Skip Navigation


NDT Advance Access originally published online on November 29, 2007
Nephrology Dialysis Transplantation 2008 23(5):1673-1681; doi:10.1093/ndt/gfm804
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/5/1673    most recent
gfm804v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Zaza, G.
Right arrow Articles by Schena, F. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zaza, G.
Right arrow Articles by Schena, F. P.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author [2007]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org


Dialysis-related systemic microinflammation is associated with specific genomic patterns

Gianluigi Zaza1, Paola Pontrelli2, Giovanni Pertosa1, Simona Granata1, Michele Rossini1, Silvia Porreca1, Frank J. T. Staal3, Loreto Gesualdo2, Giuseppe Grandaliano1 and Francesco Paolo Schena1

1 Renal, Dialysis and Transplant Unit, Department of Emergency and Transplantation, University of Bari, Bari 2 Department of Biomedical Sciences, University of Foggia, Foggia, Italy 3 Department of Immunology, Erasmus Medical Center, Rotterdam, The Netherlands

Correspondence and offprint requests to: Francesco Paolo Schena, Division of Nephrology, Department of Emergency and Transplantation, University of Bari, Piazza Giulio Cesare 11, 70124 Bari, Italy. Tel: +39-080-5592237; Fax: +39-080-5575710; E-mail: fp.schena{at}nephro.uniba.it



   Abstract
Top
 Abstract
Introduction
 Material and methods
 Results
 Discussion
 References
 
Background. Although several reports have focused on the clinical importance of the systemic microinflammatory state in the uraemic population, the relationship between the activation of a specific transcriptome and the development of this condition is still not completely defined.

Methods. Thirty haemodialysis (HD), 30 peritoneal dialysis (PD) and 30 chronic kidney disease (CKD) patients were enrolled in our study. For all patients, serum C-reactive protein (CRP) and ferritin levels were determined. In addition, the expression level of 234 inflammatory responses and oxidative stress pathway genes was measured, using oligonucleotide microarray chips (HG-U133A, Affymetrix), in peripheral blood mononuclear cells of 24 randomly selected patients (8 HD, 8 PD and 8 CKD).

Results. HD patients demonstrated higher CRP and ferritin levels compared to PD and CKD patients (P < 0.001). Statistical analysis identified 10 genes able to discriminate CKD from HD and PD patients (FDR = 5%, P < 0.001) and significantly correlated to CRP levels. All together, these genes were able to predict inflammation with an accuracy of 87% (P < 0.001). Among the selected genes there were those encoding for key regulators of inflammation and oxidative stress (e.g. RELA, GSS). Interestingly, only three inflammatory genes (MIF, IL8RB and CXCL12) were still significantly associated with inflammation when included in a multivariate analysis. RT–PCR for RELA, MIF, CXCL12 and western blots for IL8RB and GSS, using 66 patients, validated the microarray results.

Conclusions. This study may help to better understand the physiopathology of the systemic inflammatory state in CKD and dialysis patients and to identify new target genes potentially useful for future bio-molecular studies and therapeutic approaches.

Keywords: haemodialysis; microarray; microinflammation; peritoneal dialysis



   Introduction
Top
 Abstract
 Introduction
Material and methods
 Results
 Discussion
 References
 
Chronic systemic inflammation is a common feature in patients with chronic kidney disease (CKD) undergoing dialysis treatment [1]. Inflammation plays a central role in the development of atherosclerosis [2,3], and it has been identified as an epidemiologically important risk factor for cardiovascular morbidity and mortality in dialysis patients [4,5].

Renal failure per se may possibly contribute to inflammation as a result of the accumulation of pro-inflammatory compounds [6]. In addition, during dialysis, the interaction of peripheral blood mononuclear cells (PBMC) with dialytic membranes causes their activation with consequent increased synthesis and release of pro-inflammatory cytokines [6–9]. Other mechanisms involved in the release of pro-inflammatory cytokines during haemodialysis (HD) include the generation of complement fragments as a result of plasma protein membrane contact and the backfiltration of contaminated dialysate to the blood compartment [10,11].

In addition, previous in vivo studies suggest that inflammatory activity and circulating levels of cytokines are lower during peritoneal dialysis (PD) as compared with HD treatment [12]. The main cause of this lower activation may be related to the absence of any contact between blood and foreign surfaces in PD, although bioincompatibility of dialytic fluids and frequent peritonitis episodes may represent significant causes of chronic inflammatory status in PD patients [13–15].

Interestingly, recent studies suggest that chronic inflammation and oxidative stress are strongly related. In CKD patients treated with dialysis, there is an imbalance between pro- and anti-oxidant activities resulting in high oxidative stress [16,17], immune system deregulation [18] and activation of intracellular pathways that may lead to clinical complications [19,20]. HD per se has been suggested to induce oxidative stress, with reactive oxygen species (ROS) being generated on the surface of dialysis membranes by the activation of polymorphonuclear leukocytes (PMN) [21,22]. Indeed, it has been well documented that even a single session of HD significantly decreases anti-oxidant levels [23]. On the other hand, bioincompatibility of PD solutions seems to play a central role in the increased ROS synthesis in PD patients. In this situation, the plasma oxidant scavenging system is overwhelmed as substantiated by a decrease in plasma levels of glutathione, vitamins C and E, and anti-oxidant enzymes in patients receiving chronic PD treatment [24].

The present study, utilizing an innovative genomic approach, was undertaken to select a set of genes highly associated to the chronic microinflammatory state in CKD and dialysis patients, and to identify new bio-molecular markers potentially useful as target for new therapeutic interventions.



   Material and methods
Top
 Abstract
 Introduction
 Material and methods
Results
 Discussion
 References
 
Patients
Sixty stable dialysis and 30 CKD patients, having given their informed consent, were enrolled in our study. The main demographic and clinical features are summarized in Table 1. Among the dialysis population, 30 patients had been treated with HD three times/week (4–5 h/session) using synthetic membrane dialyzers [n = 8, polyamide (Gambro, Italy); n = 12, polysulphone (Fresenius, Germany); n = 4, polymethylmethacrylate (Toray Industries, Japan) and n = 6, AN 69 (Hospal, Italy)] for a mean of 4.10 years (range: 1.16–17.10), and 30 patients had been treated with PD [n = 22, continuous ambulatory peritoneal dialysis (CAPD) using a bicarbonate buffer (Physioneal, Baxter, Chicago, IL, USA) and n = 8, automated peritoneal dialysis using a bicarbonate buffer (Physioneal, Baxter)] for a mean of 3.62 years (range 0.54–9.03). All HD patients were treated with low molecular weight heparin. The 30 CKD patients on conservative therapy were in K-DOQI stage IV–V (creatinine clearance 20.5 ± 2.7 ml/min). All patients suffering from diabetes, chronic lung diseases, neoplasm or inflammatory diseases, and patients receiving antibiotics, corticosteroids, nonsteroidal anti-inflammatory agents or any medications (ACEI, statins, tocopherols, acetylcysteine) and nutrients (soy, alcohol, nuts) known to influence inflammatory and oxidative stress pathways were excluded. No patients had symptomatic coronary artery diseases or a family history of premature cardiovascular diseases. We did not include any patient with smoking habit.


View this table:
[in this window]
[in a new window]

  		Table 1 Demographic and clinical features of all patients included in the study

 
Serum C-reactive protein (CRP) and ferritin levels were measured at three time points (at the beginning of the study, 3 and 6 months thereafter) in all patients included in the study using high-sensitivity immunonephelometric (Dade Behring, Marbung, Germany) and chemiluminescence (Pierce, Rockford, IL, USA) according to the manufacturer's protocol, respectively. For the analysis, we used the mean of the three CRP and ferritin time-point measurements in each subject. In HD patients, the samples were drawn before the second dialysis of the week. CRP normal range was 0–0.4 mg/dl and ferritin normal range was 10–300 ng/ml.

For microarray analysis, we studied 24 randomly selected patients (n = 8, HD; n = 8, PD and n = 8, CKD) (Table 1) and the results were then validated using classical molecular approaches on the entire cohort of patients. The study was carried out according to the Declaration of Helsinki and approved by our Institutional Ethic Review Board.

PBMC isolation
For all patients, 20 ml samples of whole blood were collected. PBMC were isolated by density separation over a Ficoll–Hypaque (Flow Laboratories, Irvine, UK) gradient (460 g for 30 min). PBMC were washed three times with PBS pH 7.4/1 mM EDTA (Sigma, Milan, Italy). Cells were then counted and their viability was determined by trypan blue exclusion (>90% PBMC were viable).

RNA extraction and gene expression profiling
Total RNA was isolated by guanidinium–phenol–chloroform extraction, using the Trizol reagent (Gibco BRL, Gaithersburg, MD, USA) from a minimum of 5 x 106 cryopreserved PBMC of 24 patients (n = 8 HD, n = 8 PD and n = 8 CKD). Total RNA integrity was assessed by electrophoresis using the Agilent 2100 Bioanalyzer (Agilent, Palo Alto, CA, USA). RNA was processed and hybridized to the GeneChip Human Genome U133A oligonucleotide microarray (Affymetrix, Santa Clara, CA, USA) containing 22 283 gene probe sets, representing 12 357 human genes and 3800 ESTs (Affymetrix; see the manufacturer's manual for detailed protocol). We used the default settings of Affymetrix Microarray Suite software version 5 to calculate scaled gene expression values.

Statistical analysis
ANOVA, t-test and Fisher's exact test were used to assess differences in clinical and demographic features among the three study groups. Pairwise comparison t-tests were used to assess differences in CRP and ferritin levels among the three study groups. Results were expressed as mean ± SD. A value of P < 0.05 was considered to be statistically significant.

We selected a total of 234/22 283 gene probe sets (corresponding to 167 genes) involved in oxidative stress and inflammatory response pathways according to Gene ontology (GO, http://www.geneontology.org). Gene expression values for the 234 gene probe sets, scaled to the target intensity of 2500, were log transformed. ANOVA, distinction calculation (DC) [25] and Kruskal–Wallis test were used to select probe sets discriminating the three groups of patients. False discovery rate (FDR) was estimated using an empirical Bayesian approach based on permutations (n = 500) and Storey's q-value [26]. This preliminary approach, based on the identification of statistically significant modulated genes among the three treatment groups, has been performed to restrict the number of variables (genes) to be used in further analysis. Spearman's rank test and a multiple regression analysis were performed to identify genes independently associated with CRP levels. R 2.0.1 statistical software was used to perform the above analyses. Principal component analysis (PCA) and hierarchical clustering were performed using Spotfire DecisionSite 9.0 (www.spotfire.com). To assess the biological relationships among genes, we used the Ingenuity software (IPA, Ingenuity System, Redwood City, CA, USA; http://www.ingenuity.com). IPA is a knowledge database generated from the peer-reviewed scientific publications that enables discovery, visualization and exploration of functional biological networks in gene expression data and delineates the functions most significant to those networks. The 16 differentially expressed probe sets identified by microarray data, as discussed below, were used for network analyses. Affymetrix probe set IDs were uploaded into IPA and queried against all other genes stored in the IPA knowledge database to generate a set of networks. Each Affymetrix probe set ID was mapped to its corresponding gene identifier in the IPA knowledge database. Probe sets representing genes having direct interactions with genes in the IPA knowledge database are called ‘focus’ genes, which were then used as a starting point for generating functional networks. Each generated network is assigned a score according to the number of differentially regulated focus genes in our dataset. These scores are derived from negative logarithm of the P indicative of the likelihood of focus genes being found together in a network due to random chance. Scores of 4 or higher have 99.9% confidence level of significance. The resulting networks were represented in graphic format and adapted for publication.

Real-time polymerase chain reaction (RT–PCR)
Total RNA (400 ng) for 66 patients (n = 22 HD, n = 22 PD and n = 22 CKD) was reverse transcribed with the Cloned AMV first-strand cDNA synthesis kit (Invitrogen, Milan, Italy)) using Oligo(dT)20 primers (50 µM) and following the manufacturer's instructions. Real-time PCR amplification reactions were performed in triplicate in 25 µl of final volume via SYBR Green chemistry on iCycler (Bio-Rad Laboratories, Hercules, CA, USA). PCR protocol was performed using QuantiTect Primer Assays (Qiagen, Basel, Switzerland) for RELA, MIF and CXCL12: 40 cycles of 1 min at 95°C, 55°C for 30 s and 72°C for 1 min. Universal master mix obtained from Invitrogen included all reagents. The β-actin gene amplification was used as a reference standard to normalize the target signal. Amplification specificity was controlled by a melting curve and the amount of mRNA target was evaluated using the comparative Ct method.

Western blot analysis
Isolated PBMC from 66 patients (n = 22 HD, n = 22 PD and n = 22 CKD) were lysed in RIPA buffer (1 mM phenylmethylsulphonylfluoride, 5 mM EDTA, 1 mM sodium orthovanadate, 150 mM sodium chloride, 8 µg/ml leupeptin, 1.5% NonidetP-40, 20 mM Tris–HCl, pH 7.4). The lysates were kept on ice for 30 min and centrifuged at 10 000 g at 4°C for 10 min. The supernatants were collected and stored at –80°C until used. Aliquots containing 40 µg of proteins from each lysate were subjected to SDS–PAGE on a 10% gel under reducing conditions and then electrotransferred onto nitrocellulose membrane (HybondTM, Amersham, UK). The filter was blocked overnight at room temperature (RT) with 5% milk powder to detect IL-8RB or with 2% BSA to detect glutathione synthetase (GSS), both diluted in PBS containing 0.1% Tween-20 (TBS). The blot was then incubated with either anti-IL8RB monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA; 1:1000 dilution in TBS at RT for 2 h) or anti-GSS polyclonal antibody (Santa Cruz Biotechnology; 1:1300 dilution in TBS at RT for 2 h).

The membranes were washed twice in TBS and incubated for one hour at RT with horseradish peroxidase-conjugated goat anti-mouse IgG (Santa Cruz Biotechnology) at 1:1200 dilution in TBS for IL8RB or with horseradish peroxidase-conjugated goat anti-rabbit IgG (Santa Cruz Biotechnology) at 1:1500 dilution in TBS for GSS. The same membranes were stripped and immunoblotted again with anti-actin monoclonal antibody (Santa Cruz Biotechnology) at 1:500 dilution in TBS at RT for 2 h. The membranes were washed three times at RT in TBS and then once with 0.1% SDS in PBS. The ECL-enhanced chemiluminescence system (Amersham, Buckinghamshire, UK) was used for detection. The membrane was visualized using a scanner EPSON Perfection 2580 Photo and quantified by Image J 1.34 Software (http://rsb.info.nih.gov/ij/). The intensity of bands corresponding to IL8RB and GSS protein was normalized to the actin signal.



   Results
Top
 Abstract
 Introduction
 Material and methods
 Results
Discussion
 References
 
Differences in CRP and ferritin level between PD, HD and CKD patients
HD patients presented significantly higher CRP and ferritin levels compared to PD patient and CKD patients. In addition, both CRP and ferritin levels were significantly higher in PD patients compared to CKD patients (Figure 1A and B).


Figure 1
View larger version (16K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 CRP (A) and ferritin (B) serum levels in HD (n = 30), PD (n = 30) and CKD (n = 30) patients. The horizontal lines indicate the median for each subgroup; the top and bottom of each box depict the 25th and 75th percentile value. The dashed lines depict 1.5 times the interquantile range. The P-value was determined by pairwise comparison using the t-test adjusted for multiple testing.

 
Differences in gene expression between the three groups of patients
To better investigate the molecular basis underlying the microinflammatory state in our patient population, we analysed the gene-expression profiling of PBMC isolated from 8 HD, 8 PD and 8 CKD patients. According to three independent statistical algorithms (ANOVA, Kruskal–Wallis test and distinction calculation) and the estimate FDR, we identified 16 gene probe sets (Figure 2A) that discriminated the three groups of patients (P < 0.001, FDR = 5%). The 2D hierarchical clustering using the 16 selected gene probe sets clearly separated patients into three distinct groups (Figure 2A) and PCA illustrates the degree of discrimination among the three groups of patients (Figure 2B).


Figure 2
View larger version (41K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 ‘Supervised’ hierarchical clustering and principal components analysis (PCA) discriminating HD, PD and CKD patients. (A) Patients are depicted as vertical columns, with red symbols indicating HD (n = 8), green indicating PD (n = 8) and blue indicating CKD (n = 8) patients. Sixteen gene probe sets (rows, with gene names shown) were used for hierarchical clustering. The relative level of gene expression is depicted from lowest (green) to highest (red) according to the scale shown at the bottom. (B) PCA plot using the 16 selected gene probe sets discriminating the three groups of patients.

 
Discriminating genes included seven genes significantly up-regulated in HD (ATOX1 P = 0.006, RELA P = 0.004, CSDE1 P = 0.009, MIF P = 0.004, LTB4R P = 0.006, GSS P = 0.007 and NFRKB P = 0.004), six genes up-regulated in PD (HRH1 P = 0.002, OLR1 P = 0.003, CHST4 P = 0.003, S100A8 P = 0.006, CXCL12 and GPX7 P = 0.009) and three genes up-regulated in CKD patients (IL8RB P = 0.0008, HDAC5 P = 0.009 and BCL6 P = 0.008) (Figure 2A).

In addition, using Spearman's rank correlation tests, the expression levels of 10/16 gene probe sets were significantly correlated with CRP levels measured in the 24 patients included in the microarray analysis. Only three gene probe sets (CXCL12, IL8RB and MIF) were still statistically significant when included in a multivariate analysis. Interestingly, the model based on the level of expression of the 10 gene probe sets was able to predict the CRP levels with an accuracy of 87% (P < 0.001) (Table 2).


View this table:
[in this window]
[in a new window]

  		Table 2 Correlation analysis between gene expression and CRP levels using the 16 top selected genes discriminating the three groups of patients

 
Functional analysis of selected genes up-regulated in HD and PD patients
When the top selected genes up-regulated in HD patients were analysed using the Ingenuity Pathway Analysis software, we found that all the seven genes were included in the higher scored network (score 17, n = 35 associated genes, P < 0.0001). The relative connectivity diagram in Figure 3A, showed that RELA had a central role in the network being connected by direct or indirect relationships with all the other six genes discriminating HD patients. Additionally, several genes implicated in immune response, inflammation, cell proliferation and lipid metabolism were connected with the seven selected genes, indicating a possible role of our selected genes in these important biological functions. On the other hand, the top selected genes in PD patients presented biological interactions with different genes. The top ranked network (score 9, n = 38 associated genes, P < 0.001) included 4 genes up-regulated in PD patients (CHST4, OLR1, CXCL12 and S100A8) and 34 genes implicated in cell to cell signalling, immune response and cellular movement pathways (Figure 3B).


Figure 3
View larger version (27K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3 Networks of genes modulated in HD and PD patients. Networks, algorithmically generated based on their functional and biological connectivity, were graphically represented as nodes (genes) and edges (the biological relationship between genes). Shaded nodes represented genes identified by our microarray analysis and others (empty nodes) were those that IPA automatically included because biologically linked to our genes based on evidence in the literature. (A) Molecular network generated using the genes up-regulated in HD patients (indicated with red shades) (B) Molecular network generated using the genes up-regulated in PD patients (indicated with green shades). Meaning of node shapes and edges are indicated.

 
Quantitative RT-PCR for RELA, MIF and CXCL12 validated microarray results
RELA mRNA levels measured by RT–PCR were significantly higher in HD patients compared to PD (P < 0.01) and CKD patients (P < 0.01). RELA expression levels were higher in PD compared to CKD patients (P < 0.01) (Figure 4A). In addition, MIF mRNA levels were significantly higher in HD patients compared to PD (P < 0.01) and CKD patients (P < 0.01). There was no statistical difference in expression levels between PD and CKD patients (P = 0.68) (Figure 4B). Finally, CXCL12 mRNA levels were higher in PD patients compared to both HD (P < 0.01) and CKD patients (P < 0.01). CXCL12 expression levels were higher in CKD compared to HD patients (P = 0.02) (Figure 4C). These results were in line with those obtained by the gene expression array.


Figure 4
View larger version (13K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4 RELA, MIF and CXCL12 gene expression by real-time PCR in PBMC from HD, PD and CKD patients. The histograms represent the mean ± SD of (A) RELA, (B) MIF and (C) CXCL12 level of expression determined by real time-PCR in PBMC from 22 HD, 22 PD and 22 CKD patients.

 
Concordance of protein level and gene expression
To document a corresponding intracellular protein level and to validate our model, we selected two genes as representative of the 16 most discriminating genes between the three groups of patients.

GSS protein levels were significantly higher in HD (1.98 ± 0.43 GSS/actin ratio) compared with PD (1.49 ± 0.12 GSS/actin ratio; P < 0.01) and CKD patients (1.09 ± 0.07 GSS/actin ratio; P < 0.01). In addition, PD patients presented significantly higher GSS levels compared with CKD (P < 0.05) (Figure 5A and B). IL8RB protein levels were significantly higher in CKD (1.73 ± 0.19 IL8RB/actin ratio) compared with HD (1.20 ± 0.14; P < 0.01) and PD patients (1.24 ± 0.23 IL8RB/actin ratio, P < 0.01). There was no statistical difference in IL8RB protein levels between HD and PD patients (Figure 5C and D).


Figure 5
View larger version (32K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 5 GSS (A and B) and IL8RB (C and D) protein expression in PBMC from HD, PD and CKD patients. Panel A and C show a representative western blotting experiment respectively for GSS and IL8RB. The histograms represent the mean ± SD of GSS (panel B) and IL8RB (panel D) protein levels in whole-cell lysates of 22 CKD, 22 PD and 22 HD patients assessed by western blotting. GSS protein level was significantly higher in HD compared to PD and CKD patients (*P < 0.01). PD patients presented significantly higher GSS levels compared with CKD (**P < 0.05). IL8RB protein level was significantly higher in CKD compared to HD and PD patients (§P < 0.01), which is concordant with the mRNA expression results obtained by microarray analysis.

 


   Discussion
Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
References
 
During the past decade, several reports have been published regarding the multifactorial origin of microinflammation in dialysis-treated patients [10,11,13,15]. The molecular mechanisms underlying this event have been studied focusing on the role of chronic stimulation of PBMC by the exposure to a bioincompatible system. Altered expression of several genes encoding for proinflammatory cytokines and oxidative stress mediators have been reported as associated with the development and progression of the chronic microinflammatory state in uraemic patients [6–9,27]. This abundant literature, primarily based on univariate single-gene analysis, appears most of the time controversial and not completely exhaustive. To our knowledge, the present report represents the first systematic genomic study based on a relatively high number of patients, targeted to assess the relationship between the activation of a specific transcriptome and the chronic microinflammatory state of dialysis patients. Our study has been performed using a new oligonucleotide microarray technology able to evaluate simultaneously the expression of more than 15 000 genes. However, to take full advantage of the opportunities offered by this high throughput method, it is necessary to manage, integrate and interpret a huge amount of data correctly. Thus, we decided to use a pathway analysis to focus our research on candidate genes known to be associated with inflammation and oxidative stress in order to reduce the false positive rate and the confounding factors not directly associated with the aims of our research. In addition, we performed our gene selection based on the estimation of the FDR as suggested by recent reports indicating the limitations of P-value ranking [28].

Several statistical methods revealed specific genomic patterns for HD, PD and CKD patients associated with different levels of inflammation in the three treatment groups. In this setting, the higher CRP and ferritin levels observed in HD compared to PD patients may be due to the common use in our PD protocols of high biocompatible dialysis fluids [29,30]. Interestingly, using a multivariate analysis, only three genes (MIF, IL8RB and CXCL12) were independently associated with inflammation. CXCL12 and IL8RB were inversely correlated to CRP levels and highly expressed in PD and CKD patients, respectively. MIF was up-regulated in HD patients and directly correlated to CRP levels. Noteworthy, none of these genes have been previously considered as mediators of the microinflammatory state in CKD and dialysis-treated patients. The preliminary selection of inflammation–oxidative stress-related genes in the microarray analysis and the use of CRP levels as a unique marker of inflammation in the correlation study may likely represent the main limits of our study and need to be addressed with a whole genome analysis including genes involved in other pathways (e.g. metabolic, apoptosis) potentially related to the inflammatory process and to introduce in the genomic model more clinical variables of interest (e.g. albumin, cholesterol).

MIF, highly expressed in dialysis patients, encodes for an ‘early response’ cytokine that plays an important patho- genic role in numerous inflammatory disorders [31,32]. It activates macrophages to produce pro-inflammatory mediators and to migrate to the sites of inflammation [33,34]. In a mouse model of spontaneous atherosclerosis, MIF blockade led to a marked reduction of inflammation associated with the disease [35]. Therefore, it is likely that existing anti-MIF therapy (N-acetyl-p-benzoquinone imine) may in the future represent a novel therapeutic strategy to reduce chronic inflammation in dialysis patients.

IL8RB binds IL-8, a chemokine with pro-inflammatory and chemotactic activity [36]. Previous reports have shown a decreased surface expression of this receptor on PBMC of patients with severe chronic inflammatory disease [37,38]. In addition, Calcano et al. reported an increased systemic level of IL-6 in the IL8RB null mice [39]. These observations strongly support our finding of high IL8RB expression in the low-inflamed CKD patients, indicating a role for IL8RB as a regulator of inflammation in uraemic patients.

CXCL12 and its receptor, CXCR4, are important modulators of inflammation and immune response. Recently, different studies demonstrated a role for CXCL12/CXCR4 in PMN turnover [40]. PMNs are the first line of defence and are known to be functionally defective in dialysis patients [41]. Senescent PMNs return to bone marrow and undergo apoptosis. CXCL12 coordinates this event interacting with CXCR4, preferentially expressed on senescent PMNs and inducing TNF-related apoptosis-inducing ligand [40,42]. Thus, the low-CXCL12 expression observed in HD patients may induce an excessive accumulation of senescent PMN, potentially underlying PMNs dysfunction observed in these patients [41–43]. In addition, this observation suggests the necessity to improve bio-molecular studies to address this important issue in dialysis patients.

Finally, two genes highly expressed in HD patients, RELA and GSS, deserve to be mentioned for their close relationship with the inflammatory status. Surprisingly, these genes were not significantly associated with CRP levels when included in a multivariate model. It is conceivable that this observation might be due to their prominent regulation at the post-translational level. RELA is a gene encoding for the human nuclear factor kB (NF-kB) subunit p65. NF-kB is a pivotal regulator of gene expression in response to inflammatory stimuli [44–49] and is itself regulated by the cellular redox status [50,51]. Our observation of NF-kB up-regulation in HD confirms the results obtained by Rangan et al. [52] in a small number of HD patients. This evidence not only suggests a possible role for this transcription factor in the cascade of events leading to systemic microinflammation observed in the HD population, but also emphasizes the necessity to eliminate drugs and chemicals (nicotine) known to enhance its expression [44].

GSS encodes for glutathione (GSH) synthetase, the major enzymatic route for ROS scavenging [53]. GSS is involved in the second step of the de novo synthesis of GSH [54]. In our HD patients, higher GSS expression may suggest an attempt to raise the defence mechanisms against pro-oxidant activities. Moreover, this evidence indicates that anti-oxidant drugs (e.g. acetyl-L-carnitine), by modulating expression of genes involved in the oxidative stress pathway, may be beneficial to a subset of HD patients [55].

In conclusion, our data suggest that specific genomic patterns characterize the systemic microinflammation occurring in CKD, HD and PD patients. In addition, they demonstrate that genomic inflammatory fingerprints are modulated by different modalities of renal replacement therapy. Finally, the set of identified genes may be considered as potential target for future therapeutic approaches to reduce the microinflammatory state in dialysis patients.



   Acknowledgments
 
This study was supported by Ministero dell’ Università e della Ricerca Scientifica (PRIN 2003 granted to G.P., L.G. and G.G. and PRIN 2005 granted to G.P. and L.G.).

Conflict of interest statement. None declared.



   References
Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 

  1. Kaysen GA. The microinflammatory state in uremia. Causes and potential consequences. J Am Soc Nephrol (2001) 12:1549–1557.[Abstract/Free Full Text]
  2. Osterud B, Bjorklid E. Role of monocytes in atherogenesis. Physiol Rev (2003) 83:1069–1112.[Abstract/Free Full Text]
  3. Hansson GK. Inflammation, atherosclerosis and coronary artery disease. N Engl J Med (2005) 352:1685–1695.[Free Full Text]
  4. Yeun JY, Levine RA, Mantadilok V, et al. C-Reactive protein predicts all-cause and cardiovascular mortality in hemodialysis patients. Am J Kidney Dis (2000) 35:469–476.[Web of Science][Medline]
  5. Zimmermann J, Herrlinger S, Pruy A, et al. Inflammation enhances cardiovascular risk and mortality in hemodialysis patients. Kidney Int (1999) 55:648–658.[CrossRef][Web of Science][Medline]
  6. Akahoshi T, Kobayashi N, Hosaka S, et al. In vivo induction of monocyte chemotactic and activating factor in patients with chronic renal failure. Nephrol Dial Transplant (1995) 10:2244–2249.[Abstract/Free Full Text]
  7. Herbelin A, Urena P, Nguyen AT, et al. Elevated circulating levels of interleukin-6 in patients with chronic renal failure. Kidney Int (1991) 39:954–960.[Web of Science][Medline]
  8. Lonnemann G, Van Der Meer JW, Cannon JG, et al. Induction of tumor necrosis factor during extracorporeal blood purification. N Engl J Med (1987) 317:963–964.[Web of Science][Medline]
  9. Nakanishi I, Moutabarrik A, Okada N, et al. Interleukin-8 in chronic renal failure and dialysis patients. Nephrol Dial Transplant (1994) 9:1435–1442.[Abstract/Free Full Text]
  10. Gesualdo L, Pertosa G, Grandaliano G, et al. Cytokines and bioincompatibility. Nephrol Dial Transplant (1998) 13:1622–1626.[Free Full Text]
  11. Walter R, Mischak H, Haller H. Haemodialysis, atherosclerosis and inflammation-identifying molecular mechanisms of chronic vascular disease in ESRD patients. Nephrol Dial Transplant (2002) 17(Suppl 3)):24–29.[Free Full Text]
  12. Ando M, Shibuya A, Yasuda M, et al. Impairment of innate cellular response to in vitro stimuli in patients on continuous ambulatory peritoneal dialysis. Nephrol Dial Transplant (2005) 20:2497–2503.[Abstract/Free Full Text]
  13. Breborowicz A. Free radicals in peritoneal dialysis: agents of damage. Perit Dial Int (1992) 12:194–198.[Free Full Text]
  14. Douvdevani A, Rapoport J, Konforti A, et al. The effect of peritoneal dialysis fluid on the release of IL-1 beta and TNF alpha by macrophages/monocytes. Perit Dial Int (1993) 13:112–117.[Abstract/Free Full Text]
  15. Holmes CJ. Biocompatibility of peritoneal dialysis solutions. Perit Dial Int (1993) 13:88–94.[Free Full Text]
  16. Becker BN, Himmelfarb J, Henrich WL, et al. Reassessing the cardiac risk profile in chronic hemodialysis patients: a hypothesis on the role of oxidant stress and other non-traditional cardiac risk factors. J Am Soc Nephrol (1997) 8:475–486.[Web of Science][Medline]
  17. Canaud B, Cristol J, Morena M, et al. Imbalance of oxidants and antioxidants in haemodialysis patients. Blood Purif (1999) 17:99–106.[CrossRef][Web of Science][Medline]
  18. Descamps-Latscha B, Herbelin A. Long-term dialysis and cellular immunity: a critical survey. Kidney Int Suppl (1993) 41:S135–S142.[Medline]
  19. Himmelfarb J, Stenvinkel P, Ikizler TA, et al. The elephant in uremia: oxidant stress as a unifying concept of cardiovascular disease in uremia. Kidney Int (2002) 62:1524–1538.[CrossRef][Web of Science][Medline]
  20. Locatelli F, Canaud B, Eckardt KU, et al. Oxidative stress in end-stage renal disease: an emerging threat to patient outcome. Nephrol Dial Transplant (2003) 18:1272–1280.[Abstract/Free Full Text]
  21. Lazarus JM, Owen WF. Role of bioincompatibility in dialysis morbidity and mortality. Am J Kidney Dis (1994) 24:1019–1032.[Web of Science][Medline]
  22. Ritz E, Deppisch R, Stier E, et al. Atherogenesis and cardiac death: are they related to dialysis procedure and biocompatibility. Nephrol Dial Transplant (1994) 9(Suppl 2)):165–172.[Web of Science][Medline]
  23. Manns BJ, Burgess ED, Hyndman ME, et al. Hyperhomocyst(e)inemia and the prevalence of atherosclerotic vascular disease in patients with end-stage renal disease. Am J Kidney Dis (1999) 34:669–677.[Web of Science][Medline]
  24. Tarng DC, Wen CT, Huang TP, et al. Increased oxidative damage to peripheral blood leukocyte DNA in chronic peritoneal dialysis patients. J Am Soc Nephrol (2002) 13:1321–1330.[Abstract/Free Full Text]
  25. Golub TR, Slonim DK, Tamayo P, et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science (1999) 286:531–537.[Abstract/Free Full Text]
  26. Storey JD, Tibshirani R. Statistical significance for genomewide studies. Proc Natl Acad Sci USA (2003) 100:9440–9445.[Abstract/Free Full Text]
  27. Friedrich B, Alexander D, Janessa A, et al. Acute effects of hemodialysis on cytokine transcription profiles: evidence for C-reactive protein-dependency of mediator induction. Kidney Int (2006) 70:2124–2130.[Web of Science][Medline]
  28. Hu J, Zou F, Wright FA. Practical FDR-based sample size calculations in microarray experiments. Bioinformatics (2005) 21:3264–3272.[Abstract/Free Full Text]
  29. Rippe B, Simonsen O, Heimburger O, et al. Long-term clinical effects of a peritoneal dialysis fluid with less glucose degradation products. Kidney Int (2001) 59:348–357.[CrossRef][Web of Science][Medline]
  30. Fusshoeller A, Plail M, Grabensee B, et al. Biocompatibility pattern of a bicarbonate/lactate-buffered peritoneal dialysis fluid in APD: aprospective, randomized study. Nephrol Dial Transplant (2004) 19:2101–2106.[Abstract/Free Full Text]
  31. Calandra T, Bernhagen J, Metz CN, et al. MIF as a glucocorticoid-induced modulator of cytokine production. Nature (1995) 377:68–71.[CrossRef][Medline]
  32. Rossi AG, Haslett C, Hirani N, et al. Human circulating eosinophils secrete macrophage migration inhibitory factor (MIF). Potential role in asthma. J Clin Invest (1998) 101:2869–2874.[Web of Science][Medline]
  33. Onodera S, Nishihira J, Koyama Y, et al. Macrophage migration inhibitory factor up-regulates the expression of interleukin-8 messenger RNA in synovial fibroblasts of rheumatoid arthritis patients: common transcriptional regulatory mechanism between interleukin-8 and interleukin-1beta. Arthritis Rheum (2004) 50:1437–1447.[CrossRef][Web of Science][Medline]
  34. Onodera S, Kaneda K, Mizue Y, et al. Macrophage migration inhibitory factor up-regulates expression of matrix metalloproteinases in synovial fibroblasts of rheumatoid arthritis. J Biol Chem (2000) 275:444–450.[Abstract/Free Full Text]
  35. Burger-Kentischer A, Gobel H, Kleemann A, et al. Reduction of the aortic inflammatory response in spontaneous atherosclerosis by blockade of macrophage migration inhibitory factor (MIF). J Atheroscler Res (2006) 184:28–38.[CrossRef]
  36. Baggiolini M, Walz A, Kunkel SL. Neutrophil-activating peptide-1/interleukin 8, a novel cytokine that activates neutrophils. J Clin Invest (1989) 84:1045–1049.[Web of Science][Medline]
  37. Asagoe K, Yamamoto K, Takahashi A, et al. Down-regulation of CXCR2 expression on human polymorphonuclear leukocytes by TNF-alpha. J Immunol (1998) 160:4518–4525.[Abstract/Free Full Text]
  38. Jones SA, Wolf M, Qin S, et al. Different functions for the interleukin 8 receptors (IL-8R) of human neutrophil leukocytes: NADPH oxidase and phospholipase D are activated through IL-8R1 but not IL-8R2. Proc Natl Acad Sci USA (1996) 93:6682–6686.[Abstract/Free Full Text]
  39. Cacalano G, Lee J, Kikly K, et al. Neutrophil and B cell expansion in mice that lack the murine IL-8 receptor homolog. Science (1994) 265:682–684.[Abstract/Free Full Text]
  40. Martin C, Burdon PC, Bridger G, et al. Chemokines acting via CXCR2 and CXCR4 control the release of neutrophils from the bone marrow and their return following senescence. Immunity (2003) 19:583–593.[CrossRef][Web of Science][Medline]
  41. Chonchol M. Neutrophil dysfunction and infection risk in end-stage renal disease. Semin Dial (2006) 19:291–296.[Web of Science][Medline]
  42. Renshaw SA, Parmar JS, Singleton, et al. Acceleration of human neutrophil apoptosis by TRAIL. J Immunol (2003) 170:1027–1033.[Abstract/Free Full Text]
  43. Lum JJ, Bren G, McClure R, et al. Elimination of senescent neutrophils by TNF-related apoptosis-inducing ligand. J Immunol (2005) 175:1232–1238.[Abstract/Free Full Text]
  44. Baeuerle PA, Baltimore D. NF-kB: ten years after. Cell (1996) 87:13–20.[CrossRef][Web of Science][Medline]
  45. Baldwin AS Jr. The NF-kB and IkB proteins: new discoveries and insights. Ann Rev Immunol (1996) 14:649–683.[CrossRef][Web of Science][Medline]
  46. Baeuerle PA, Henkel T. Function and activation of NF-kB in the immune system. Annu Rev Immunol (1994) 12:141–179.[Web of Science][Medline]
  47. Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature (1993) 362:801–809.[CrossRef][Medline]
  48. Cybulsky MI, Gimbrone MA Jr. Endothelial expression of a mononuclear leukocyte adhesion molecule during atherogenesis. Science (1991) 251:788–791.[Abstract/Free Full Text]
  49. Poston RN, Haskard DO, Coucher JR, et al. Expression of intercellular adhesion molecule-1 in atherosclerotic plaques. Am J Pathol (1992) 140:665–673.[Abstract]
  50. Staal FJ, Roederer M, Herzenberg LA, et al. Intracellular thiols regulate activation of nuclear factor kappa B and transcription of human immunodeficiency virus. Proc Natl Acad Sci (1990) 87:9943–9947.[Abstract/Free Full Text]
  51. Staal FJ, Anderson MT, Staal GE, et al. Redox regulation of signal transduction: tyrosine phosphorylation and calcium influx. Proc Natl Acad Sci (1994) 91:3619–3622.[Abstract/Free Full Text]
  52. Rangan GK, Kan G, Steer JH, et al. Altered expression of nuclear factor-B in peripheral blood mononuclear cells in chronic haemodialysis patients. Nephrol Dial Transplant (2006) 21:1137–1139.[Free Full Text]
  53. Alhamdani MS. Impairment of glutathione biosynthetic pathway in uraemia and dialysis. Nephrol Dial Transplant (2005) 20:124–128.[Abstract/Free Full Text]
  54. Griffith OW. Biologic and pharmacologic regulation of mammalian glutathione synthesis. Free Radic Biol Med (1999) 27:922–935.[CrossRef][Web of Science][Medline]
  55. Pertosa G, Grandaliano G, Simone S, et al. Inflammation and carnitine in hemodialysis patients. J Renal Nutr (2005) 15:8–12.[CrossRef][Web of Science][Medline]
Received for publication: 30. 5.07
Accepted in revised form: 16.10.07


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/5/1673    most recent
gfm804v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Zaza, G.
Right arrow Articles by Schena, F. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zaza, G.
Right arrow Articles by Schena, F. P.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?