Advanced Glycosylation End-Products in Diabetic Rats on Peritoneal Dialysis Using Various Solutions

Abstract

Objective

To evaluate and compare the effects of glucose-based solutions to those of icodextrin with respect to peritoneal transport characteristics and advanced glycosylation end-product (AGE) formation in the peritoneal membrane in a diabetic rat model of peritoneal dialysis (PD).

Design

Thirty-three male Sprague–Dawley rats weighing between 275 – 300 g were divided into five groups: group C (n = 6), control rats implanted with a catheter but not dialyzed; group D (n = 5), diabetic rats implanted with a catheter but not dialyzed; group G (n = 7), diabetic rats implanted with a catheter and dialyzed with standard 2.5% glucose solution for daytime exchanges and 4.25% glucose solution for overnight exchanges; group H (n = 8), diabetic rats implanted with a catheter and dialyzed with standard 2.5% glucose solution for daytime exchanges and 7.5% icodextrin solution for overnight exchanges; group I (n = 7), diabetic rats implanted with a catheter and dialyzed with 7.5% icodextrin solution for all exchanges. Dialysis exchanges (25 mL per exchange) were performed three times daily for a period of 12 weeks. Tissue sections were stained using a monoclonal anti-AGE antibody. One-hour peritoneal equilibration tests (PET) were performed every 4 weeks for comparison of transport characteristics.

Results

The level of immunostaining was lowest in group C and highest in group G. Significant differences in immunostaining were seen between group C and group G (p < 0.001), group C and group H (p = 0.001), and group C and group I (p < 0.05). Significant differences were also found between group G and group D (p < 0.05), and between group G and group I (p < 0.05). Over time, the ratio of glucose concentration after 1 hour to glucose concentration at instillation (D/D₀) decreased and the dialysate-to-plasma ratio (D/P) of urea increased. Significant differences in D/D₀ glucose and D/P urea were found between group C and group H (D/D₀: 0.40 ± 0.01 vs 0.35 ± 0.01, p < 0.05; D/P urea: 0.87 ± 0.03 vs 0.97 ± 0.02, p < 0.05).

Conclusions

These results suggest that AGE formation is lower with the use of peritoneal dialysis solution containing icodextrin than with glucose-based solution. We conclude that use of icodextrin may help to slow the deterioration of the peritoneal membrane, prolonging its use for dialysis.

Keywords

Glycosylation advanced glycosylation end-products icodextrin dialysis solutions

Continuous ambulatory peritoneal dialysis (CAPD) has been used to treat patients with end-stage renal failure for about 20 years now. Glucose has been the osmotic agent used in nearly all peritoneal dialysis solutions. Problems associated with the use of glucose include: a short duration of effective ultra-filtration, particularly in high transporters (1); potential for disruption of the peritoneal membrane (2); impairment of peritoneal cellular defense mechanisms (3); and adverse metabolic and nutritional consequences from the daily absorption of 150 – 300 g glucose (4). Dobbie and colleagues suggested the need for a CAPD solution that is iso-osmolar with uremic serum and devoid of the damaging effects associated with glucose solutions (5). Icodextrin was developed in response to these requirements.

Icodextrin is a starch-derived glucose polymer with an average molecular weight of 16,200 Da and with chain lengths varying between 4 and 250 glucose units. Extraneal (Baxter Healthcare Corporation, Castlebar, Ireland), a 7.5% icodextrin-based dialysis solution, has an osmolality of 284 mOsm/kg, and a pH of 5.3 (6). The iso-osmolar Extraneal solution should be less damaging to the peritoneum and to local host defense as compared to hyperosmolar glucose solutions. Extraneal has a colloid-induced osmotic convective transport as compared to the crystalloid osmosis of glucose-based solutions. Extraneal solution provides sustained ultrafiltration over long dwell times of 8 – 12 hours in CAPD (7) and up to 16 hours in automated peritoneal dialysis (8). Owing to these properties, Extraneal can be considered a suitable alternative for glucose-based solutions in the treatment of patients with ultrafiltration failure (9).

Glycosylation refers to a non enzymatic process in which reducing sugars covalently bond to molecules such as peptides and proteins, which contain free amino groups. In vivo, the sugar prominently involved is glucose, and the rate of glycosylation depends primarily on glucose concentration. The carbonyl group of an open-chain glucose molecule reacts with an amino group to form an unstable, reversible Schiff base intermediate. The Schiff base may then undergo an Amadori rearrangement to form a stable ketoamine (Amadori product). This reaction reaches equilibrium in approximately 28 days (10). Over a period of months, the Amadori product subsequently undergoes further chemical reactions involving oxidation, dehydration, cyclization, rearrangement, and condensation with other reactive amino groups to form advanced glycosylation end-products (AGEs), which irreversibly bind to protein (11).

Advanced glycosylation end-products are a heterogeneous group of compounds characterized by the ability to polymerize proteins, a fluorescent spectrum, and a yellow-brown color—hence the term “browning” reaction. Once the initial glycosylation reaction has occurred, polymerization and cross-linking continue even in the absence of glucose (12). The main sites of AGE cross-linking are the long-lived structural proteins, such as those found in tissue matrix and basement membrane. Tissue concentrations of AGEs increase naturally with aging and have been shown to be accelerated in patients with diabetes mellitus as a consequence of hyperglycemia. Elevated serum concentrations of deoxyglucosones, which are reactive intermediates in AGE formation, have been detected in diabetic patients (13,14) and in patients on dialysis (15).

Advanced glycosylation end-products have been implicated in the pathogenesis of some secondary complications of diabetes (16), especially vascular disease. Effects of AGEs may also be mediated through interaction with receptors on endothelial, smooth muscle, mesangial, and inflammatory cells, among others (17). Increased serum AGEs have been shown to contribute to the accelerated vascular dysfunction associated with uremia and diabetes (18,19).

The formation of AGEs in diabetes has been proposed to result in cross-linking of collagen and distortion of subcelluar structures, causing irreversible tissue damage in peripheral nerves, the macrovasculature, and the microvasculature (20,21). Prolonged exposure to the glucose-based dialysis fluids currently in use results in production and deposition of AGEs in the subendothelial area of the peritoneum, microvascular changes similar to those seen in diabetes, and eventually to loss of peritoneal function.

Several in vivo and in vitro studies have demonstrated that use of glucose-based solutions results in higher AGE levels as compared with the levels seen with the use of icodextrin-based solutions. In vivo studies on the formation of AGEs are limited to retina, nerve, aorta, and glomerulus (10,22). Long-term studies of peritoneal dialysis in an animal model, comparing icodextrin with glucose with regard to the formation of AGEs, have not been done. The present study was performed to compare, in vivo, in the rat model of peritoneal dialysis, the effects of glucose-based solutions with those of icodextrin-based solutions—specifically, AGE formation in the peritoneal membrane and subsequent effects on peritoneal transport characteristics.

Materials and Methods

Animals

The study was performed on 42 male Sprague-Dawley rats weighing between 275 g and 300 g, randomly divided into 5 groups:

•

Group C (n = 6), control (catheter but no dialysis);

•

Group D (n = 6), diabetic rats with catheter but no dialysis;

•

Group G (n = 10), diabetic rats dialyzed with standard 2.5% glucose solution (Dianeal: Baxter Healthcare Corporation, Deerfield, IL, U.S.A.) for daytime exchanges (0800 hr, 1400 hr) and with 4.25% glucose solution (Dianeal) for the overnight exchange (2200 hr);

•

Group H (n = 10), diabetic rats dialyzed with standard 2.5% glucose solution for daytime exchanges and with 7.5% icodextrin solution (Extraneal) for the overnight exchange;

•

Group I (n = 10), diabetic rats dialyzed with 7.5% icodextrin solution for all exchanges.

Before completion of the study, 1 animal in group D, 3 animals in each of groups G and I, and 2 animals in group H were euthanized owing to non functional catheters, leaving 33 animals for the analysis.

Induction of Diabetes

Diabetes was induced using streptozotocin (Sigma, St. Louis, MO, U.S.A.) 65 mg/kg dissolved in citrate buffer (0.1 mol/L, pH 4.5) and injected into the tail vein under methoxyflurane anesthesia (Metofane: Pitman-Moore Inc., Madison, NJ, U.S.A.) in all groups except group C (control). Diabetic rats with serum glucose greater than 500 mg/dL were given regular subcutaneous insulin (0.5 - 2 units) twice daily to maintain a serum glucose level between 250 -500 mg/dL.

Catheter Implantation

All catheters were implanted according to a previously described technique (23). When a catheter malfunctioned owing to omental wrapping, that catheter was left in place, and a second catheter was implanted contra-laterally using the same technique.

Dialysis Exchange

Dialysis was started on the first postoperative day, with an instillation volume that was gradually increased from 15 mL to 25 mL. Prophylactic antibiotics (vancomycin 125 mg/L and gentamicin 32 mg/L) were given intraperitoneally in the solution until the fifth postoperative day. The prophylactic antibiotics were changed every 5 days thereafter in this sequence: gentamicin 8 mg/L; vancomycin 25 mg/L; ciprofloxacin 25 mg/L; ceftazidime 125 mg/L; and sulfamethoxazole–trimethoprim, 200 mg/L and 40 mg/L respectively. Dialysis exchanges were performed 3 times daily at 0800 hr, 1400 hr, and 2200 hr, 7 days per week for 3 months.

Peritonitis

Samples for culture were collected once weekly, before instillation of the morning dialysis solution. Diagnosis of peritonitis was based solely on culture results (more than 10 colonies in a blood agar plate) because no accepted definition of peritonitis has been established for rats. The infection was treated according to the sensitivity-test result.

Peritoneal Membrane Function and Blood Tests

Peritoneal membrane function was assessed on day 8 and at 1 month, 2 months, and 3 months by a 1-hour peritoneal equilibration test (PET) using 1.5% glucose solution. A 1.5% glucose solution was used for the overnight exchange preceding the PET.

At time 0, 25 mL of dialysis solution was instilled. At 1 hour, the animal was anesthetized, and a sample of dialysate was taken for measurement of urea nitrogen, creatinine, and glucose. Immediately thereafter, a blood sample (2 mL) was taken by direct cardiac puncture for measurement of urea nitrogen, creatinine, glucose, lipid profile, and hematocrit. In cases of peritonitis, the test was deferred until cultures were negative.

Peritoneal membrane transport rate was assessed by dialysate-to-plasma ratio (D/P) of urea nitrogen and D/D₀ glucose, where D is the glucose concentration in dialysate after the 1-hour dwell, and D₀ is the glucose concentration in the dialysis solution before infusion into the peritoneal cavity. High levels of D/P urea and low levels of D/D₀ glucose indicate high transport.

Weight

All animals were weighed at the time of induction of diabetes, on the 8th day after induction, and at 1, 2, and 3 months after induction. Weighings after induction were carried out before the morning dialysis exchange.

Histologic Analysis

All animals were humanely killed during the 12th week, and specimens for histologic examination were obtained. Peritoneal tissues were taken from four locations: a loop of gut with mesentery, abdominal wall, diaphragm, and surface of liver. The peritoneal membrane was evaluated by gross inspection and light microscopy. Gomori trichrome stain was used to detect collagen deposition (fibrosis) in the peritoneal membrane. This procedure differentially stains connective tissue elements as follows: collagen, green; muscle tissue, red; and nuclei, blue to black.

Immunohistochemical Analysis

At the end of the study, the animals were anesthetized and trans-cardially perfused with phosphate-buffered saline (PBS) followed by 4% paraformaldehyde. Isolated tissues were put into 30% sucrose solution. The tissues were then rapidly frozen with isopentane (2-methylbutane) in liquid nitrogen. Frozen tissue sections were cut sequentially at a 10 μm thickness and thaw-mounted onto slides.

Immunostaining of AGEs in peritoneal tissue sections was conducted using the streptoavidin– biotinylated peroxidase complex method. The sections were washed three times with PBS. Endogenous per-oxidase activity was blocked by incubating the sections with 0.3% H₂O₂ at room temperature for 30 minutes. The sections were washed with PBS and incubated with normal goat serum in PBS at room temperature for 2 hours. After the normal goat serum was removed, the sections were incubated in a humid chamber at room temperature overnight with anti-AGE mouse monoclonal antibody (Dojindo Laboratories, Tokyo, Japan) diluted 100x in PBS. The sections were washed with PBS and incubated with secondary antibody, biotinylated anti-mouse immunoglobulin G [(IgG) Vector Laboratories, Burlingame, CA, U.S.A.], at room temperature for 1 hour. After being washed three times with PBS, the sections were incubated with streptoavidin peroxidase complex (Vector Laboratories) at room temperature for 30 minutes. After being washed three times with PBS, the reaction was completed by the addition of diaminobenzidine for 15 minutes. After being washed with PBS, the slides were counterstained with hematoxylin. Control staining without primary antibody was performed on 2 sections in each group to ascertain that no antibody was present in the tissues.

The slides were read by 5 separate examiners in a blinded fashion and staining intensity was graded semi-quantitatively from 0 to 3, where 0 represented no staining, 1 represented weak staining, 2 represented moderate staining, and 3 represented strong staining.

Statistical Analysis

Data are presented as mean ± standard error of mean (SEM). Statistical analysis of values at various time points among the groups were assessed by oneway analysis of variance (ANOVA), with Tukey correction for multiple comparisons, using the software SigmaStat 2.0 for Windows (SPSS Inc., Chicago, IL, U.S.A.). Differences were considered significant at p < 0.05.

Results

Immunohistochemical Detection of Ages in Peritoneal Membrane

Immunohistochemical staining showed that AGE accumulation was minimal or absent in group C [Figure 1(a)].

In group D, the intensity of staining of the mesothelial layers ranged from very weak to moderate. On the other hand, staining of the vascular wall was weak to moderately positive in all cases. In interstitial and connective tissues, AGEs were not found.

In group G, the mesothelial layers showed a consistent staining pattern ranging from moderate to strong. Staining of the vascular wall was moderately to strongly positive in all cases. No AGEs were found in interstitial or connective tissues [Figure 1(d)].

In group H, the mesothelial layers were stained in varying degrees of intensity ranging from weak to moderate. Staining of the vascular wall was moderately to strongly positive in all cases. In interstitial and connective tissues, AGEs were not found [Figure 1(c)].

In group I, the mesothelial layers showed inconsistent staining patterns ranging from weak to moderate. Staining of the vascular wall was also weakly to moderately positive in all cases. No AGEs were found in interstitial and connective tissues [Figure 1(b)].

Figure 2 shows the scores for AGE staining in the various groups. The immunostaining grade was lowest in group C and highest in group G. Significant differences were seen in comparisons of group C (0.08 ± 0.08) with group G (2.27 ± 0.28, p < 0.001), with group H (1.72 ± 0.19, p < 0.001), and with group I (1.26 ± 0.21, p = 0.008); group G (2.27 ± 0.28) with group D (1.12 ± 0.29, p = 0.016) and with group I (1.26 ± 0.21, p < 0.05); and group C with group D (p < 0.05). A strong correlation was noted between serum glucose concentration and AGE stain score [r = 0.894, Figure 3].

Peritoneal Equilibration Test

Table 1 shows changes in D/D₀ glucose from baseline to 12 weeks. A tendency toward a decrease in D/D₀ glucose over this period was noted. Significant differences in D/D₀ were seen at 12 weeks between group C (0.40 ± 0.01) and group H (0.35 ± 0.01, p < 0.05).

Table 1

Changes in D/D₀ Glucose

Group	1 week	4 weeks	8 weeks	12 weeks
C	0.43±0.03	0.41±0.02	0.40±0.02	0.40±0.01
D	0.42±0.02	0.42±0.01	0.40±0.01	0.39±0.04
G	0.44±0.02	0.41±0.01	0.39±0.01	0.36±0.01
H	0.45±0.01	0.42±0.02	0.38±0.03^a	0.35±0.01
I	0.44±0.02	0.43±0.03	0.40±0.02	0.36±0.01

Data are presented as mean ± standard error of mean (SEM). C = intact control rats; D = diabetic control rats; G = diabetic rats dialyzed with glucose solution; H = diabetic rats dialyzed with glucose and icodextrin; I = diabetic rats dialyzed with icodextrin only.

p < 0.05 C versus H.

Table 2 shows changes in D/P urea over the same time period. A tendency toward an increased D/P urea was noted. At 12 weeks, a significant difference was seen between group C (0.87 ± 0.03) and group H (0.97 ± 0.02, p < 0.05).

Table 2

Changes in D/P Urea

Group	1 week	4 weeks	8 weeks	12 weeks
C	0.83±0.05	0.84±0.03	0.85±0.03	0.87±0.03
D	0.87±0.02	0.88±0.04	0.89±0.04	0.92±0.02
G	0.88±0.02	0.90±0.01	0.92±0.02	0.96±0.04
H	0.88±0.01	0.93±0.03	0.96±0.02^a	0.97±0.02^a
I	0.88±0.02	0.90±0.02	0.93±0.02	0.95±0.01

p < 0.05 C versus H.

Peritonitis Episodes

The incidence of peritonitis was greatest in group G, at 0.25 episodes/rat-month. Group H experienced 0.14 episodes/rat-month, and group I, 0.11 episodes/rat-month.

Body Weight

Table 3 shows changes in weight over time. The mean baseline weight of the rats was approximately 304 g. Weight increased in all groups over the period of study. Significant differences were seen between group C and groups G, H, and I (4 weeks: p = 0.001, p < 0.001, and p = 0.003 respectively; 8 weeks: p < 0.001, p < 0.001, and p < 0.001 respectively; 12 weeks: p < 0.001, p = 0.001, and p < 0.05, respectively). As expected, diabetic rats (groups D, G, H, and I) gained less weight than did non diabetic rats (p < 0.05). Control rats (groups C and D) gained more weight than rats on dialysis (groups G, H, and I; p < 0.05). Rats in group I gained more weight than rats in groups G and H, but the difference was not statistically significant.

Table 3

Changes in Body Weight (grams)

Group	–2 weeks	1 week	4 weeks	8 weeks	12 weeks
C	303.8±2.8	340.0±7.3	396.8±9.1	422.1±12.2	489.6±13.3
D	304.9±3.8	326.1±4.4	377.1±13.7	387.2±10.9	458.5±14.8
G	303.3±2.9	315.5±7.1	342.2±8.1^a	351.9±9.2^a	408.7±9.4^a
H	302.5±1.9	317.4±6.2	348.0±6.2^b	353.2±9.3^b	415.6±10.3^b
I	302.8±1.8	314.8±5.9	343.5±7.4^c	349.7±9.9^c	438.7±12.5^c

p < 0.05 C versus G.

p < 0.05 C versus H.

p < 0.05 C versus I.

Table 4

Changes in Serum Glucose Levels (mg/dL) After Injection of Streptozotocin

Group	1 week	4 weeks	8 weeks	12 weeks
C	149.7±5.6	151.8±9.8	161.5±9.7	184.7±8.9
D	276.0±5.7^a	278.6±17.3	290.2±19.8	286.0±14.8
G	271.9±15.9^b	360.9±19.6^b	487.6±57.9^b,e	418.4±45.0^b
H	272.4±18.9^c	359.13±9.5^c	481.06±2.4^c	384.1±62.7^c
I	273.5±8.9^d	330.02±3.6^d	390.1±30.8^d	345.4±29.4^d

p < 0.05 C versus D.

p < 0.05 C versus G.

p < 0.05 C versus H.

p < 0.05 C versus I.

p < 0.05 D versus G.

Blood Tests

Table 4 shows changes in serum glucose levels over time. Serum glucose levels were maintained between 250 mg/dL and 500 mg/dL. At 1 week, non diabetic rats (group C) had lower serum glucose levels than diabetic rats (groups D, G, H, and I; p < 0.001, p < 0.001, p < 0.001, and p < 0.001, respectively). Significant differences in serum glucose were seen between group C and groups G, H, and I (p < 0.001, p < 0.001, and p < 0.05) at 4 weeks. At 8 weeks, significant differences were seen between group C and groups G, H, and I (p < 0.001, p < 0.001, and p < 0.05), as well as between group D and group G (p < 0.05). At 8 weeks, the serum glucose level in group I was lower than that in groups G and H, but the difference was not statistically significant. Significant differences were seen between group G and group C at 12 weeks (p < 0.05).

Table 5 shows changes in total cholesterol over time. Group G had higher levels of serum cholesterol as compared with the other groups. At 12 weeks, a significant difference was seen between group C and group G (p < 0.05).

Table 5

Changes in Serum Total Cholesterol (mg/dL)

Group	4 weeks	8 weeks	12 weeks
C	62.0±2.5	60.8±6.7	71.3±3.3
D	67.8±4.2	64.2±3.4	71.8±3.2
G	68.3±4.2	102.6±3.8	87.7±3.7^a
H	68.0±3.5	85.8±3.3	84.9±3.2
I	68.0±3.5	86.3±3.4	73.0±4.5

p < 0.05 C versus G.

Table 6 shows changes in serum triglycerides over time. Group G had higher levels of serum triglycerides as compared with the other groups. At 12 weeks, significant differences were seen between group G and groups D and C (p < 0.05 and p < 0.05, respectively).

Table 6

Changes in serum triglycerides (mg/dL)

Group	4 weeks	8 weeks	12 weeks
C	58.0±5.2	58.0±5.2	59.0±6.6
D	58.4±6.0	67.4±7.0	71.4±14.9
G	56.8±6.8	135.5±17.6	136.3±25.9^a,b
H	57.9±4.8	112.4±17.1	114.4±19.9
I	55.4±7.8	79.0±7.8	98.7±27.7

p < 0.05 C versus G.

p < 0.05 D versus G.

Gross Findings on Study Termination

The catheters in groups C and D were completely encased by fibrous tissue, with severe adhesions to the liver and abdominal wall. The abdominal wall and mesentery not directly involved with the catheter were clear.

In group G, two catheters were completely encased in fibrous tissue with severe adhesions. All other catheters in group G were encased, but adhesions were few. Two animals in this group had had a second catheter implanted; the second catheter was moderately covered by fibrous tissue as compared with the first. The abdominal wall in two animals and the mesentery in one animal showed white patchy lesions.

Two catheters in group H were completely encased by fibrous tissue with severe adhesions; all other catheters in this group were covered by fibrous tissue with only some adhesions. Two animals in this group had had a second catheter implanted; the second catheter was moderately covered by fibrous tissue as compared with the first. The abdominal walls of these animals showed white patchy lesions, but the mesentery was relatively clear.

In group I, one catheter was partially encased by fibrous tissue with severe adhesions; the rest of the catheters were encased by fibrous tissue with only some adhesions. One animal had had a second catheter implanted; the second catheter was only moderately covered by fibrous tissue as compared with the original catheter. In one animal, the abdominal wall showed white patchy lesions, but the mesentery was relatively clear.

Histological Findings

Collagen deposition in the peritoneal membrane was not seen in groups C and D, except that 1 animal in group D had mild fibrous tissue in the hepatic visceral peritoneum. All sections in group G had mild-to-moderate fibrous tissue in the hepatic visceral peritoneum; mild fibrosis was noticed in the mesentery of 2 rats from this group. In group H, 5 of 8 sections had mild-to-moderate fibrous tissue in the hepatic visceral peritoneum. In group I, 3 of 7 sections had mild fibrous tissue in the hepatic visceral peritoneum.

Discussion

The results of this study show that AGE accumulation in peritoneal tissue corresponded to the concentration of glucose in peritoneal dialysis fluid. Several in vitro and ex vivo studies have demonstrated that glycosylation and AGE formation occurs with use of conventional glucose-containing solutions (22,24). It has been established that the rate of glycosylation is related to the concentration of glucose in dialysis fluid and that it also depends on other components of the fluid. Determination of the chemistry and other factors involved in AGE formation in glucose-based solutions has been established (25). In this study, AGE accumulation was seen in all diabetic rats to degrees that varied in relation to the type of dialysis solution used.

Information on glycosylation and AGE formation under physiological conditions in the presence of glucose polymers and their final breakdown product, maltose, is scarce. Recently reported in vitro studies have shown a time- and concentration-dependent increase in human serum albumin glycosylation with maltose, icodextrin, and glucose (26). The glycosylation rate was comparable for equimolar maltose and icodextrin, and twofold to threefold lower than with equimolar glucose. The glycosylation rate with the standard 7.5% icodextrin solution used in peritoneal dialysis was more than tenfold lower than the rate with the lowest glucose concentration (1.36%) used in standard peritoneal dialysis solutions. The results of the present study accord with these findings.

Loss of ultrafiltration is typically associated with increased permeability to small solutes such as creatinine (27). This observation could be explained if AGE formation occurred in the peritoneal membrane, resulting in increased permeability to solutes. Increased permeability might occur owing to vascular effects or to basement membrane thickening, as is the situation in diabetes (28). It is also possible that modification of collagen in the submesothelial interstitium by AGEs might diminish diffusion resistance to small solutes and account for a more pronounced increase in permeability to small solutes than to macromolecules (29).

In the present study, rats dialyzed with glucose-based solution had significantly higher permeability to small solutes such as glucose and urea as compared with control groups. Cross-linking of basement membrane components by AGEs such as laminin and heparin sulfate proteoglycan or type IV collagen disturbs the integrity of the basement membrane (30,31), increasing vascular permeability. On other hand, the vascular endothelium has AGE receptors. Deposition of AGEs on the receptors is reported to induce an increase in permeability of the vascular wall (32). Nakayama et al have shown that AGEs accumulate in the vascular walls after continuous ambulatory peritoneal dialysis (CAPD) with glucose-based solutions (33). This deposition might play a role in the increased permeability of the peritoneal membrane.

In the present study, AGE staining was evident in the vascular walls of all rats except those from group C. Accordingly, the accumulation of AGEs in the vascular walls of these rats might possibly play a role in the increased peritoneal permeability shown.

Monitoring changes in body weight over time is a useful way of assessing the general condition of experimental rats. Our rats gained weight throughout the study. These results suggest that the rats were in good condition. Chronic dialysis with glucose solution is associated with various metabolic disadvantages, especially obesity (34). In the present study, diabetic rats gained less weight as compared with non diabetic rats. Rats on dialysis gained less weight than rats not on dialysis. These results suggest that diabetes and dialysis concomitantly induced a stress that resulted in a smaller weight gain.

Prolonged exposure to non physiological dialysis solutions results in certain inevitable morphologic changes, mainly peritoneal fibrosis (35). Morphologic analysis of the peritoneal membrane was introduced as a useful way of assessing peritoneal fibrosis (36). Our results showed that the group using hyperosmolar glucose solutions had more episodes of peritonitis. The fibrous tissue covering the catheters of rats dialyzed with glucose solution was thicker and showed more adhesion to the abdominal wall as compared with other groups. In addition, whitish fibrotic patchy lesions were comparatively more frequent in rats dialyzed with glucose-based solution. These clinical observations showed that dialysis with higher concentrations of glucose resulted in more peritonitis episodes and was more irritating to the peritoneal membrane, producing peritoneal fibrosis and catheter blockage.

Maltose and related fractions neither affect glucose metabolism nor lead to hyperinsulinemia (6). Furthermore, the lack of hyperglycemia and hyperinsulinemia associated with icodextrin as compared with glucose may offer some long-term metabolic advantages and possibly simplify management of diabetic patients (6). A favorable impact on lipid profile is possible. Cholesterol levels were lower in the icodextrin group as compared with the glucose-based groups in the present study, but the differences did not reach statistical significance. However, serum glucose levels also increased in the icodextrin group in this study, though not to the extent seen in glucose-based groups. This observation requires further evaluation.

Conclusions

The present study demonstrates that AGE accumulation in the peritoneal tissues was lower in rats dialyzed with icodextrin as compared to those dialyzed with glucose-based solutions. At present, in clinical practice, icodextrin solution is used in conjunction with glucose-based solutions. The exclusive use of icodextrin solution holds promise and needs further evaluation. Pharmacological intervention may inhibit or reduce AGE formation. Eventually, this study of the pharmacological, mechanistic, and morphological effects of peritoneal membrane exposure to various compounds may provide insights into new therapeutic strategies for the prevention of deterioration of peritoneal function during peritoneal dialysis.

Footnotes

Acknowledgments

The authors wish to express their gratitude to J.T. Cunningham of the Department of Physiology and Dalton Cardiovascular Research Center, University of Missouri, for technical assistance with internal fixation and cryosection of animal tissues. The authors also gratefully acknowledge Carol S. Okamura of the Department of Veterinary Medicine and Surgery, Dalton Cardiovascular Research Center, University of Missouri, for technical assistance with immunohistochemical staining.

References

Feriani

, LaGreca

, Kriger

F.L.

, Winchester

J.F.

CAPD systems and solutions. In: Gokal

, Nolph

K.D.

, eds. The textbook of peritoneal dialysis. London: Kluwer Academic Publishers; 1994: 233–70.

Honda

, Nitta

, Horita

, Yumura

, Nihei

Morphological changes in the peritoneal vasculature of patients on CAPD with ultrafiltration failure.

Nephron 1996; 72: 171–6.

de Fijter

C.W.H.

, Oe

P.L.

, Donker

A.J.M.

, Verhoef

, Verbrugh

H.A.

Compatibility of peritoneal dialysis fluids containing alternative osmotic agents with cells present in the peritoneal cavity.

Perit Dial Int 1994; 14(Suppl 2): S33–8.

Hain

, Kessel

Aspects of new solutions for peritoneal dialysis.

Nephrol Dial Transplant 1987; 2: 67–72.

Dobbie

, Lloyd

J.K.

, Gall

C.A.

Categorization of the ultrastructural changes in peritoneal mesothelium, stroma and blood vessels in uremia and CAPD patients.

Adv Perit Dial 1990; 6: 3–12.

Davies

D.S.

Kinetics of icodextrin.

Perit Dial Int 1994; 14(Suppl 2): S45–50.

Mistry

C.D.

, Gokal

, Peers

E.M.

, and the MIDAS study group. A randomized multicenter clinical trial comparing isoosmolar icodextrin with hyperosmolar glucose solution in CAPD. Kidney Int 1994; 46: 496–503.

Posthuma

, ter Wee

P.M.

, Donker

A.J.M.

, Oe

L.P.

, Verbrugh

H.A.

, Peers

Improved ultrafiltration in CCPD patients using icodextrin (I) instead glucose (G) during the long daytime dwell (Abstract).

J Am Soc Nephrol 1995; 6: 513, Abstract 406.

Krediet

R.T.

, Ho-dac-Pannekeet

M.M.

, Imholz

A.L.

, Struijk

D.G.

Icodextrin's effects on peritoneal transport.

Perit Dial Int 1997; 17: 35–41.

10.

Dawnay

A.B.S.

, Millar

D.J.

Glycation and advanced glycation end-product formation with icodextrin and dextrose.

Perit Dial Int 1997; 17: 52–8.

11.

Bucala

, Vlassara

, Cerami

Advanced glycosylation end-products. In: Harding

J.J.

, Crabbe

M.J.C.

, eds. Post-translational modifications of proteins. Boca Raton: CRC Press; 1992: 53–79.

12.

Eble

A.S.

, Thorpe

S.R.

, Baynes

J.W.

Nonenzymatic glycosylation and glucose-dependent cross-linking protein.

J Biol Chem 1983; 258: 9406–12.

13.

Wells–Knecht

K.J.

, Lyons

T.J.

, McCance

D.R.

, Thorpe

S.R.

, Feather

M.S.

, Baynes

J.W.

3-Deoxyfructose concentrations are increased in human plasma and urine in diabetes.

Diabetes 1994; 443: 1152–6.

14.

Yamada

, Miyata

, Igaki

, Yatabe

, Miyauchi

, Ohara

Increase in 3-deoxyglucosone levels in diabetic rat plasma.

J Biol Chem 1994; 269: 20275–80.

15.

Niwa

, Miyazaki

, Momoi

, Ogawa

, Maeda

3-Deoxyglucosone, a reactive intermediate of Maillard reaction, accumulating in uremic serum modifies β₂-microglobulin with advanced glycation end-products (Abstract).

J Am Soc Nephrol 1995; 6: 554.

16.

Vlassara

, Bucala

, Striker

Pathogenic effects of advanced glycosylation: Biochemical, biologic, clinical implications for diabetes and aging.

Lab Invest 1994; 70: 138–51.

17.

Vlassara

Recent progress on the biologic and clinical significance of advanced glycosylation end-products.

J Lab Clin Med 1994; 124: 19–30.

18.

Makita

, Bucala

, Rayfield

E.J.

, Friedman

E.A.

, Kaufman

A.M.

, Korbet

S.M.

Reactive glycosylation endproducts in diabetic uraemia and treatment of renal failure.

Lancet 1994; 343: 1519–22.

19.

Vlassara

Serum advanced glycosylation end-products: A new class of uremic toxins?

Blood Purif 1994; 12: 54–9.

20.

Kennedy

, Baynes

J.W.

Non-enzymatic glycosylation and the chronic complications of diabetes. An overview.

Diabetologia 1984; 26: 93–8.

21.

Brownlee

, Cerami

, Vlassara

Advanced products of nonenzymatic glycosylation and the pathogenesis of diabetic vascular diseases.

Diabetes Metab Rev 1988; 4: 437–51.

22.

Dobbie

J.W.

Advanced glycosylation end products in peritoneal tissue with different solutions.

Perit Dial Int 1997; 17(Suppl 2): S27–30.

23.

Moore

H.L.

A chronic catheter model for PD exchanges in the ambulatory rat (Abstract).

Perit Dial Int 1992; 12(Suppl 1): S141.

24.

Lamb

, Cattell

W.R.

, Dawnay

Glycated albumin in serum and dialysate of patients on continuous ambulatory peritoneal dialysis.

Clin Sci 1993; 84: 619–26.

25.

Dawney

Advanced glycation end products in peritoneal dialysis.

Perit Dial Int 1996; 16(Suppl 1): S50–3.

26.

Millar

D.J.

, Dawnay

A.B.S.

Glycation of albumin with icodextrin (Abstract).

J Am Soc Nephrol 1995; 6: 551.

27.

Struijk

D.G.

, Krediet

R.T.

, Koomen

G.C.M.

, Hoek

F.J.

, Boeschoten

E.W.

, Reijden

H.J.

Functional characteristics of the peritoneal membrane in long-term continuous ambulatory peritoneal dialysis.

Nephron 1991; 59: 213–20.

28.

Tooke

J.E.

Microcirculation and diabetes.

Br Med Bull 1989; 45: 206–23.

29.

Rippe

Pathophysiological description of the ultra-structural changes of the peritoneal membrane during long-term continuous ambulatory peritoneal dialysis.

Blood Purif 1994; 12: 211–20.

30.

Tsilibary

E.C.

, Charonis

A.S.

, Reger

L.A.

, Wohlhueter

R.M.

, Furcht

L.T.

The effect of nonenzymatic glycosylation on the binding of the main noncollagenous NC1 domain to type IV collagen.

J Biol Chem 1988; 263: 4302–8.

31.

Charonis

A.S.

, Reger

L.A.

, Dege

J.E.

, Kouzi–Koliakos

, Furcht

L.T.

, Wohlhueter

R.M.

Laminin alterations after in vitro nonenzymatic glycosylation.

Diabetes 1990; 39: 807–14.

32.

Esposito

, Gerlach

, Brett

, Stern

, Vlassara

Endothelial receptor–mediated binding of glucose-modified albumin is associated with increased mono-layer permeability and modulation of cell surface coagulant properties.

J Exp Med 1989; 170: 1397–407.

33.

Nakayama

, Kawaguchi

, Yamada

, Hasegawa

, Takazoe

, Katoh

Immunohistochemical detection of advanced glycosylation end products in the peritoneum and its possible pathophysiological role in CAPD.

Kidney Int 1997; 51: 182–6.

34.

Jörres

, Topley

, Gahl

G.M.

Biocompatibility of peritoneal dialysis fluids (Editorial).

Artif Organs 1992; 15: 79–83.

35.

Dobbie

J.W.

Pathogenesis of peritoneal fibrosing syndromes (sclerosing peritonitis) in peritoneal dialysis.

Perit Dial Int 1992; 12: 14–27.

36.

Frazier–Jessen

M.R.

, Kovacs

E.J.

Abdominal wall thickness as a means of assessing peritoneal fibrosis in mice.

J Immunol Methods 1993; 162: 115–21.