Glucose Degradation Products and Peritoneal Membrane Function

Glucose Degradation in Dialysis Fluids

The spontaneous breakdown of glucose is very slow but increases rapidly at high temperatures. Thus, the majority of GDPs are formed during heat-sterilization of PDF (H-PDF) (2). The accumulation of GDPs is reflected by an increase in ultraviolet absorbance (mainly at 284 nm and 228 nm) and a corresponding appearance of several peaks during high performance liquid chromatography (HPLC) analysis (3). However, only a few of these products could be identified and quantified (Table 1). They turned out to be small molecular weight aldehydes including 5-hydroxymethylfuraldehyde, which is often regarded as a general indicator of glucose breakdown.

OPEN IN VIEWER

Table 1

Glucose Degradation Products (GDPs) Identified in Peritoneal Dialysis Solutions

GDP	Concentration (mmol/L)	Reference
Acetaldehyde	120–420	(3)
3-Deoxyglucosone	118–154	(4)
Formaldehyde	6–15	(3)
2-Furaldehyde	0.05–2	(3)
Glyoxal	3–14	(3)
5-Hydroxymethylfuraldehyde	6–30	(3)
Methylglyoxal	2–23	(3)
Valeraldehyde	ND	(3)

ND = not determined.

Biological Effects of Gdps

While significant buildup of GDPs can be detected in H-PDF, the same solutions sterilized by filtration contain virtually no GDPs (3,5). Therefore, the most common experimental approach to assess the biological impact of GDPs is to compare the function of cells exposed to either heat-or filtered-sterilized fluids. Typically, cells are incubated in the culture medium mixed with an equal volume of the solution tested (and supplemented with fetal calf serum if required). The pH of such a mixture is adjusted to neutral in order to exclude a well-known inhibitory impact of PDF's acidic milieu (6,7).

In a seminal paper from 1991, Wieslander et al. reported that under these conditions H-PDF inhibited proliferation of L929 murine fibroblasts to a significantly greater extent compared to filter-sterilized PDF (F-PDF) (8). These results were later confirmed not only in L929 cells (9) but also in RAW 264.7 macrophages, and SH-SY5Y neuroblastoma cells (10). The difference between H-PDF and F-PDF was observed not only in the proliferation assays. It was also detected in tests examining the functions Peritoneal dialysis fluids were laboratory prepared and sterilized either by heat (H-PDF: 121°C, 0.2 MPa, 20 min) or filtration (F-PDF: 0.2 m), as described previously (16). The accumulation of GDP was confirmed by UV absorbance at 284 nm. After the incubation, HPMC were treated with 1.25 mg/mL MTT salt [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolinum bromide] for 4 hours and then lysed with acidic solution of 20% sodium dodecyl sulfate. The amount of formazan product generated was assessed by the absorbance at 595 nm with a reference wavelength of 690 nm. Data represent the means ± SEM of 10 experiments performed of neutrophils and mononuclear leukocytes. These studies demonstrated that, under in vitro conditions, the ability of leukocytes to mount a respiratory burst, phagocytose, and release inflammatory cytokines was much better preserved in cells treated with F-PDF compared to those exposed to autoclaved solutions (11,5,12). It is possible that similar effects occur in vivo, since peritoneal macrophages isolated from rats injected intraperitoneally with H-PDF were found to generate less superoxide than cells obtained from animals receiving F-PDF (13). In the present study we examined the effect of differently sterilized PDF on human peritoneal mesothelial cells (HPMC). We found that the presence of H-PDF in culture media resulted in a significant reduction in HPMC viability compared to F-PDF (Figure 1). Loss of cell viability was also accompanied by a decreased ability of HPMC to synthesize cytokines in response to interleukin (IL)-1β stimulation (Figure 2). Mesothelial cells treated with H-PDF released significantly less monocyte chemotactic protein-1 (MCP-1) than HPMC exposed to F-PDF. Importantly, the inhibitory effects seen in response to H-PDF were more pronounced in solutions with high glucose concentration and therefore with increased GDP content. On the other hand, some suppression of MCP-1 secretion was also evident in cells exposed to F-PDF with 4.25% glucose, which suggests that elevated glucose per se or corresponding hyperosmolality may also inhibit the release of chemokines from HPMC.

OPEN IN VIEWER

Figure 1

Viability of human peritoneal mesothelial cells (HPMC) exposed to heat- and filter-sterilized peritoneal dialysis fluids (H-PDF, F-PDF). Cells were incubated for 24 hours in the standard M199 (with final fetal calf serum concentration of 0.3%) mixed with an equal volume of PDF containing either 1.5% or 4.25% glucose.

OPEN IN VIEWER

Figure 2

Release of monocyte chemotactic protein-1 (MCP-1) by human peritoneal mesothelial cells (HPMC) exposed to heat- and filter-sterilized peritoneal dialysis fluids (H-PDF, F-PDF). Cells were incubated for 24 hours in standard M199 (with final fetal calf serum concentration of 0.3%) mixed with an equal volume of PDF containing either 1.5% or 4.25% glucose, and stimulated with 1000 pg/mL interleukin-1β. After the incubation, the supernates were analyzed for MCP-1 with specific immunoassay, and the protein content in solubilized HPMC monolayers was measured with the Bradford method. Data represent the mean ± SEM MCP-1 release from five experiments performed with cells from different donors. Asterisks represent a statistically significant difference (p < 0.05) compared to the control.

Wieslander et al. have also performed studies in which cells were exposed directly to aldehydes that had been detected in PDF (14). These GDPs inhibited growth of L929 cells in a dose-dependent fashion, however, the concentrations that significantly impaired cell proliferation were much higher than those found in PDF. More importantly, when clinically relevant doses of GDPs were added to PDF sterilized by filtration, they did not produce the inhibition seen in response to H-PDF. This rather surprising observation may indicate that toxicity exerted by autoclaved solutions is primarily mediated by other not yet identified products.

Glucose Degradation Products and Peritoneal Mesothelium

At first glance the response of HPMC to GDP-containing dialysis solutions appears to be very similar to that observed in established cells lines. Peritoneal mesothelial cells are usually isolated from specimens of omentum obtained during elective abdominal surgery (15). The procedure is laborious, the yield of cells is often very poor, and sometimes the attempts to propagate HPMC end in failure. In contrast, cells of established and transformed lines are easy to maintain and can be purchased all over the world, offering a highly reproducible experimental tool. Why is it then important to study primary HPMC cultures? Our recent observations indicate that the degree to which the data obtained with cell lines can be extrapolated to HPMC may be limited (16).

We exposed L929 fibroblasts and HPMC to a broad range of GDP concentrations under the same experimental conditions. We found that in both cell systems GDPs suppressed cell growth in a dose-dependent manner. However, some of the GDPs studied (acetaldehyde, glyoxal, furaldehyde) inhibited HPMC proliferation to a greater degree compared to L929 cell proliferation. In addition, when cell viability was studied with either lactate dehydrogenase release or MTT conversion tests, none of the GDPs was found to impair the viability of L929 cells, while several GDPs significantly decreased the viability of mesothelial cells, with formaldehyde, glyoxal, and methylglyoxal being most toxic. These data clearly indicate that the real impact of GDPs on peritoneal mesothelial cells may be more serious than judged on the basis of results obtained with cell lines.

As mentioned earlier, clinically relevant doses of GDPs identified in PDF had no significant impact on L929 cells (14). Although mesothelial cells appear to be more sensitive to the insult from GDPs, these doses were also ineffective in the HPMC system. Unfortunately, this observation does not necessarily prove that toxicity associated with GDPs (or at least with those identified so far) is negligible. In most in vitro experiments performed to date, cells were treated with GDPs for a very short period of time (typically for 24 – 72 hours), while in the course of clinical CAPD, the peritoneal membrane is continuously exposed to GDP-containing fluids for years. Thus, one may hypothesize that even very low doses of GDPs may produce deleterious effects in HPMC when the exposure period is long enough.

To address this issue we have recently examined the function of HPMC treated chronically with GDPs. We designed an in vitro system in which HPMC were incubated in the presence of clinically relevant doses of GDPs for up to 6 weeks (17). Exposure to GDPs under these conditions resulted in a progressive loss of cell viability and gradual detachment of HPMC from cell monolayers so that, after 6 weeks, the number of cells in GDP-treated cultures was reduced by more than 50% (18). It was also apparent that incubation with GDPs significantly impaired the biosynthetic potential of HPMC. Reduced secretion of IL-6 and fibronectin was detected already within the first 2 weeks of exposure (18). Likewise, a similar decline in cell viability and cytokine secretion was detected when HPMC were exposed “chronically” to culture media supplemented with H-PDF rather than with pure GDPs (19). In contrast, no significant toxicity was associated with the use of PDF sterilized by filtration.

Glucose Degradation and Advanced Glycation End-Product Formation

It has been suggested for some time that chronic contact of glucose-containing dialysate and proteins results in the deposition of advanced glycation end-products (AGEs) on peritoneal tissues. Indeed, increased levels of AGEs were detected in peritoneal effluent from CAPD patients (20). More importantly, immunohistochemistry revealed accumulation of AGEs in the peritoneal membrane, the degree of which was clearly related to the duration of CAPD (21). More recently, increased peritoneal deposition of AGEs was shown to correlate with the development of peritoneal fibrosis and ultrafiltration failure (22). In 1997, Lamb et al. demonstrated that in vitro formation of AGEs in peritoneal effluent obtained from CAPD patients was significantly greater than that in phosphate-buffered saline normalized for glucose and protein concentration (23). These results indicated that factors other than glucose that were present in dialysate could modulate AGE accumulation. Millar and Dawnay observed that heat sterilization of PDF intensified AGE formation (24). Later, Linden et al. demonstrated that 3-deoxyglucosone, a potent promoter of AGE formation, is generated as a GDP during heat sterilization of PDF (4). Schalkwijk et al. showed rapid in vitro AGE formation in response to 3-deoxyglucosone, glyoxal, or methylglyoxal (25). These data indicate that pathological effects of chronic exposure to GDPs are likely to relate to accumulation of AGEs. In this respect, it will be interesting to know if GDPs modulate the function of specific receptors for AGEs, the presence of which was detected in HPMC (26).

Dual-Chambered Bag: Solution to the Problem of Gdps in Peritoneal Dialysis?

It is now well recognized that the presence of GDPs in dialysis solutions is related primarily to the way these fluids are manufactured. The only technology by which bulk volumes of PDF can be effectively sterilized is autoclaving, but this gives rise to GDPs. However, researchers have demonstrated that breakdown of glucose and GDP formation during heat sterilization might be substantially reduced by separating highly concentrated glucose at approximate pH 3.2 from catalyzing electrolytes and buffers at approximate pH 6.5 (7). This led to the development of a PDF bag consisting of two compartments, the contents of which are mixed immediately prior to use. Dialysis solutions in dual-chambered bags are now offered by all major manufacturers (Table 2).

OPEN IN VIEWER

Table 2

Multichambered Bags of Glucose-Containing Dialysis Solutions (with Different Buffering Systems) with Reduced GDP Content

Fluid		Manufacturer
Gambrosol	Bio/Unica Bag System	Gambro
Balance		Fresenius
Bicarbonate	Bic20/30/40	Fresenius
Physioneal		Baxter

Increasing evidence suggests that a biocompatibility profile of these solutions is better than that of conventionally autoclaved PDFs. Dual-chambered PDF appeared to be less toxic to L929 cells and human peripheral monocytes (5). In addition, in vitro formation of AGEs was found to be greater in PDF heat-sterilized in a standard single-compartment bag compared with that of equivalent glucose strength but sterilized in a two-compartment system (27). We recently examined the function of HPMC incubated chronically in media supplemented with differently sterilized PDF. These experiments showed that, while exposure to conventional H-PDF resulted in a progressive reduction in cell viability with time, no significant toxicity was observed in cells treated with either F-PDF or H-PDF sterilized in a dual-chambered bag (28). Taken together, these results indicate that introduction of a new generation of dialysis solutions with low GDP content may be an important step toward more biocompatible peritoneal dialysis fluids.