proliferation (3), PDGF-A and B chain gene expression (4), and IL-6 release in several cell lines (5). Altieri et al. (6) identified effector protease receptor-1 (EPR-1) as the main FXa receptor on leukocytes and endothelial cells. However, recent obser-

Received July 30, 2004. Accepted May 12, 2005. Published online ahead of print. Publication date available at www.jasn.org. Address correspondence to: Dr. Giuseppe Grandaliano, Division of Nephrology, Department of Emergency and Transplantation, University of Bari, Policlinico, Piazza Giulio Cesare 11, Bari 70124, Italy. Phone: ⫹39-080-5592787; Fax: ⫹39-0805593227; E-mail: [email protected] Copyright © 2005 by the American Society of Nephrology

vations support the hypothesis that FXa may induce cell activation primarily interacting with protease-activated receptor-2 (PAR-2). PAR-2 belongs to a growing family of G protein– coupled receptors activated by limited proteolysis (7–9). Although Vu et al. (10) characterized the first of these receptors in the early 1990s, very little is known about the effect of HD on PAR expression in peripheral blood mononuclear cells (PBMC). Leukocyte activation during HD is the central event in the pathogenesis of bioincompatibility, but the mechanisms of this phenomenon remain unclear. Monocyte chemoattractant protein-1 (MCP-1), a member of the CC family of chemokines, has been suggested to play a pivotal role in HD-induced mononuclear cell activation (11,12). We and others (13,14) reported that thrombin may induce MCP-1 expression in different cell lines, suggesting a close relationship between coagulation and inflammatory response. MCP-1 effects are mediated by a specific cell-surface receptor, CC chemokine receptor 2 (CCR2), mainly expressed in monocytes, basophils, and certain subsets of T cells (15,16). There are several pieces of experimental evidence ISSN: 1046-6673/1608-2477

2478

Journal of the American Society of Nephrology

that CCR-2 may play a key role in the pathogenesis of atherosclerosis (17,18). Although the potential inflammatory effects of clotting proteases now are clear, the question of whether a reduced priming of coagulation cascade may improve HDinduced PBMC activation is still unanswered. To address this issue, we evaluated in vivo CCR2 gene expression and PBMC activation in uremic patients who were treated with low-flux cellulose acetate (CA) or Ethylen-vinyl-alcohol (EVAL) membrane, a synthetic membrane characterized by a reduced thrombogenic activity (19).

Materials and Methods Patients Ten stable HD patients (five men and five women; mean age 45.3; range 25 to 67 yr), having given their informed consent, were enrolled in the study; 10 healthy subjects (five men and five women; mean age 43.2; range 25 to 62 yr) who were matched for gender and age represented the control group. Patients were stabilized on renal replacement therapy for ⬎6 mo before the study (mean dialytic age 24.2; range 12 to 55 mo) and were treated by bicarbonate HD using a CA dialyzer for a mean of 4.0 h thrice weekly when entering the study. None of them had signs of infection, active immunologic processes, or malignancy at the time of the study. Underlying diseases leading to end-stage renal failure were chronic glomerulonephritis (five patients), interstitial nephritis (three patients), and cystic disease (two patients).

Study Design All HD patients, in a randomized manner, were treated for two subsequent periods of 3 mo either with a CA hollow-fiber dialyzer (CA 180; Althin, Milan, Italy; membrane surface 1.4 m2) or with EVAL membrane (KF101–1.6; Kuraray, Tokyo, Japan; membrane surface 1.6 m2). After 36 sequential treatments, the patients were switched from CA to EVAL and vice versa for another 36 treatments. HD efficiency, as indicated by urea reduction rate, remained unchanged during the study periods. Dialyzers were not reused. Endotoxin content of the dialysate, as shown by colorimetric Limulus Amebocyte Lysate assay (Coatest Kabi Vitrum, Stockholm, Sweden), was constantly ⬍0.05 EU/ml.

Isolation of Human PBMC and Western Blot Analysis At the end of each treatment period with CA or EVAL, blood samples (25 ml) were drawn in sterile, heparinized vacuum tubes from the patients’ arteriovenous fistulae just before the second HD session of the week (T0) and subsequently from the efferent line of the dialyzer after 15 (T15) and 180 min (T180) of the same dialysis session. PBMC were separated on a Ficoll/Hypaque gradient (Pharmacia, Uppsala, Sweden). Cells were washed with 0.9% NaCl solution, and remaining erythrocytes were destroyed by hypotonic lysis using 0.2% and 1.6% NaCl solutions. Isolated PBMC were lysed with RIPA buffer (1 mM PMSF, 5 mM EDTA, 1 mM sodium orthovanadate, 150 mM sodium chloride, 8 ␮g/ml leupeptin, 1.5% Nonidet P-40, and 20 mM tris-HCl [pH 7.4]) and centrifuged at 10,000 ⫻ g for 30 min at 4°C. The lysate supernatant was collected, and its protein content was determined by the Bradford method (Bio-Rad, Hercules, CA). Twenty micrograms of protein from each lysate was mixed with SDS sample buffer, and aliquots were separated by SDS-PAGE (7.5% acrylamide) and electrophoretically transferred onto a nitrocellulose filter. The filter was blocked overnight at room temperature with 2% BSA in PBS that contained 0.1% Tween-20 (TBS) and incubated with anti-phosphotyrosine (Santa Cruz Biotechnology, Santa Cruz, CA), anti–phospho-JNK, anti-JNK (Santa Cruz Biotechnology), anti-CCR2 (R&D, Abingdon,

J Am Soc Nephrol 16: 2477–2486, 2005

UK), or anti–PAR-2 (Zymed, San Francisco, CA) mAb at room temperature for 4 h. The membranes were washed twice in TBS and incubated for 2 h at room temperature with horseradish peroxidase– conjugated sheep anti-mouse IgG at 1:1500 dilution in TBS. The membranes were washed three times at room temperature in TBS and then once with 0.1% SDS in PBS. The ECL enhanced chemiluminescence system (Amersham, Buckinghamshire, UK) was used for detection.

RNA Extraction and Reverse-Transcription Polymerase Chain Reaction (RT-PCR) Total RNA was isolated from freshly isolated PBMC by the singlestep method, using phenol and chloroform/isoamylalcohol. One microgram of total RNA was used in a reverse transcription (RT) reaction. Twenty microliters of the reaction mixture that contained 1 ␮g of total RNA, PCR buffer (10 mM Tris-HCl [pH 8.3] and 50 mM KCl), 5 mM MgCl2, 1 mM dNTPs, 20 U of RNase inhibitor, 2.5 mM oligo (dT), and 100 U of Moloney murine leukemia virus reverse transcriptase was incubated at 42°C for 30 min and then heated to 95°C for 5 min to inactivate the enzyme activity and to denature RNA-cDNA hybrids. PCR was performed with separate sets of oligonucleotide primers, specific for each gene studied. For CCR2, we used two sets of primers that recognize the two splicing isoforms of the receptor, CCR2A and CCR2B (16):EPR-1 5⬘-TTACGCCAGACTTCAGCCTG-3⬘, 5⬘-TGGGTAACAGTGGCTGCTTC-3⬘; PAR-1 5⬘-CTCAGAGGCTGTAGAACTAAGCAG-3⬘, 5⬘-GGTACACTTCCATAGACTTAGGCC-3⬘; PAR-2 5⬘-TGGGTTTGCCAAGTAACGGC-3⬘, 5⬘-GGGAGATGCCAATGGCAATG-3⬘; PAR-4 5⬘-CTACGACGAGAGCGGGAGCAC-3⬘, 5⬘-CAGCGCCAGGGCCAGCAGGAGGTC-3⬘; glyceraldehyde-3-phosphate dehydrogenase (GAPDH) 5⬘-TGGTATCGTGGAAGGACTCATGAC-3⬘, 5⬘-ATGCCAGTGAGCTTCCCGTTCAGC-3⬘; CCR2 A 5⬘-AAGAGGCATAGGGCAGTGAGA-3⬘, 5⬘AGCCTTCCTGCCTGGTAACGTAG-3⬘; and CCR2B 5⬘-CTCCCATTGTGGGCTCACTCTGCTGCAA-3⬘, 5⬘-TTCCTGGGCACCTGCTTTACAGG-3⬘. EPR-1, PAR-1, PAR-2, PAR-4, and GAPDH cDNA amplification was performed in two separate sets of reactions at a final concentration of 1⫻ PCR buffer, 1.5 mM MgCl2, 200 ␮M dNTPs, 0.15 ␮M PAR-2 primers or 0.15 ␮M GAPDH primers, and 1.25 U of AmpliTaq DNA polymerase (Perkin Elmer, Monza, Italy) in a total volume of 50 ␮l. The amplification profile involved denaturation at 95°C for 1 min, primer annealing at 61°C for PAR-2 or at 55°C for GAPDH for 1 min, and extension at 72°C for 1 min. PCR products were electrophoresed in 1.6% agarose gel in Tris borate/EDTA buffer, loading 10 ␮l of either PAR-2 or GAPDH PCR products for each sample. Intensity of the PAR-2 and GAPDH bands was quantified by computer-assisted densitometry (Optilab 2.6.1; Graftek, Villanterio, Italy). Results were expressed as PAR-2 to GAPDH ratios.

Assay for Terminal Complement Complex EDTA anticoagulated samples were drawn before dialysis (T0) and after 15 min (T15), 180 min (T180), and 480 min (T480) of HD. The specimens were collected aseptically, centrifuged at 4°C, and stored at ⫺70°C until processed. Detection of terminal complement complex (TCC) was based on a double-sandwich ELISA system using a specific antibody for the activated complement component (Quidel, San Diego, CA). Mean normal value was 0.16 ⫾ 0.07 ␮g/ml.

In Vitro Study Twenty milliliters of heparinized whole blood, drawn from healthy subjects, were used to isolate PBMC as described above. The cells were resuspended at 106/ml, preincubated in serum-free RPMI overnight, and then exposed to FXa (10 nM; Calbiochem, La Jolla, CA). At the

J Am Soc Nephrol 16: 2477–2486, 2005

indicated time points, the PBMC were lysed in RIPA buffer as described previously and used for Western blot studies. In a separate set of experiments, freshly isolated PBMC were resuspended in serum-free RPMI at 106 cells/ml, stimulated with FXa for the indicated time period, and lysed in guanidium isothiocyanate. Total RNA then was extracted by the single-step method and used for RT-PCR as described above. Experiments were performed with blood from three different healthy subjects.

Statistical Analyses All results are given as mean ⫾ SD. Significance was assessed using paired or unpaired t test as well as repeated measures ANOVA, as appropriate. Differences were considered to be significant at P ⬍ 0.05.

Results Effect of Dialytic Membranes on PBMC Activation and CCR2 Gene Expression The first step of the study was to investigate the ability of the two membranes to activate circulating mononuclear cells. To this purpose, we evaluated whether dialysis treatment with either CA or EVAL may induce the activation of one or more

Coagulation Priming and CCR2 Expression in HD

2479

cytoplasmic tyrosine kinases. In addition, we investigated the phosphorylation and subsequent activation of JNK, a stressactivated kinase that plays a key role in several leukocyte functions. In vivo interaction of PBMC with CA membrane induced an increase in cytoplasmic levels of tyrosine-phosphorylated proteins, an indirect measure of tyrosine kinase activation, that peaked at the end of the dialysis session (Figure 1, A and B). During the EVAL treatment, however, the low levels of tyrosine phosphorylated proteins observed at T0 remained unchanged throughout the dialysis session, with only a slight increase at the end of HD (Figure 1, A and B). Simultaneously, at T180, we observed a significant increased phosphorylation of both JNK isoforms in CA-treated patients but not in those who underwent dialysis with EVAL (Figure 1, C and D). To investigate whether the activation of intracellular signaling pathways influenced the phenotype of circulating mononuclear cells, we evaluated by RT-PCR the expression of CCR2, the specific MCP-1 receptor. Because there are two splicing isoforms of CCR2, A and B, which have been hypothesized to

Figure 1. Modulation of peripheral blood mononuclear cell (PBMC) protein tyrosine phosphorylation (A and B) and JNK phosphorylation (C and D) by dialysis with cellulose acetate (CA) and ethylen-vinyl-alcohol (EVAL) membrane. PBMC were isolated and lysed in RIPA buffer at T0 and T180. Tyrosine-phosphorylated protein and JNK phosphorylated levels were studied by Western blotting using a specific monoclonal anti-phosphotyrosine and anti–phospho-JNK antibody, respectively, as described in Materials and Methods. The anti–phospho-JNK blots were stripped and reprobed with mAb that recognize the total form of the enzyme. All gels shown are from the same patient and are representative of 10 patients. Quantification of the total tyrosine phosphorylated protein as well as the phospho-JNK/total JNK ratio are provided in B and D, respectively. *P ⬍ 0.001 versus EVAL membrane.

2480


J Am Soc Nephrol 16: 2477–2486, 2005

play a different role in cell activation, we used two sets of primers that are able to recognize CCR2 A and B (16). A weak expression of both isoforms was observed in normal PBMC (Figure 2A). By contrast, PBMC from CA-treated uremic patients presented an increased mRNA expression for the CCR2A at the beginning of and during the dialysis session (Figure 2, B and C). Conversely, CCR2A gene expression was barely detectable in PBMC from EVAL-treated patients (Figure 2, B and C). Similarly, at all times of dialysis, the CCR2B gene expression was markedly upregulated in patients who were treated with CA. In contrast, PBMC from patients who underwent dialysis with EVAL membrane showed a significantly lower CCR2B gene expression (Figure 2, B and C). To confirm that the increased gene expression for CCR-2 did correspond to an increased protein synthesis, we investigated the protein levels of this receptor in PBMC by Western blotting. As shown in Figure 3, CCR2 protein expression mirrored the changes in CCR2 mRNA abundance observed at the different time points with the two membranes. Finally, the changes (%) in leukocyte number during HD were taken as an additional parameter of membrane biocompatibility. Leukocyte drop (expressed as percentage of initial white blood cell count) was most pronounced after 15 min of HD, particularly when CA membrane was used (47% of pre-

Figure 2. CCR2 gene expression in PBMC that were isolated from normal subjects (A) and from CA- and EVAL-treated HD patients (B). Freshly isolated PBMC were lysed in guanidinium isothiocyanate, and total RNA was obtained by the single-step method. Monocyte chemoattractant protein-1 (MCP-1) receptor gene expression was investigated by reverse transcription–PCR (RT-PCR) using two different sets of primers that specifically recognize the two splicing variants of CCR2 as described in Materials and Methods. All gels shown are from the same patient and are representative of six healthy subjects and 10 HD patients. Quantification of CCR2/glyceraldehyde-3-phosphate dehydrogenase (GAPDH) ratio is provided in C. *P ⬍ 0.001 versus EVAL membrane.

Figure 3. CCR2 protein expression in PBMC that were isolated from normal subjects and from CA- and EVAL-treated patients (A). PBMC were isolated and lysed in RIPA buffer at T0 and T180. CCR2 protein levels were studied by Western blotting using a specific mAb as described in Materials and Methods. All gels shown are from the same patient and are representative of 10 patients. Quantification of CCR2 protein expression is provided in B. *P ⬍ 0.001 versus EVAL membrane.

J Am Soc Nephrol 16: 2477–2486, 2005

Figure 4. Leukocyte count during dialysis with CA and EVAL membrane. Whole blood samples were drawn at the indicated time points, and total leukocytes were counted as described in Materials and Methods. *P ⬍ 0.01 versus CA.

treatment levels), whereas lower leukocyte change was recorded with EVAL membrane (64%; Figure 4). In addition, at the same time, HD monocytopenia was prominent with CA membranes only (data not shown). A return to pretreatment levels was observed by the end of the third and fourth hours of HD.

Effect of CA and EVAL Membranes on C5b-9 Generation The contact between blood constituents and bioincompatible membranes triggers leukocyte activation and migration into the


2481

Figure 5. C5b-9 generation during hemodialysis (HD) with CA and EVAL membrane. Plasma concentrations of C5b-9 were measured by ELISA as described in Materials and Methods. The dotted line represents the mean of C5b-9 concentration in the control group (0.16 ⫾ 0.07 ␮g/ml). *P ⬍ 0.02 versus EVAL membrane.

pulmonary capillaries (20,21). Part of these events is thought to be a consequence of dialyzer-associated activation of the complement system through the alternative pathway (21). Recently, the terminal complement complex C5b-9 was suggested to modulate several cell functions in a variety of cell types (22,23). Thus, at the end of each experimental period, we first evaluated the ability of the different membranes to generate C5b-9 complex. At the beginning of HD, a significantly higher level of

Figure 6. Effector protease receptor-1 (EPR-1; A) and protease-activated receptor-2 (PAR-2; C) gene expression in healthy subjects (Control) and uremic patients who underwent HD with either CA or EVAL. Freshly isolated PBMC that were obtained from healthy subjects and HD patients then were lysed in guanidium isothiocyanate. EPR-1 and PAR-2 gene expression was evaluated by RT-PCR and normalized to GAPDH as described in Materials and Methods. All gels shown are from the same patient and are representative of six healthy subjects and 10 HD patients. Quantification of EPR-1/GAPDH and PAR-2/GAPDH ratios are provided in B and D, respectively. *P ⬍ 0.001 versus EVAL membrane.

2482


J Am Soc Nephrol 16: 2477–2486, 2005

plasma C5b-9 was found in CA-treated (0.22 ⫾ 0.07 ␮g/ml) as compared with EVAL-treated patients (0.15 ⫾ 0.02 ␮g/L; P ⬍ 0.02; normal value 0.16 ⫾ 0.07 ␮g/ml). C5b-9 generation was significantly increased at T15 and T180 by both types of membranes and returned to baseline values only in EVAL-treated patients (Figure 5).

Gene Expression of Specific Coagulation Factor Receptors on Normal and Uremic PBMC

Figure 7. PAR-2 protein expression in PBMC that were isolated from normal subjects and from CA- and EVAL-treated patients (A). PBMC were isolated and lysed in RIPA buffer at T0 and T180. PAR-2 protein levels were studied by Western blotting using a specific mAb as described in Materials and Methods. All gels shown are from the same patient and are representative of 10 patients. Quantification of PAR-2 protein levels is provided in B. *P ⬍ 0.001 versus EVAL membrane.

Because EVAL has been shown to reduce the activation of the coagulation system, we hypothesized that the improved biocompatibility of this dialytic membrane may be due to a reduced generation of coagulation factors. Indeed, it is well established that at least two of these serine protease, FXa and thrombin, may modulate mononuclear cell activation through the interaction of several specific cell surface receptors. Thus, we investigated whether the expression levels of these receptors were modulated by the uremic state and/or by dialysis treatment with a specific membrane. The expression of EPR-1, the FXa receptor identified by Altieri et al. (6), was comparable between normal subjects and HD patients at T0 (Figure 6, A and B). However, EPR-1 mRNA abundance was strikingly increased at T180 only in patients who were treated with CA. The prolonged use of EVAL significantly reduced EPR-1 gene expression at T180 (Figure 6, A and B). PAR-2 was highly expressed in normal PBMC (Figure 6, C and D). PAR-2 mRNA levels were increased in PBMC from CA, in particular at T180, whereas in EVAL-treated patients, both basal and T180 gene

Figure 8. PAR-4 (A) and PAR-1 (C) in healthy subjects (Controls) and uremic patients who underwent HD with either CA or EVAL. Freshly isolated PBMC that were obtained from healthy subjects and HD patients then were lysed in guanidium isothiocyanate. PAR-4 and PAR-1 gene expression was evaluated by RT-PCR and normalized to GAPDH as described in Materials and Methods. All gels shown are from the same patient and are representative of six healthy subjects and 10 HD patients. Quantification of PAR-4/GAPDH and PAR-1/GAPDH ratios are provided in B and D, respectively. *P ⬍ 0.001 versus EVAL membrane.

J Am Soc Nephrol 16: 2477–2486, 2005

expression was significantly reduced (Figure 6, C and D). Although PAR-2 protein expression at T0 mirrored the mRNA levels, at T180 in CA-treated patients, it was dramatically reduced, suggesting a significant degradation of this receptor very likely as a result of an extensive proteolytic activation (Figure 7). PAR-4, the recently cloned thrombin receptor, was significantly increased only in CA-treated patients (Figure 8, A and B), whereas gene expression of PAR-1, the first identified thrombin receptor, was not modulated by different dialytic membranes or uremic state (Figure 8, C and D).


2483

levels of cytoplasmic tyrosine-phosphorylated proteins (data not shown) and a striking and time-dependent activation of JNK (Figure 10). It is interesting that inhibition of tyrosine kinases, as well as JNK, by genistein and curcumin, respectively, completely abolished FXa-induced CCR2B gene expression (Figure 11).

Discussion

To define the role of FXa and thrombin in bioincompatibility phenomena, we investigated in vitro the effects of these coagulation factors on CCR-2 gene expression in normal PBMC. FXa, at a concentration of 10 nM, induced a striking increase in CCR2A (data not shown) and CCR2B (Figure 9) gene expression, particularly after 6 and 24 h of incubation, whereas thrombin did not modify the mRNA levels of MCP-1 receptor (data not shown). In addition, FXa caused a significant increase in the

Priming of the coagulation cascade was the first bioincompatibility event to be considered in HD, but its role in the complex inflammatory process primed by the contact between mononuclear cells and dialytic membranes long has been neglected. Even if clinically efficient anticoagulation during a dialysis session is obtained, biochemical markers of platelet and coagulation cascade activation significantly increase during dialysis (2). Hofbauer et al. (24) demonstrated in a group of patients who were undergoing HD with polysulphone the presence of a fibrin network covering a large portion of the fiber surface and protruding into the fiber lumen. This observation suggests also that membranes that are considered highly bio-

Figure 9. Activated factor X (FXa)-induced CCR2 gene expression in normal PBMC (A). Freshly isolated PBMC that were obtained from healthy subjects were resuspended in serum-free RPMI at 106/ml and incubated with FXa (10 nM) for the indicated periods. CCR2 gene expression was evaluated by RT-PCR and normalized to GAPDH as described in Materials and Methods. Representative of three independent experiments. Quantification of CCR2/GAPDH ratio is provided in B. *P ⬍ 0.001 versus unstimulated.

Figure 10. FXa-induced JNK phosphorylation in normal PBMC (A). Freshly isolated PBMC that were obtained from healthy subjects were resuspended in serum-free RPMI at 106/ml and incubated with FXa (10 nM) for the indicated periods. The cells then were lysed in RIPA buffer. Cellular proteins were separated by PAGE, and JNK phosphorylation was evaluated by Western blotting as described in Materials and Methods. The anti–phosphoJNK blots were stripped and reprobed with mAb that recognize the total form of the enzyme. Representative of three independent experiments. Quantification of phospho-JNK/total JNK ratio is provided in B. *P ⬍ 0.001 versus unstimulated.

In Vitro Study of PBMC Activation

2484


Figure 11. The effect of tyrosine kinase and JNK inhibition on FXa-induced CCR2 gene expression in normal PBMC (A). Freshly isolated PBMC that were obtained from healthy subjects were resuspended in serum-free RPMI at 106/ml, incubated for 3 h with curcumin or genistein, and then stimulated with FXa (10 nM) for 6 h. The cells were lysed in guanidium isothiocyanate. Representative of three independent experiments. Quantification of CCR-2/GAPDH ratio is provided in B. *P ⬍ 0.001 versus unstimulated; **P ⬍ 0.001 versus FXa stimulated.

compatible and used worldwide for their ability to reduce complement and monocyte activation may still prime the coagulation cascade. EVAL is a synthetic membrane that has been shown to present a reduced coagulation priming (25). In this study, we demonstrated for the first time that dialysis treatment with this dialyzer compared with dialysis performed using a modified cellulosic membrane may reduce PBMC activation. Indeed, EVAL treatment blunted the activation of both tyrosine kinase(s) and JNK and the subsequent CCR2 gene expression strikingly induced by HD with CA membrane. It is interesting that tyrosine-phosphorylated protein levels are significantly elevated already at the beginning of HD in CAtreated patients, suggesting a prolonged activation of PBMC in this group of patients. Activated coagulation factors have a short half-life and are rapidly inactivated. However, thrombin and FXa bound within the fibrin clot may be protected by the effect of their circulating inhibitors (26). Therefore, it is conceivable that the fibrin net-

J Am Soc Nephrol 16: 2477–2486, 2005

work that covers the dialysis membrane may represent a reservoir for these proteases. In this study, we demonstrated that uremic PBMC that are exposed to EVAL are less prone to express specific FXa protease receptors. By contrast, the in vivo contact between PBMC and cellulosic membranes is followed by an increased expression of EPR-1 and PAR-2. EPR-1 is necessary to localize FXa in close proximity to the cellular membrane, where it then selectively cleaves and activates PAR-2. The latter behaves as the transducing receptor component, triggering the intracellular signaling pathways (7). The dramatic reduction of PAR-2 protein expression in CA-treated patients in the presence of high mRNA levels suggests that this receptor is degraded extensively during an HD session, most likely as the result of a significant proteolytic activation. Thus, CA-treated uremic patients should be significantly more exposed to the proinflammatory FXa cellular effects. FXa but not thrombin induced a significant increase of JNK phosphorylation and CCR2 gene expression in cultured normal PBMC. This is the first report that a coagulation factor can modulate the expression of the main MCP-1 receptor in circulating mononuclear cells. It is widely known that MCP-1 is a powerful and specific chemotactic and activating factor for monocytes, and we previously reported a striking increase of MCP-1 expression by PBMC during HD (11,12). MCP-1 and CCR2 play a key role in the development of atherosclerosis. Indeed, both MCP-1 and CCR2 knockout mice do not develop atherosclerotic lesions (18,27). Thus, a reduced CCR-2 expression may lower the chemotactic response to MCP-1 and likely prevent the recruitment of monocytes into vascular atherosclerotic lesions (18). It is conceivable that the activation of the coagulation cascade and the subsequent increase in CCR2 expression may be involved in another acute effect of bioincompatible HD: Pulmonary leukosequestration. Leukopenia after blood-membrane contact, as a result of a significant sequestration of leukocytes within the pulmonary capillary, is generally thought to be associated with complement system activation through the release of chemotactic fragments (28,29). This hypothesis would explain the migration and subsequent activation of granulocytes but does not fully account for the monocyte sequestration often observed during bioincompatible treatments (30). In our study, EVAL significantly reduced early monocytopenia when compared with CA treatment, despite a similar generation of C5b-9 in the first 15 min of HD. Considering the striking reduction in CCR2 expression by PBMC in EVAL-treated patients, we hypothesized that this chemokine receptor may play a role in the recruitment of monocytes in the lung microcirculation. This hypothesis assumes that lung microvascular endothelial cells express MCP-1 at a significant level during HD session. There is well-established evidence that HD patients present high circulating levels of a variety of proinflammatory cytokines, including IL-1, TNF-␣, and IL-6, that are widely known to induce MCP-1 gene expression in different microvascular endothelial cell lines (1,31,32). Thus, it is conceivable that lung microvascular endothelial cells in these patients secrete a significant amount of MCP-1. The molecular mechanisms underlying the modulation of CCR2 gene expression in leukocytes still are largely unclear,

J Am Soc Nephrol 16: 2477–2486, 2005

and there are no observations on the signaling pathways and/or transcription factors involved in this event. It is interesting that in our in vitro study, we demonstrated that tyrosine kinase as well as JNK inhibition completely abolished FXainduced CCR2 gene expression in normal PBMC, suggesting a role for the axis tyrosine kinase–stress-activated kinases in the modulation of this key proinflammatory molecule. The tyrosine kinase signaling module has a pivotal role in the regulation of both the innate and the acquired immune response (33–35). Transgenic mice with a constitutively active HCk, a member of the src family of tyrosine kinases expressed in myelomonocytic cells, spontaneously acquired a lung pathology characterized by extensive mononuclear cell infiltration within the lung parenchyma and alveolar airspaces and around blood vessels (36), a condition that can resemble HD-induced pulmonary leukosequestration. It is interesting that src-related cytoplasmic tyrosine kinases are widely known activators of JNK (37). Once activated, JNK, in turn, can phosphorylate jun and facilitate its interaction with fos to form AP-1. This transcription factor then may modulate the gene expression of a variety of proinflammatory cytokines, a hallmark of bioincompatibility (1). In conclusion, our data support the hypothesis that subclinical clotting activation may cause an increased CCR2 gene expression on uremic PBMC and contribute to HD-related chronic microinflammation. Thus, it is conceivable that the use of a less coagulation activating membrane such as EVAL might improve this condition, reducing specific FXa and MCP-1 cellsurface receptors. In addition, the central role of JNK in the upregulation of CCR2 by FXa may suggest this enzyme as a molecular target for therapeutic intervention in the treatment of HD-related systemic microinflammation. Finally, long-term prospective studies are warranted to confirm the link between the molecular mechanisms identified in our study and the HD-related cardiovascular complications.

Acknowledgments This study was supported by MIUR (PRIN 2003 to G.P., L.G., and G.G.) and Centro d’Eccellenza Genomica in Campo Biomedico ed Agrario. We acknowledge the technical and financial support of Kuraray (Japan) and Bieffe (Italy).

References 1. Gesualdo L, Pertosa G, Grandaliano G, Schena FP: Cytokines and bioincompatibility. Nephrol Dial Transplant 13: 1622–1626, 1998 2. Sagedal S, Hartmann A, Sundstrom K, Bjornsen S, Brosstad F: Anticoagulation intensity sufficient for haemodialysis does not prevent activation of coagulation and platelets. Nephrol Dial Transplant 16: 987–993, 2001 3. Gasic GP, Arenas CP, Gasic TB, Gasic GJ: Coagulation factors X, Xa and protein S as a potent mitogens of cultured aortic smooth muscle cells. Proc Natl Acad Sci U S A 89: 2317–2320, 1992 4. Ko FN, Yang YC, Huang SC, Ou JT: Coagulation factor Xa stimulates platelet-derived growth factor release and mitogenesis in cultured vascular smooth muscle cells of rat. J Clin Invest 98: 1493–1501, 1996


2485

5. Papapetropoulos A, Piccardoni P, Cirino G, Bucci M, Sorrentino R, Cicala C, Johnson K, Zachariou V, Sessa WC, Altieri DC: Hypotension and inflammatory cytokine gene expression triggered by factor Xa-nitric oxide signaling. Proc Natl Acad Sci U S A 95: 4738 – 4742, 1998 6. Altieri DC, Edgington TS. Identification of effector cell protease receptor-1: A leukocyte-distributed receptor for the serine protease factor Xa. J Immunol 145: 246 –253, 1990 7. Bono F, Schaeffer P, He´rault J-P, Michaux C, Nestor AL, Guillemot JC, Herbert JM: Factor Xa activates endothelial cells by a receptor cascade between EPR-1 and PAR-2. Arterioscler Thromb Vasc Biol 20: 107–112, 2000 8. Coughlin SR. Protease-activated receptors start a family. Proc Natl Acad Sci U S A 91: 9200 –9202, 1994 9. Nystedt S, Emilsson K, Wahlestedt C, Sundelin J: Molecular cloning of a potential proteinase activated receptor. Proc Natl Acad Sci U S A 91: 9208 –9212, 1994 10. Vu TK, Hung DT, Wheaton VI, Coughlin SR: Molecular cloning of a functional thrombin receptor reveals a novel proteolytic mechanism of receptor activation. Cell 64: 1057– 1068, 1991 11. Leonard EJ, Yoshimura T: Human monocyte chemoattractant protein-1 (MCP-1). Immunol Today 11: 97–101, 1990 12. Pertosa G, Grandaliano G, Gesualdo L, Ranieri E, Monno R, Schena FP: Interleukin-6, interleukin-8 and monocyte chemotactic peptide-1 gene expression and protein synthesis are independently modulated by hemodialysis membranes. Kidney Int 54: 570 –579, 1998 13. Grandaliano G, Valente AJ, Abboud HE: A novel biologic effect of thrombin: Stimulation of monocyte chemotactic peptide-1 production. J Exp Med 179: 1737–1741, 1994 14. Colotta F, Sciacca FL, Sironi M, Luini W, Rabiet MJ, Mantovani A: Expression of monocyte chemotactic protein-1 by monocytes and endothelial cells exposed to thrombin. Am J Pathol 144: 975–985, 1994 15. Charo IF, Myers SJ, Herman A, Franci C, Connolly AJ, Coughlin SR: Molecular cloning and functional expression of two monocyte chemoattractant protein 1 receptors reveals alternative splicing of the carboxyl-terminal tails. Proc Natl Acad Sci U S A 91: 2752–2756, 1994 16. Kurihara T, Bravo R: Cloning and functional expression of mCCR2, a murine receptor for the C-C chemokines JE and FIC. J Biol Chem 271: 11603–11607, 1996 17. Rayner K, Van Eersel S, Groot PH, Reape TJ: Localisation of mRNA for JE/MCP-1 and its receptor CCR2 in atherosclerotic lesions of the ApoE knockout mouse. J Vasc Res 37: 93–102, 2000 18. Boring L, Gosling J, Cleary M, Charo IF: Decreased lesion formation in CCR2⫺/⫺ mice reveals a role for chemokines in the initiation of atherosclerosis. Nature 394: 894 – 897, 1998 19. Ishii Y, Yano S, Kanai H, Maezawa A, Tsuchida A, Wakamatsu R, Naruse T: Evaluation of blood coagulation-fibrinolysis system in patients receiving chronic hemodialysis. Nephron 73: 407– 412, 1996 20. Pertosa G, Grandaliano G, Gesualdo L, Schena FP: Clinical relevance of cytokine production in hemodialysis. Kidney Int Suppl 76: S104 –S111, 2000 21. Pertosa G, Tarantino EA, Gesualdo L, Montinaro V, Schena FP: C5b-9 generation and cytokine production in hemodialysed patients. Kidney Int Suppl 43: S221–S225, 1993 22. Rus HG, Niculescu FI, Shin ML: Role of the C5b-9 comple-

2486

23.

24.

25.

26.

27.

28.

29.

30.


ment complex in cell cycle and apoptosis. Immunol Rev 180: 49 –55, 2001 Viedt C, Hansch GM, Brandes RP, Kubler W, Kreuzer J: The terminal complement complex C5b-9 stimulates interleukin-6 production in human smooth muscle cells through activation of transcription factors NF-kB and AP-1. FASEB J 14: 2370 –2372, 2000 Hofbauer R, Moser D, Frass M, Oberbauer R, Kaye AD, Wagner O, Kapiotis S, Druml W: Effect of anticoagulation on blood membrane interactions during hemodialysis. Kidney Int 56: 1578 –1583, 1999 Cicchetti T, Senatore RP, Frandina F, Ferrari S, Striano U, Milei M, Cosentino S: Dialysis treatment using an ethylene vinyl alcohol membrane and no anticoagulation for chronic uremic patients. Artif Organs 17: 816 – 819, 1993 Wiliner GD, Danitiz MP, Mudd MS, Hsiedh K-H, Fenton JW: Selective immobilization of thrombin by surfacebound fibrin. J Lab Clin Med 97: 403– 411, 1980 Gosling J, Slaymaker S, Gu L, Tseng S, Zlot CH, Young SG, Rollins BJ, Charo IF: MCP-1 deficiency reduces susceptibility to atherosclerosis in mice that overexpress human apolipoprotein B. J Clin Invest 103: 773–780, 1999 Grooteman MP, Nube MJ, van Houte AJ, Schoorl M, van Limbeek J: Granulocyte sequestration in dialysers: A comparative elution study of three different membranes. Nephrol Dial Transplant 10: 1859 –1864, 1995 Blomquist S, Malmros C, Martensson L, Thorne J: Absence of lung reactions after complement depletion during dialysis: An experimental study in pigs. Artif Organs 15: 397– 401, 1991 Sester U, Sester M, Heine G, Kaul H, Girndt M, Kohler H: Strong depletion of CD14(⫹)CD16(⫹) monocytes during

J Am Soc Nephrol 16: 2477–2486, 2005

31.

32.

33.

34.

35.

36.

37.

haemodialysis treatment. Nephrol Dial Transplant 16: 1402– 1408, 2001 Franscini N, Bachli EB, Blau N, Leikauf MS, Schaffner A, Schoedon G: Gene expression profiling of inflamed human endothelial cells and influence of activated protein C. Circulation 110: 2903–2909, 2004 Brown Z, Gerritsen ME, Carley WW, Strieter RM, Kunkel SL, Westwick J: Chemokine gene expression and secretion by cytokine-activated human microvascular endothelial cells. Differential regulation of monocyte chemoattractant protein-1 and interleukin-8 in response to interferongamma. Am J Pathol 145: 913–921, 1994 Suzuki T, Kono H, Hirose N, Okada M, Yamamoto T, Yamamoto K, Honda Z: Differential involvement of Src family kinases in Fc gamma receptor-mediated phagocytosis. J Immunol 165: 473– 482, 2000 Cannons JL, Schwartzberg PL: Fine-tuning lymphocyte regulation: What’s new with tyrosine kinases and phosphatases? Curr Opin Immunol 16: 296 –303, 2004 Zamoyska R, Basson A, Filby A, Legname G, Lovatt M, Seddon B: The influence of the src-family kinases, Lck and Fyn, on T cell differentiation, survival and activation. Immunol Rev 191: 107–118, 2003 Ernst M, Inglese M, Scholz GM, Harder KW, Clay FJ, Bozinovski S, Waring P, Darwiche R, Kay T, Sly P, Collins R, Turner D, Hibbs ML, Anderson GP, Dunn AR: Constitutive activation of the SRC family kinase Hck results in spontaneous pulmonary inflammation and an enhanced innate immune response. J Exp Med 196: 589 – 604, 2002 Tateno M, Nishida Y, Adachi-Yamada T: Regulation of JNK by Src during Drosophila development. Science 287: 324 –327, 2000