Aquaporin-1 plays an essential role in water permeability and ultrafiltration during peritoneal dialysis.

The water channel aquaporin-1 (AQP1) is considered as the molecular counterpart of the ultrasmall pore predicted by the three-pore model of fluid transport across the peritoneal membrane. However, the definitive proof of the implication of AQP1 in solute-free water transport, sodium sieving, and ultrafiltration (UF) during peritoneal dialysis (PD) is lacking, and the effects of its deletion on the structure of the membrane are unknown. Using real-time reverse transcriptase-polymerase chain reaction and immunogold electron microscopy, we showed that AQP1 is the most abundant member of the AQP gene family expressed in the mouse peritoneum, and the only one located in the capillary endothelium. Transport studies during a 2-h dwell demonstrated that, in comparison with Aqp1(+/+) littermates, Aqp1(-/-) mice had no sodium sieving; an approximately 70% decrease in the initial, solute-free UF; and an approximately 50% decrease in cumulative UF. These modifications occurred despite unchanged osmotic gradient and transport of small solutes in the Aqp1(-/-) mice. Heterozygous Aqp1(+/-) mice showed intermediate values in sodium sieving and initial UF, whereas cumulative UF was similar to Aqp1(+/+) mice. The deletion of AQP1 had no effect on the expression of other AQPs and on the density, structure, or diameter of peritoneal capillaries. These data provide direct evidence for the role of AQP1 during PD. They validate essential predictions of the three-pore model: (i) the ultrasmall pores account for the sodium sieving, and (ii) they mediate 50% of UF during a hypertonic dwell.

Effects of glomerular filtration rate on Ficoll sieving coefficients (theta) in rats.

The purpose of the present study was to assess the role of diffusion and convection during filtration of Ficoll across the glomerular filter by comparing glomerular sieving coefficients (theta) to neutral fluorescein isothiocyanate (FITC)-Ficoll 70/400 obtained at low (hydropenic) vs raised (normal) glomerular filtration rates (GFRs). The theta for FITC-Ficoll was determined in anesthetized Wistar rats (304 +/- 18 g) following laparotomy and cannulation of the ureters, used for urine sampling. After surgery, GFR was 1.2 +/- 0.16 ml/min (+/- s.e.), assessed using the plasma to urine clearance of FITC-inulin and (51)Cr-ethylenediaminetetraacetic acid. FITC-Ficoll 70/400 was infused intravenously (i.v.) following an initial bolus dose. To raise GFR, to an average of approximately 2 ml/min, 5 ml of serum together with glucagon (3 microg/min) was given i.v. FITC-inulin and FITC-Ficoll were determined in plasma and urine using size-exclusion high-performance liquid chromatography. The theta for Ficoll as a function of Stokes-Einstein radius was significantly reduced in the range of 13-43 A when GFR was raised. The maximal theta lowering effect, in relative terms, of raising GFR was obtained for a Ficoll a(e) of approximately 32 A. For Ficoll(36 A) (cf. albumin), theta was reduced from 0.111+/- 0.009 to 0.081+/- 0.012 (P < 0.05; n = 7) for the GFR increment imposed. The reduction in theta for Ficoll after raising GFR indicates the presence of a high diffusive component of glomerular Ficoll filtration in rats in vivo and contradicts the notion of a significant concentration polarization effect in the glomerular filter upon Ficoll molecules < 50 A in radius.

Very high daily intraperitoneal doses of carbonyl compounds affect the morphology, but not the exchange characteristics, of rat peritoneum.

Glucose degradation products (GDP) are carbonyl compounds, that are formed by heat sterilization of conventional peritoneal dialysis (PD) fluids. Carbonyl compounds are known to be toxic in vitro and potentially toxic also in vivo. The aim of this study was to evaluate the effects of daily, short-term exposure of the peritoneum to very high concentrations of GDP in vivo on peritoneal transport parameters and on peritoneal morphology in a well-established rat model of PD. Rats were exposed to three daily intraperitoneal (IP) injections (10 ml) for 9 days of a largely neutral (pH 7.2) PD fluid containing 1.5% glucose and sterilized by filtration, with (n = 8) or without (n = 8) the presence of different carbonyl compounds in concentrations 100 times higher than those reported in commercial PD fluids. Seven rats, not subjected to any exposure, served as controls. After the exposure, the rats were subjected to acute PD in 4-hour dwells. Twenty milliliters of 4% glucose dialysis fluid were instilled into the rat peritoneal cavity. Blood and dialysate samples were taken during the dwell for measurements of dialysate sodium, and for assessments of the mass transfer area coefficient (PS) for glucose and 51Cr-EDTA and of transperitoneal clearance (Cl) or radiolabelled albumin (RISA). At the end of the dwell, parts of the liver, diaphragm and peritoneum were removed for measurements of tissue cell density and thickness of the submesothelial peritoneal tissue. The exposure of the peritoneum to very high doses of carbonyl compounds did not affect the peritoneal transport of fluid and small solutes significantly, but seemed to slightly reduce lymph flow and albumin clearance out of the peritoneal cavity. Assessed after a hypertonic dwell, and compared to the situation in nontreated rats after the same kind of dwell, there was a significant thinning of the submesothelial tissue, but no difference in tissue cell density. It is concluded that short-term exposure of the peritoneum in vivo to very high doses of GDP resulted in almost no signs of acute toxicity.

In vivo microvascular clearance of albumin in renal and extrarenal tissues in puromycin aminonucleoside (PAN) induced nephrotic syndrome.

BACKGROUND: The nephrotic syndrome is characterized by generalized oedema considered to be due to the fall in serum albumin and to sodium retention. The aim of the present study was to investigate whether a generalized disturbance in vascular integrity contributes to the oedema formation. METHODS: We used the PAN-(puromycin aminonucleoside) nephritis model in order to induce the nephrotic syndrome in female Wistar rats. Eight rats were given PAN, 15 mg/100 g body weight, intraperitoneally 10 days prior to the study, whereas 21 rats served as controls. Albumin clearance to tissues was measured using a dual isotope technique. Repeated blood samples as well as samples from various muscles, kidney, liver, lung, heart, abdominal wall and from ascites fluid were taken to determine radioactivity and tissue dry-to-wet weights. Clearance of albumin (Clalb) from plasma to interstitium was calculated from the (linear) increment in 'plasma equivalent tissue albumin space' as a function of time, corrected for intravascular volume and oedema. The plasma and urine concentrations of albumin were determined in a parallel study by single radial diffusion using monospecific rabbit anti-rat antiserum in seven PAN animals and 13 controls. RESULTS: A marked fall in dry-to-wet weight ratios together with pronounced proteinuria, oedema and ascites were found in the PAN animals. Haematocrit decreased from 45% (32-51) to 30% (28-38) and serum albumin from 22.0 g/l (16.3-25.2) to 4.94 g/l (3.20-6.72) in control and PAN animals, respectively. However, Clalb apparently remained unchanged in the PAN animals in comparison to controls in most tissues examined. Thus, in these in vivo experiments there was no direct evidence of an increased extravasation of albumin in extrarenal tissues. CONCLUSIONS: There was no strong support for the contention that a generalized disturbance of capillary integrity outside the renal vasculature would contribute to the oedema formation in the PAN nephrotic syndrome.

Effects of acidity, glucose degradation products, and dialysis fluid buffer choice on peritoneal solute and fluid transport in rats.

OBJECTIVE: To evaluate the effects of acidity, glucose degradation products (GDP), and different solution buffer systems on solute and fluid transport during acute peritoneal dialysis (PD) in rats. DESIGN: Dialysis fluid (16 mL) containing 2.5% glucose as the osmotic agent was instilled intraperitoneally in Wistar rats (280 g) via a thin catheter in dwells lasting 4 hours. Blood and dialysis fluid samples (25 microL) were taken for measurement of glucose, sodium, and radioactive markers. The mass transfer area coefficient (MTAC or PS) for glucose and for 51Cr-EDTA (given as an intravenous infusion) and the peritoneal clearance (Cl) of 125I albumin (RISA), as well as the clearance of RISA to plasma (Cl --> P) were assessed for a commercial, heat-sterilized, acidic PD solution (2.5% glucose, pH 5.5; Gambrosol, Gambro, Lund, Sweden), containing GDP, and for four filter-sterilized solutions containing either lactate (40 mmol/L, pH 5.5 or 7.2), bicarbonate (38 mmol/L, pH 7.2), or pyruvate (40 mmol/L, pH 7.2) as buffers and being devoid of GDP. RESULTS: The initial pH of the acidic solutions increased rapidly, and attained physiological levels within 40 minutes. The initial drop of sodium, which is expected during the first part of the dwell, was significantly more pronounced with neutral than with acidic lactate. The PS for glucose and 51Cr-EDTA were slightly, but significantly, higher with the acidic and heat-sterilized solution (Gambrosol) than with the neutral, sterile-filtered lactate-buffered solution (p < 0.01), especially early during the dwell. Such an increase may be due to initial vasodilatation, and hence, recruitment of capillaries by the combination of acidity and GDP. However, there were no significant differences with respect to small solute PS values among sterile-filtered solutions, regardless of the presence of acidity or of buffer choice. CONCLUSION: There were no major differences in fluid and solute transport among sterile-filtered PD solutions having differing buffer systems and pH. Neither were there any effects of GDP alone. However, the combination of a low pH and the presence of GDP in the PD solutions seemed to cause significant increases in peritoneal small solute transport.

Transport of tracer albumin from peritoneum to plasma: role of diaphragmatic, visceral, and parietal lymphatics.

Using a technique to acutely seal off various parts of the peritoneal membrane surface, with or without evisceration, we investigated the role of diaphragmatic, visceral, and parietal peritoneal lymphatic pathways in the drainage of 125I-labeled albumin (RISA) from the peritoneal cavity to the plasma during acute peritoneal dialysis in artificially ventilated rats. The total RISA clearance out of the peritoneal cavity (Cl) as well as the portion of this Cl reaching the plasma per unit time (Cl--> P) were assessed. Under non-steady-state conditions, the Cl was fivefold higher than the Cl--> P. Evisceration caused a 25-30% reduction in both Cl--> P and Cl. Sealing of the diaphragm, however, reduced the Cl--> P by 55% without affecting the Cl. A further reduction in the Cl--> P was obtained by combining sealing of the diaphragm with evisceration, which again markedly reduced the Cl. However, the greatest reduction in the Cl was obtained when the peritoneal surfaces of the anterior abdominal wall were sealed off in eviscerated rats. The discrepancy between the Cl and the Cl--> P can be explained by the local entrance of fluid and macromolecules into periabdominal tissues, where fluid is rapidly absorbed through the capillary walls via the Starling forces, while macromolecules are accumulating due to their very slow uptake by tissue lymphatics under non-steady-state conditions. Of the portion of the total Cl that rapidly entered the plasma, conceivably by lymphatic absorption, 55% could be ascribed to diaphragmatic lymphatics 30% to visceral lymphatics, and only some 10-15% to parietal lymphatics.