TY - JOUR
AU1 - Amore, Alessandro
AU2 - Cappelli, Giorgio
AU3 - Cirina, Paola
AU4 - Conti, Giovanni
AU5 - Gambaruto, Caterina
AU6 - Silvestro, Leandra
AU7 - Coppo, Rosanna
AB - Abstract Background. The heat sterilization of glucose solutions for peritoneal dialysis (PDS) induces the formation of glucose degradation products (GDPs), a phenomenon amplified by lactate and neutral pH. In the new three‐compartment bag (3CB) PDS, a glucose solution at pH 3 is kept apart from the buffer until use, and the final solution delivers glucose concentrations that are similar to traditional PDS (TPD), with pH 6 and a lower content of GDPs. As GDPs have oxidant activity that may favour apoptosis, we investigated mesothelial cell apoptosis modulation by 12 h cultures in media supplemented with: (i) two relevant GDPs, methylglyoxal (MGly) and formaldehyde (For) in time and dose‐dependence assays, (ii) GDPs at concentrations detected in TPD and 3CB, and (iii) commercial TPD and 3CB PDS, both with 1.36% glucose. Methods. Apoptosis was evaluated by terminal 3′ uridine labelling. Key proteins involved in the apoptotic pathway were investigated by reverse transcription polymerase chain reaction (RT–PCR) mRNA expression and immunoperoxidase staining (caspase 9, tumour suppressor protein p53, inducible cyclooxygenase COX‐2). Results. The apoptotic effects of MGly and For were dose and time dependent. GPDs at concentrations detected in TPD induced greater transcription and translation of apoptotic pathway proteins (caspase 9, p53 and COX‐2) than GPDs in 3CB. This resulted in a higher apoptotic rate, which was not influenced by addition of sterile glucose. A similar enhancement of apoptosis was detected when mesothelial cells were incubated with TPD, whereas incubation in 3CB PDS resulted in less enhanced apoptosis. The 12 h incubation effect of PDS on cultured mesothelial cells was not related to advanced glycosylated end‐product formation. Conclusions. As the rate of mesothelial cell apoptosis is lower in 3CB than in TPD solutions, the 3CB appears to provide improved biocompatibility. peritoneal dialysis, biocompatibility, glucose degradation products, apoptosis, peritoneal fibrosis Introduction Glucose is employed as an osmotic agent in peritoneal dialysis solutions (PDS) because it is inexpensive, safe and readily metabolized. Heat sterilization and storage of PDS induces glucose auto‐oxidation, causing formation of glucose degradation products (GDPs) [1]. The reaction is enhanced in dilute solutions by high temperatures at neutral pH, and in the presence of catalysing substances such as lactate, calcium and magnesium [2]. GDPs show strong oxidant activity and are able to decrease in vitro cell regeneration [3,4]. Linden et al. [5] demonstrated that some GDPs, including methylglyoxal (MGly), formaldehyde (For), acetaldehyde (Ac), 3‐deoxyglucosone and other dicarbonyl compounds, behave as reactive aldehydes, which participate in the non‐enzymatic glycosylation of proteins and lipids, generating advanced glycosylated/lypo‐oxidated end‐products (AGEs and ALEs). In order to reduce GDP formation, a new three‐compartment bag (3CB) has recently been conceived and developed [6]. The system allows maintenance of glucose at high concentrations and low pH, and is free from substances that catalyse the degradation of glucose during heat sterilization and storage. By mixing the compartments, the system enables patients to prepare different glucose concentrations using the same bag, obtaining a peritoneal solution that has a pH above 6 and is low in GDP levels [7]. Wieslander et al. [7] reported that concentrations of various GDPs in traditional peritoneal dialysis solutions (TPD) were much higher than in PDS obtained by separating glucose from other peritoneal dialysis components during heat sterilization and storage, as is done using the 3CB [7]. An Italian multi‐centre study reported increases in CA125 in effluents using PDS obtained from the 3CB system. As Ca125 is a marker of mesothelial cell mass, the data indicate that this PDS may have better biocompatibility than conventional solutions [8]. A loss of viable mesothelial cells without signs of inflammation is the characteristic pathologic feature of peritoneal membrane use in long‐term dialysed patients, and supports the hypothesis of apoptotic death over that of necrosis. Apoptosis has recently been implicated in progressive scarring that is due to various pathologic events. Programmed cell death is comprised of an active complex machinery requiring gene transcription and translation, and is triggered by different stimuli to monotonically result in activation of specific endonucleases leading to DNA fragmentation [9]. Apoptotic cells promptly undergo phagocytosis by professional and non‐professional phagocytic cells, expressing αvβ3 integrins that are able to recognize the phosphatidylserine residues expressed on cell surfaces during apoptotic death [9]. Apoptosis is characterized by an altered mithocondrial function leading to an abnormal intracellular redox state [8]. Oxidants are strong apoptotic inductors [10] and act by stimulating the neo‐transcription of the tumour suppressor protein, p53 [9]. Carbonyl compounds are formed in uraemia as a result of enhanced oxidative stress caused by uraemic toxin. Moreover, carbonyl compounds amplify oxidant damage in a vicious circle to elicit a series of events resulting in the activation of p53 tumour suppressor protein, of the caspase system, and finally producing internucleosomal DNA breaks that are characteristic of apoptotic death. In recent papers, GDPs incubated with human mesothelial cells induced significant depression of cell growth, viability and function [11,12]. From these observations, we speculated that the 3CB, having lower detectable amounts of GDPs [7], may be less active than TPD in inducing apoptosis of mesothelial cells in vitro. To this aim, we tested the mesothelial cell apoptotic rate induced by scalar doses of relevant GDPs (MGly and For) incubated at different times. Moreover, we extended the analysis by testing each GDP detected previously in 3CB and TPD PDS at the concentrations reported by Wieslander et al. [7]. Finally, we incubated mesothelial cells in both commercial 3CB and TPD PDS in a Transwell system mimicking peritoneal dialysis. To obtain a deeper insight into the apoptotic pathways we evaluated the gene transcription and translation of some key molecules, including caspase 9, the tumour suppressor protein p53, and the inducible isoform of cyclo‐oxygenase (COX‐2). Subjects and methods Experimental design To evaluate the modulation of cell apoptosis by GDPs, human mesothelial cells were obtained from four different healthy donors subjected to non‐neoplastic surgery, were grown at confluence, and were cultured for 12 h in medium supplemented with: (i) two GDPs, MGly and For in both time and dose‐dependence assays, (ii) GDPs at concentrations reported by Wieslander et al. [7] in TPD and 3CB (detailed in Table 1) and with (iii) commercial TPD (Gambro Renal Products) and 3CB (Gambro Renal Products) PDS, both with 1.36% glucose. Mesothelial cell cultures were carried out in 25 cm2 flasks (Becton Dickinson Labware, Franklin Lakes, NY, USA) for 12 h in a humidified atmosphere with 5% CO2–95% air at 37°C. Time (2–96 h) and dose‐dependent tests of apoptosis were carried out using concentrations of MGly ranging from 70 to 1.4 µM and For from 19 to 0.7 µM, which included the concentrations reported by Wieslander et al. for TPD and in 3CB PDS [7] (Table 1). The additional effect of glucose on GDP effects was investigated by supplementing the culture medium with filter sterilized glucose solutions to obtain final glucose concentrations of 1.36%, which matches commercial PDS. To test commercial PDS, mesothelial cells were incubated for 12 h in Transwell plates (Lab‐Tek, Miles Scientific Inc., Naperville, IL, US), while avoiding PDS dilution in culture medium and while mimicking conditions of peritoneal dialysis (schematically represented in Figure 1) with commercial 3CB or TPD PDS (either at final 1.36% glucose concentration) by following previously described procedures to test mesothelial cells with PDS [13]. Transwell plates are transparent, microporous filter culture devices (24.5 mm diameter, with a membrane thickness of 10 mm, pore size 3.0 mm). The growth medium was first added to the lower reservoir followed by addition of the Transwell. Then, the mesothelial cells were seeded into the upper reservoir. The microporous membrane was completely sealed by mesothelial cells within 24 h. Prior to addition of PDS, unattached cells were removed by rinsing the Transwells with culture medium. PDS (either 3CB or TPD) were then added in the upper reservoir of the Transwell. This device has been shown to allow an equilibration between the two solution compartments during the 12 h period [13]. To investigate whether the modification in mesothelial cell apoptosis was mediated by AGE formation and possibly by enhanced GDPs, we pre‐treated mesothelial cells with 50 mg/ml of rabbit anti human RAGE (Research Diagnostics Inc., Flanders, NJ, USA) for 2 h before the 12 h Transwell incubation with commercial 3CB or TPD PDS. Table 1.  Concentrations of GDPs detected by Wieslander et al. [7] in TDP and 3CB; these concentrations for each GDP were used in our experiments   TPD (µM)   3CB (µM)   Acetaldehyde (Ac)  264  1  5 Hydroxymethylfurfural (5HMF)  7  24  Glyoxal (Gly)  7.5  6.9  Methylglyoxal (MGly)  35  2.8  Formaldehyde (For)  9.5  1.7  2‐Furaldehyde (Fur)  0.8  0.2    TPD (µM)   3CB (µM)   Acetaldehyde (Ac)  264  1  5 Hydroxymethylfurfural (5HMF)  7  24  Glyoxal (Gly)  7.5  6.9  Methylglyoxal (MGly)  35  2.8  Formaldehyde (For)  9.5  1.7  2‐Furaldehyde (Fur)  0.8  0.2  View Large Fig. 1.  View largeDownload slide Cross section of a Transwell device. (1) Transwell insert. (2) Upper reservoir for dialysis solutions. (3) Microporus membrane. (4) Lower reservoir for culture medium. Fig. 1.  View largeDownload slide Cross section of a Transwell device. (1) Transwell insert. (2) Upper reservoir for dialysis solutions. (3) Microporus membrane. (4) Lower reservoir for culture medium. Preparation and characterization of human mesothelial cells Human peritoneal specimens were obtained from four consenting non‐uraemic patients undergoing abdominal surgery for non‐neoplastic diseases. Peritoneal samples were washed 3× in PBS and pieces of ∼1 cm2 were incubated in 2 ml trypsin/EDTA (0.125%/0.01%) (Sigma, St Louis, MO, USA) for 15 min at 37°C in continuous rotation. The suspension was centrifuged at 180 g at 4°C for 5 min. The cells obtained were re‐suspended and then plated in 25 cm2 culture flasks in RPMI 1640 medium (Sigma) supplemented with 20% fetal calf serum (FCS), penicillin (100 U/ml), streptomycin (100 µg/ml), l‐glutamine (2 mM) (all reagents from Sigma). The medium was supplemented with 5% d‐valine to inhibit contaminating fibroblast growth. Cells were gently rinsed to remove unattached cells and propagated to obtain 90% confluence prior to being harvested and sub‐cultured. The purity of mesothelial cell culture was assessed by immunoperoxidase using positive staining for cytokeratin, vimentin and the negativity for factor VIII. In the culture experiments, mesothelial cell purity attained 98%. Before use, cells were kept in a quiescent (G0–S1) state by culturing them in 0.1% FCS. The experiments were performed in media supplemented with 20% FCS. Description of GDPs and relative concentrations The GDPs and their concentrations, detailed in Table 1, were the same as detected by Wieslander et al. [7] in 3CB and TPD PDS. All reagents were purchased from Sigma (Sigma Chemical Co.). To avoid evaporation of volatile aldehydes, culture plates were sealed with Parafilm A (American Can Co., Greenwich, CT, USA). A time course (from 2 to 96 h of incubation) at graded doses (MGly from 1.4 to 70 µM and For from 0.7 to 19 µM) was performed. Terminal uridine nick 3′ end labelling (TUNEL) method After a 5 min incubation with 1% paraformaldehyde, cells were fixed in 2:1 v/v ethanol/acetic acid for 10 min at room temperature. After three washings with PBS, cells were incubated with 100 U/ml terminal deoxyuridine transferase, 0.5 µg/ml biotinylated uridine in 1 M potassium cacodylate and 125 mM Tris–HC1, 2.5 mM cobalt chloride (Boeringher, Mannheim, Germany), at pH 6.6 for 1 h at 37°C in a humidified chamber. After washes, a 1:40 solution of fluoresceinated streptavidin (Boeringher) was incubated for 30 min at room temperature. Slides were counter‐stained with 0.3 µg/ml propidium iodine (Sigma) in PBS for 1 min at room temperature. An epi‐fluorescent microscope (Ernst Leitz, Inc., Rockleigh, NJ, USA) was used to detect apoptotic cells, which were quantified by counting the number of fluorescein‐positive cells relative to the total number of cells in at least 10 microscopic fields. For each experiment 200 cells were counted. Cell viability Cell viability was ensured by Trypan Blue staining, and by the effective gene transcription and protein translation of house‐keeping genes. Immunoperoxidase staining for caspase 9 and p53 For immunoperoxidase studies cells were cultured in chamber slide plates (Lab‐Tek, Miles Scientific Inc., Naperville, IL, USA) and fixed after extensive washing with phosphate buffer 0.15 M, pH 7.4 (PBS). Cells were stained using a standard protocol. Briefly, cells were air dried and fixed in chilled acetone for 10 min. Endogenous peroxidase activity was inhibited by incubation for 30 min in 0.025% H2O2 in methanol. Endogenous biotin activity was inhibited by sequential 30 min exposures to avidin D and biotin blocking solutions (Vector Laboratories Inc., Burlingame, CA, USA). The sections were placed in dilute goat serum as blocking agents, followed by application of 1:40 rabbit IgG anti p53 (Dako, Milan, Italy), or 1:100 rabbit anti‐caspase 9 (Santa Cruz Bio‐technologies, Santa Cruz, CA, USA) or normal rabbit IgG at the same final concentrations for 30 min; antibodies were diluted in PBS containing 10% goat serum. After washing, the slides were developed with the Vectastain ABC rabbit IgG detection kit (Vector Laboratories). According to the manufacturer's directions, slides were sequentially exposed to biotinylated goat anti‐rabbit IgG for 1 h, then streptavidin–peroxidase for 30 min, and finally 5,5′diaminobenzidine in 0.05 M Tris buffer at pH 7.2, and hydrogen peroxide 3×10−5% for 5 min. Three washes in PBS were performed between each step. The cells, after counter‐staining with haematoxylin and mounting under coverslips, were qualitatively evaluated for staining. RNA extraction and RT–PCR analysis for tumour suppressor p53 mRNA expression For mRNA, extraction mesothelial cells were washed thrice with sterile PBS, trypsinized and stored in the RNA STAT‐60 (TEL‐TEST B, Inc., Friendswood, TX, USA) at −80°C until use. Total RNA was extracted from cell cultures using the RNA STAT‐60 (TEL‐TEST B, Inc.). Cells grown in monolayers, with or without in vitro challenge, were lysed directly in a culture flask by adding extraction reagent (which contains guanidinium thiocyanate and phenol). The resultant homogenates were incubated for 5 min at room temperature to permit complete dissociation of nucleoprotein complexes. Next, 0.2 ml chloroform was added per ml of extraction reagent; the mixture was shaken vigorously for 15 s, and incubated for 2–3 min at 4°C. Centrifugation separated the homogenates into two phases: a lower red phenol–chloroform phase and a colorless upper aqueous phase containing RNA. The aqueous upper phase was transferred to a fresh tube, and 0.5 ml isopropanol was added per 1 ml of the extraction reagent used for homogenization. This mixture was incubated at room temperature for 10 min and then centrifuged for 10 min at 12 000 g. Supernatants were removed, and the RNA pellet was washed once with 75% ethanol by vortexing and subsequent centrifugation at 7500 g for 5 min at 4°C. The extracted RNA in the pellet was air dried and dissolved in DEPC–H2O for use in RT–PCR analysis. RT–PCR was performed using specific oligonucleotides that detect the p53 and COX‐2 transcripts. The p53 sense and antisense primers were: 5′‐CTGAGGTTGGCTCTGACTGTAC‐3′ and 5′‐CTCATTCAGCTCTCGGAACATCT‐3′, respectively; COX‐2 sense and antisense primers: 5′‐TTCAAATGAGATTGTGGGAAAATTGCT‐3′ and 5′‐AGATCATCTCTGCCTGAGTATCTT‐3′, respectively (TIB MOLBIOL CBA, Genoa, Italy). The RT–PCR was performed by using a commercial kit (Gene Amp RNA‐PCR, Perkin‐Elmer, Roche Molecular System Inc., Branchburg, NJ, USA) following strictly the instructions furnished by the manufacturer. Briefly, 2 µg of total RNA were added to a master mix for reverse transcription to obtain a final concentration of 5 mM MgCl2, 1 mM dNTPs, 1 U RNase inhibitor, 2.5 µM of Oligo(T)16 and 2.5 U of MuLV reverse transcriptase. The tubes were allowed to incubate at room temperature for 10 min, allowing the extension of Oligo(T)16 by reverse transcriptase. RT was performed at 42°C for 45 min and at 99°C for 5 min on a Gene Amp PCR System 9700 (PE Applied Biosystem, Foster City, NJ, USA). The reaction mixture was cooled at 4°C, then 20 µl of reverse transcription products were amplified by 0.0125 U Taq DNA polymerase in a mix reaction containing 2 mM MgCl2 and 0.15 µM of the sense and antisense primers. The PCR profile for both p53 and COX‐2 consisted of a 5 min initial denaturation at 94°C, followed by 40 cycles of 30 s of denaturation at 94°C, 1 min of annealing at 56°C, 1 min of polymerization at 72°C, and finally a 7 min extension at 72°C. Ten microlitres of the PCR amplification products were separated by electrophoresis at 100 V for 1 h in 1% agarose gel in TBE buffer (45 mM Tris–borate, 1 mM EDTA, pH 8). Primers specific for GAPDH were used in each experiment to normalize the results: sense and antisense primers were 5′‐CGGAGTCAACGGATTTGGTCGTAT‐3′; and 5′‐AGCCTTCTCCATGGTGGTGAAGAC‐3′, respectively. The gels were exposed to a UV scanner using the Gel DOC 2000 (Bio‐Rad, Software Quantity One, Hercules, CA, USA). Both p53 and COX‐2 mRNA, normalized for the housekeeping gene GAPDH mRNA, were expressed as percentage increase vs unconditioned cells. Four experiments for each condition were performed in order to allow a statistical analysis of densitometry data. Statistical analysis The values reported in the results represent means±SD of four independent experiments from four different peritoneal cell donors, each done in triplicate to obtain a sufficient quantity of data for statistical analysis. RT–PCR was done on four individual experiments. Statistical significance was analysed by one‐way analysis of variance (ANOVA) using a post‐hoc analysis with Dunnet's multiple comparison test when appropriate. Values of P<0.05 were considered statistically significant. Results Mesothelial cell apoptotic rate By using TUNEL analysis we demonstrated a time‐ and dose‐dependent enhancement of apoptosis induced by incubation of mesothelial cells with graded concentrations of MGly (Figure 2) and For (Figure 3). The effect was visible after 6 h and increased during long‐term incubation. The range of tested GDP concentrations included those reported by Wieslander et al. [7] for TPD and 3CB PDS (MGly 35 and 2.8 µM, For 9.5 and 1.7 µM) [7]. When testing the other relevant GDPs, we also found that glyoxal (Gly), 2‐furaldehyde (Fur) and Ac were more active in the concentrations detected in TPD PDS (Gly 7.5 µM, Fur 0.8 µM, Ac 264 µM) than at the concentrations present in 3CB (Gly 6.9 µM, Fur 0.2 µM, Ac 1 µM) in inducing mesothelial cell programmed cell death (Figure 4). Of note, the apoptotic effect of 5‐hydroxymethyl‐furfural, which is detectable at higher levels in 3CB than in TPD (24 vs 7 µM), was not significantly different at the two concentrations tested. Figure 5 shows the apoptotic bodies induced by incubation of mesothelial cells with 35 µM MGly. When mesothelial cells were incubated in Transwell plates with commercial 3CB and TPD solutions, a significantly lower enhancement of mesothelial cell apoptosis was detected for 3CB vs TPD PDS (Figure 4). Pre‐treatment of mesothelial cells with rabbit anti‐human RAGE failed to modify the apoptotic rate induced by TPD or 3CB PDS (decrease in percent apoptotic cells of 2±2.3% and 0.5±0.5%, respectively, P not significant). Culture media supplemented with 1.36% sterile glucose failed to modify the apoptotic rate induced by MGly (1.4–70 µM) and For (0.7–19 µM) indicating a lack of additive effect (data not shown, P not significant). Even though glucose undergoes auto‐oxidation with final production of carbonyl compounds, the short time of our experiments (12 h) did not allow a sufficient production of GDPs to modify the apoptotic rate induced by directly supplemented GPDs. Fig. 2.  View largeDownload slide Time‐course and dose‐dependent effect of the GDP MGly. Percentage of apoptotic human mesothelial cells cultured with scalar concentrations of MGly (range 70–1.4 µM) for 2–96 h. Data are the mean of four experiments and bars represent SD. The concentrations of MGly detected in the TPDs and in the 3CB PDS were 35 and 2.8, respectively (see Table 1) [7]. Fig. 2.  View largeDownload slide Time‐course and dose‐dependent effect of the GDP MGly. Percentage of apoptotic human mesothelial cells cultured with scalar concentrations of MGly (range 70–1.4 µM) for 2–96 h. Data are the mean of four experiments and bars represent SD. The concentrations of MGly detected in the TPDs and in the 3CB PDS were 35 and 2.8, respectively (see Table 1) [7]. Fig. 3.  View largeDownload slide Time and dose‐dependent effect of the GDP For. Percentage of apoptotic human mesothelial cells cultured with scalar concentrations of For (range 19–0.7 µM) for 2–96 h. Data are the mean of four experiments and bars represent SD. The concentrations of For detected in the TPD and 3CB PDS were 9.5 and 1.7, respectively (see Table 1) [7]. Fig. 3.  View largeDownload slide Time and dose‐dependent effect of the GDP For. Percentage of apoptotic human mesothelial cells cultured with scalar concentrations of For (range 19–0.7 µM) for 2–96 h. Data are the mean of four experiments and bars represent SD. The concentrations of For detected in the TPD and 3CB PDS were 9.5 and 1.7, respectively (see Table 1) [7]. Fig. 4.  View largeDownload slide Modulation of apoptosis of cultured mesothelial cells by GDP and PDS. Percentage of apoptotic mesothelial cells after in vitro incubation with GDPs at the concentrations detected in TPD and 3CB [7] (abbreviations and concentrations for each GDP in TPD and in 3CB are reported on Table 1). The last two columns on the right of the figure represent the percentage of apoptotic cells after incubation in Transwell devices with commercial PDS (TPD and 3CB). The percentage of apoptotic mesothelial cells was significantly higher when cells were cultured in TPD than in 3CB PDS. Cells were cultured for 12 h and apoptotic bodies were detected by TUNEL analysis. In each of the four experiments performed in triplicate, 200 cells were counted. *P<0.01 between conditions in TDP and in 3CB Fig. 4.  View largeDownload slide Modulation of apoptosis of cultured mesothelial cells by GDP and PDS. Percentage of apoptotic mesothelial cells after in vitro incubation with GDPs at the concentrations detected in TPD and 3CB [7] (abbreviations and concentrations for each GDP in TPD and in 3CB are reported on Table 1). The last two columns on the right of the figure represent the percentage of apoptotic cells after incubation in Transwell devices with commercial PDS (TPD and 3CB). The percentage of apoptotic mesothelial cells was significantly higher when cells were cultured in TPD than in 3CB PDS. Cells were cultured for 12 h and apoptotic bodies were detected by TUNEL analysis. In each of the four experiments performed in triplicate, 200 cells were counted. *P<0.01 between conditions in TDP and in 3CB Fig. 5.  View largeDownload slide (A) Viable mesothelial cells incubated with culture medium. (B) Apoptotic mesothelial cells after 12 h incubation with medium supplemented with MGly 35 mM. The arrows indicate chromatin condensation, specific for apoptotic bodies. Fig. 5.  View largeDownload slide (A) Viable mesothelial cells incubated with culture medium. (B) Apoptotic mesothelial cells after 12 h incubation with medium supplemented with MGly 35 mM. The arrows indicate chromatin condensation, specific for apoptotic bodies. Caspase 9 and p53 expression Caspase 9 and p53 expression in human mesothelial cells conditioned with GDPs at the concentrations detected in TPD and 3CB fluids showed similar profiles to the modulation of apoptosis (Figure 6) with a higher percentage of caspase 9 and p53 positively stained mesothelial cells when GDPs were used at the concentrations present in TPD solutions compared with concentrations in 3CB PDS (P<0.01). The commercial TPD PDS were more active than 3CB in inducing the neo‐expression of caspase 9 and p53, which are involved in the apoptotic pathway in mesothelial cells. The expression of caspase 9 and p53 in unconditoned mesothelial cells was negligible and was rarely present in trace amounts. Figure 7 shows the cytoplasmatic positive staining for p53 when 35 µM MGly was incubated with mesothelial cells. Fig. 6.  View largeDownload slide Modulation of apoptotic pathways in cultured mesothelial cells by GDP and PDS. Percentage of p53 or caspase 9 positive mesothelial cells in immunoperoxidase after in vitro incubation with GDPs at the concentrations detected in TPD and 3CB [7] (abbreviations and concentrations for each GDP in TPD and 3CB are reported on Table 1). The last two columns on the right represent the percentage of cells positive for p53 or caspase 9 after incubation in Transwell devices with commercial PDS (TPD and 3CB). The percentage of p53 and caspase 9 positive mesothelial cells was significantly higher when cells were cultured in TPD than in 3CB PDS. In each of the four experiments, performed in triplicate, 200 cells were counted. *P<0.01 between conditions in TDP and in 3CB. Fig. 6.  View largeDownload slide Modulation of apoptotic pathways in cultured mesothelial cells by GDP and PDS. Percentage of p53 or caspase 9 positive mesothelial cells in immunoperoxidase after in vitro incubation with GDPs at the concentrations detected in TPD and 3CB [7] (abbreviations and concentrations for each GDP in TPD and 3CB are reported on Table 1). The last two columns on the right represent the percentage of cells positive for p53 or caspase 9 after incubation in Transwell devices with commercial PDS (TPD and 3CB). The percentage of p53 and caspase 9 positive mesothelial cells was significantly higher when cells were cultured in TPD than in 3CB PDS. In each of the four experiments, performed in triplicate, 200 cells were counted. *P<0.01 between conditions in TDP and in 3CB. Fig. 7.  View largeDownload slide Immunoperoxidase staining for p53 in unconditioned mesothelial cells (A) and after incubation with MGly 35 mM (B). The arrow indicates cytoplasmatic positive staining for p53. Fig. 7.  View largeDownload slide Immunoperoxidase staining for p53 in unconditioned mesothelial cells (A) and after incubation with MGly 35 mM (B). The arrow indicates cytoplasmatic positive staining for p53. p53 and COX‐2 gene expression GDPs elicited the transcription of the gene encoding for p53. The concentrations of GDPs in TPD solutions strongly activated the transcription of p53 mRNA compared with the concentrations present in 3CB PDS (Figures 8 and 9). Accordingly, p53 mRNA transcription was significantly more enhanced by TPD than by 3CB PSD. GDPs similarly modulated the transcription of the gene encoding for COX‐2 (Figures 11 and 12). The protein was effectively translated, as shown after treatment with MGly 35 µM in immunoperoxidase staining (Figure 10). Fig. 8.  View largeDownload slide Modulation of p53 mRNA expression in mesothelial cells by GDPs and PDS. Representative electrophoresis of the RT–PCR amplification product of p53 mRNA in mesothelial cells incubated with GDPs at concentrations detected in TPD and 3CB PDS [7] (abbreviations and concentrations detailed in Table 1). The last two lines on the right represent results after cell incubation in Transwell devices with commercial PDS (TPD and 3CB). Lane 1, basal; lane 2, Ac 264 µM; lane 3, Ac 1 µM; lane 4, 5HMF 7 µM; lane 5, 5HMF 24 µM; lane 6, Gly 7.5 µM; lane 7, Gly 6.9 µM; lane 8, MGly 35 µM; lane 9, MGly 2.8 µM; lane 10, For 9.5 µM; lane 11, For 1.7 µM; lane 12, Fur 0.8 µM; lane 13, Fur 0.2 µM; lane 14, TDP PDS; lane 15, 3CB PDS. Fig. 8.  View largeDownload slide Modulation of p53 mRNA expression in mesothelial cells by GDPs and PDS. Representative electrophoresis of the RT–PCR amplification product of p53 mRNA in mesothelial cells incubated with GDPs at concentrations detected in TPD and 3CB PDS [7] (abbreviations and concentrations detailed in Table 1). The last two lines on the right represent results after cell incubation in Transwell devices with commercial PDS (TPD and 3CB). Lane 1, basal; lane 2, Ac 264 µM; lane 3, Ac 1 µM; lane 4, 5HMF 7 µM; lane 5, 5HMF 24 µM; lane 6, Gly 7.5 µM; lane 7, Gly 6.9 µM; lane 8, MGly 35 µM; lane 9, MGly 2.8 µM; lane 10, For 9.5 µM; lane 11, For 1.7 µM; lane 12, Fur 0.8 µM; lane 13, Fur 0.2 µM; lane 14, TDP PDS; lane 15, 3CB PDS. Fig. 9.  View largeDownload slide Densitometric analysis of bands from RT–PCR amplification products of p53 mRNA reported in Figure 8. For each RT–PCR amplification product from the experimental conditions reported in Figure 8, results were calculated as the p53/GAPDH mRNA optical density ratio divided by a similar ratio in basal control cells and multiplied ×100 in order to obtain per cent increases from basal expression. Columns represent means of four separate experiments. Bars represent SD. For abbreviations and symbols see Table 1. Fig. 9.  View largeDownload slide Densitometric analysis of bands from RT–PCR amplification products of p53 mRNA reported in Figure 8. For each RT–PCR amplification product from the experimental conditions reported in Figure 8, results were calculated as the p53/GAPDH mRNA optical density ratio divided by a similar ratio in basal control cells and multiplied ×100 in order to obtain per cent increases from basal expression. Columns represent means of four separate experiments. Bars represent SD. For abbreviations and symbols see Table 1. Fig. 10.  View largeDownload slide Immunoperoxidase staining for COX‐2. Unconditioned mesothelial cells (A) and cells after incubation with MGly 35 µM (B). The arrow indicates positive cytoplasmic staining for COX 2. Fig. 10.  View largeDownload slide Immunoperoxidase staining for COX‐2. Unconditioned mesothelial cells (A) and cells after incubation with MGly 35 µM (B). The arrow indicates positive cytoplasmic staining for COX 2. Fig. 11.  View largeDownload slide Modulation of COX 2 mRNA expression in mesothelial cells by GDPs and PDS. Representative electrophoresis of the RT–PCR amplification product of COX 2 mRNA in mesothelial cells incubated with GDPs at concentrations detected in TPD and 3CB PDS (7) (abbreviations and concentrations detailed in Table 1). The last two lines on the right represent results after cell incubation in Transwell devices with commercial PDS (TPD and 3CB). Lane 1, basal; lane 2, Ac 264 µM; lane 3, Ac 1 µM; lane 4, 5HMF 7 µM; lane 5, 5HMF 24 µM; lane 6, Gly 7.5 µM; lane 7, Gly 6.9 µM; lane 8, MGly 35 µM; lane 9, MGly 2.8 µM; lane 10, For 9.5 µM; lane 11, For 1.7 µM; lane 12, Fur 0.8 µM; lane 13, Fur 0.2 µM; lane 14, TDP PDS; lane 15, 3CB PDS. Fig. 11.  View largeDownload slide Modulation of COX 2 mRNA expression in mesothelial cells by GDPs and PDS. Representative electrophoresis of the RT–PCR amplification product of COX 2 mRNA in mesothelial cells incubated with GDPs at concentrations detected in TPD and 3CB PDS (7) (abbreviations and concentrations detailed in Table 1). The last two lines on the right represent results after cell incubation in Transwell devices with commercial PDS (TPD and 3CB). Lane 1, basal; lane 2, Ac 264 µM; lane 3, Ac 1 µM; lane 4, 5HMF 7 µM; lane 5, 5HMF 24 µM; lane 6, Gly 7.5 µM; lane 7, Gly 6.9 µM; lane 8, MGly 35 µM; lane 9, MGly 2.8 µM; lane 10, For 9.5 µM; lane 11, For 1.7 µM; lane 12, Fur 0.8 µM; lane 13, Fur 0.2 µM; lane 14, TDP PDS; lane 15, 3CB PDS. Fig. 12.  View largeDownload slide Densitometric analysis bands from RT–PCR amplification products of COX 2 mRNA reported in Figure 11 For each RT–PCR amplification product from the experimental conditions reported in Figure 11, results were calculated as the COX 2/GAPDH mRNA optical density ratio divided by a similar ratio in basal control cells and multiplied ×100 in order to obtain per cent increases from basal expression. Columns represent means from four separated experiments. Bars represent SD. For abbreviations and symbols see Table 1. Fig. 12.  View largeDownload slide Densitometric analysis bands from RT–PCR amplification products of COX 2 mRNA reported in Figure 11 For each RT–PCR amplification product from the experimental conditions reported in Figure 11, results were calculated as the COX 2/GAPDH mRNA optical density ratio divided by a similar ratio in basal control cells and multiplied ×100 in order to obtain per cent increases from basal expression. Columns represent means from four separated experiments. Bars represent SD. For abbreviations and symbols see Table 1. Discussion In this study we demonstrated that MGly and For, which are two relevant GDPs known to be produced in PDS, induced a dose‐ and time‐dependent apoptosis of cultured human mesothelial cells. Apoptotic effects were also exerted by each of the GDPs detected previously in PDS [7]. From our data, we conclude that the GPD concentrations reported in TPD PDS are more active than those contained in the new formulation from 3CB in inducing apoptosis of mesothelial cells in culture. Moreover, mesothelial cells incubated with commercial TPD PDS are more prone to undergo apoptosis than when incubated with 3CB PDS, even with the same glucose concentration. Even though a direct demonstration of a GDP concentration effect is lacking, this is highly probable, considering that the two PDS preparations had identical pH and glucose content, but differed mainly in GDP concentration. Pre‐treatment of mesothelial cells with rabbit anti‐human RAGE failed to modify the apoptotic rate induced by TPD or 3CB PDS, suggesting that the short‐term incubation of 12 h was not sufficient to enhance AGE formation in amounts suitable to modify the apoptotic effect of GDPs. GDPs induce direct modifications of the cell redox state, interfering with the complex mithocondrial machinery to result in DNA oxidative stress that is responsible for cellular suicide. It is possible that the toxic effects of GDPs we found during in vitro conditions may be further enhanced in vivo during repeated exposure to long‐term PD. To investigate the modulation of either PDS or pure aldehydes on mesothelial cell apoptosis, we decided to use mesothelial cells from healthy subjects instead of uraemic patients to avoid possible underlying damage due to uraemic toxins and naturally occurring AGEs. The effects we found would eventually be amplified in patients undergoing regular PD treatment, as their mesothelial cells should be relatively depleted in protective scavengers. Peritoneal fibrosis with loss of ultrafiltration capability represents one of the major cause of peritoneal dialysis patient drop‐out. The pathogenetic mechanisms responsible for peritoneal fibrosis are still largely unknown and are under investigation. In addition to peritonitis, much recent attention has been devoted to the role of biocompatibility of PDS. The negative effects of PDS, pH, and lactate buffer [14] have been extensively evaluated, and studies attempting to elucidate the negative effects of glucose are still ongoing. An anti‐proliferative effect of glucose per se is well known and has been ascribed to a specific inhibition of cyclin‐dependent kinases responsible for cell‐cycle arrest in the G0–S1 phase [15]. In our experiments, culture media supplemented with 1.36% sterile glucose failed to modify the apoptotic rate induced by MGly and For, indicating a lack of additive effect. Even though glucose undergoes auto‐oxidation to produce carbonyl compounds, the short time of our experiments (12 h) did not allow a production of GDPs sufficient to modify the apoptotic rate induced by directly supplemented GPDs. At high concentrations glucose may alter the intracellular redox state and this effect is mediated by a phosphokinase C activation (polyol pathway) or by altering the NADH/NAD ratio responsible for pseudo hypoxia conditions (esosomonophosphate‐shunt). The pro‐oxidant activity exhibited by glucose is enhanced by the tendency of glucose at high concentration to undergo auto‐oxidation [16], a process that is accelerated and catalysed by high temperature and neutral pH, such as during heat sterilization of glucose containing‐PDS [7]. The glucose oxidation results in formation of the so called ‘carbonylic stress products’, having an aldehyde biochemical structure and strong oxidative effects [17]. Aldehydes are able to induce DNA oxidative damage that could represent the starting point of the apoptotic process. Moreover, a modification of the intracellular redox state with increases in oxidants and eventual reductions in scavenger molecules or enzymes (such as gluthatione, gluthatione peroxidase, superoxidodismutase, characteristic of uraemia), through activation of different phosphorylation systems may induce IkB phosphorylation and consequent activation and nuclear translocation of the transcriptional factor NFkB. This factors uses consensus sequences initiating the transcription of different genes, such as TGF‐β, COX‐2, nitric oxide synthase and collagens involved in the progressive scarring processes. Apoptosis has been recently implicated in the ‘silent’ cell loss occurring in different scarring processes [9]. The redox imbalance is pivotal in the process of apoptosis, inducing modifications in genes (such as Bcl‐2, Bad, Bax, p53) involved in programmed cell death [9,18,19]. Oxidative agents provoke apoptosis, and scavengers such as superoxide dismutase or glutathione prevent apoptotic death. We demonstrated that GPDs are able to induce the neo‐transcription of p53 and that the effect is dose‐dependent. In fact, we demonstrated that GDPs at the concentrations present in TPD were significantly more active than the lower concentrations of GDPs present in 3CB, and that the commercially available TPD PDS significantly up‐regulated p53 gene expression when compared with 3CB. The effective translation of p53 mRNA was demonstrated by the higher percentage of p53 positive mesothelial cells in immunoperoxidase induced by GDPs in TPD fluids with respect to the percentage of positive cells following treatment with GDPs at the concentrations in 3CB fluids. Similarly, COX‐2 gene transcription paralleled the modulation of p53 induced by GDPs in TPD fluids and 3CB and by commercial TPD fluids and 3CB. The translated p53 protein is known to modulate the expression of the WAF1/CIP1 gene [9], which in turn induces the synthesis of the regulator protein p21. Finally, p21 inhibits the activity of cyclin E. Apoptosis then occurs following the consequent inhibition of the delta DNA polymerase isoform, the most active enzyme involved in DNA replication. The finding that caspase 9 expression occurs in immunoperoxidase strongly supports the hypothesis of a pro‐apoptotic role for GDPs. Further supporting this was the higher percentage of caspase 9 positive mesothelial cells when treated with GDPs at concentrations in TPD or in commercial TPD PDS compared with GDPs in 3CB and in commercial 3CB PDS. In this regard, it is of note that the percentage of caspase 9 positive mesothelial cells was slightly higher than that of p53 positive cells. This finding could be explained by the timing of activation of the different pro‐apoptotic molecules, as caspase 9 is activated earlier than p53 during the apoptotic cascade. TUNEL analysis revealed that caspase 9 expression, p53 gene transcription and translation, and COX‐2 de novo synthesis result in an increased percentage of apoptotic cells. GDPs concentrations detected in TPD and in commercial TPD PDS induced significant increases in the percentage of TUNEL positive mesothelial cells compared with GDP concentrations found in 3CB PDS. Two recent papers investigated the effects of pure GDPs and of commercial PDS on mesothelial cell viability, growth and synthetic activity [11,12]. While the data reported on pure GDPs were in agreement with our apoptosis results [11], they failed to show significant inhibition of mesothelial cell proliferation when cells were incubated during short‐term experiments in freshly prepared glucose solutions, either heat sterilized or pore membrane filtered [11]. We were able to perform our experiments with commercial PDS by adopting the Transwell device (Figure 1), which mimics peritoneal dialysis conditions. This device exposes one cell surface to culture medium and the other to commercial PDS. We tested commerial PDS that had been stocked for months before the use and was likely to contain naturally formed GDPs. For these reasons we believe that our current experimental conditions more closely reproduced in vivo PD conditions. The long‐term experiments performed by Witowski et al. [12], showing a gradual loss of cell viability, decreased adhesion, and cell shedding after prolonged incubation of mesothelial cells with GDPs (up to 36 days), fit very well with our data of enhanced apoptosis. We cannot directly compare the apoptotic rate induced by single GDPs with that induced by commercial PDS, as we used different culture procedures. We directly supplemented the culture medium of mesothelial cells grown in plastic flasks with pure GDPs whereas commercial PDS were tested on mesothelial cells in Transwell devices. The significantly higher apoptotic rate induced by TPD compared with 3CB PDS is in agreement with previous reports showing that mesothelial cells from patients using 3CB formed a confluent layer in a significantly shorter time than mesothelial cells from patients using TPD [20]. This finding also agrees with the increased levels of CA125, a marker of mesothelial cell mass in peritoneal dialysis effluent, in 3CB compared with TPD PSD [8]. Our data support the hypothesis of a mesothelial cell layer loss induced by GPDs at the high concentrations that are detected in TPD solutions. The resulting exposure of matrix components and vascular endothelial cells to the oxidative action of GDPs would produce in peritoneal barrier damage. The acceleration of non‐enzymatic AGE formation (Namiki's pathway) by GDPs further favours peritoneal fibrosis and loss of ultrafiltration capability. Being 3CB PDS and their corresponding GPDs concentrations less active in inducing apoptosis of mesothelial cells, a better biocompatibility of this new formulation of PDS is expected in respect to enhancement of peritoneal fibrosis in long‐term dialyzed patients. Conflict of interest statement. None declared. Correspondence and offprint requests to: Alessandro Amore, MD, U.O.A. Nefrologia Dialisi e Trapianto, Ospedale Regina Margherita, Piazza Polonia, 94, I‐10126 Torino, Italy. Email: amorenefro@oirmsantanna.piemonte.it References 1 Kjellstrand P, Martinsson E, Wieslander AP, Holmqvist B. Development of toxic degradation products during heat sterilization of glucose‐containing fluids for peritoneal dialysis: influence of time and temperature. Perit Dial Int  1995; 15: 26–32 Google Scholar 2 Trzeciak M, Zakrewski Z, Siedlecka E, Furmancyk Z. Studies on dependence of stability of glucose in solutions applied to peritoneal dialysis upon their electrolyte composition. Acta Pol Pharm  1989; 2: 174–178 Google Scholar 3 Wieslander AP, Nordin MK, Kiellstrand PT, Boberg U. Toxicity of peritoneal dialysis gives rise to several aldehydes. Perit Dial Int  1993; 13: 208–213 Google Scholar 4 Wieslander AP, Nordin MK, Martinsson E. Heat‐sterilized PD fluids impair growth and inflammatory responses of cultured cell lines and human leukocytes. Clin Nephrol  1993; 39: 343–348 Google Scholar 5 Linden T, Forsbach G, Deppish R, Henle T, Wieslander A. 3‐Deoxyglucosone, a promoter of advanced glycation end products in fluids for peritoneal dialysis. Perit Dial Inter  1998; 18: 290–293 Google Scholar 6 Cappelli G. Dialisato più fisiologico con una nuova sacca per dialisi peritoneale. In: Tecniche Nefrologiche e Dialitiche. Bios, Cosenza, 1997 281–288 Google Scholar 7 Wieslander AP, Deppisch R, Svensson E, Forsback G, Speidel R, Rippe B. In vitro biocompatibility of a heat‐sterilized, low‐toxic and less acid fluid for peritoneal dialysis. Perit Dial Int  1995; 15: 158–164 Google Scholar 8 Cappelli G, Bandiani G, Cancarini GC et al. Low concentration of glucose degradation products in peritoneal dialysis fluids and their impact on biocompatibility parameters: prospective cross‐over study with a three‐compartment bag. Adv Perit Dial  1999; 15: 238–242 Google Scholar 9 Amore A, Coppo R. Role of apoptosis in pathogenesis and progression of renal diseases. Nephron  2000; 86: 99–104 Google Scholar 10 Buttke TM, Sandstrom PA. Oxidative stress as mediator of apoptosis. Immunol Today  1994; 15: 7–10 Google Scholar 11 Witowski J, Wisniewska J, Korybalska K et al. Prolonged exposure to glucose degradation products impairs viability and function of human peritoneal mesothelial cells. J Am Soc Nephrol  2001; 12: 2434–2441 Google Scholar 12 Witowski J, Korybalska K, Wisniewska J et al. Effect of glucose degradation products on human peritoneal mesothelial cell function. J Am Soc Nephrol  2000; 11: 729–739 Google Scholar 13 Yang AH, Chen JY, Lin YP, Huang TP, Wu CW. Peritoneal dialysis solutions induce apoptosis of mesothelial cells. Kidney Int  1997; 51: 1280–1288 Google Scholar 14 Chung SH, Stenvinkel P, Bergstrom J, Lindholm B. Biocompatibility of new peritoneal dialysis solutions: what can we hope to achieve? Perit Dial Int  2000; 20 [Suppl 5]: S57–S67 Google Scholar 15 Wolf G. Cell cycle regulation in diabetic nephropathy. Kidney Int  2000; 77: S59–66 Google Scholar 16 Ha H, Kim KH. Pathogenesis of diabetic nephropathy: the role of oxidative stress and protein kinase C. Diabetes Res Clin Pract  1999; 45: 147–151 Google Scholar 17 Suzuki D, Miyata T. Carbonyl stress in the pathogenesis of diabetic nephropathy. Intern Med  1999; 38: 309–314 Google Scholar 18 Miyashita T, Krajewski S, Krajewska M et al. Tumor suppressor protein p53 is a regulator of bcl‐2 and bax gene expression in vitro and in vivo. Oncogene  1994; 9: 1799–1805 Google Scholar 19 Savill J, Mooney A, Hughes J. Apoptosis and renal scarring. Kidney Int  1996; 54: S14–S17 Google Scholar 20 Jarkelid LE, Deppish R, Wieslander A. Presence of GDP in PD fluids decreases ex vivo mesothelial cell regeneration. Perit Dial Inter  1999; 19 [Suppl 1]: S54. Google Scholar European Renal Association–European Dialysis and Transplant Association
TI - Glucose degradation products increase apoptosis of human mesothelial cells
JF - Nephrology Dialysis Transplantation
DO - 10.1093/ndt/gfg003
DA - 2003-04-01
UR - https://www.deepdyve.com/lp/oxford-university-press/glucose-degradation-products-increase-apoptosis-of-human-mesothelial-AnS5hRq7du
SP - 677
EP - 688
VL - 18
IS - 4
DP - DeepDyve
ER -