Fluid reabsorption in proximal convoluted tubules of mice with gene deletions of claudin-2 and/or aquaporin1

Abstract

Deletions of claudin-2 (Cldn2) and aquaporin1 (AQP1) reduce proximal fluid reabsorption (PFR) by about 30% and 50%, respectively. Experiments were done to replicate these observations and to determine in AQP1/claudin-2 double knockout mice (DKO) if the effects of deletions of these established water pores are additive. PFR was determined in inactin/ketamine-anesthetized mice by free-flow micropuncture using single-nephron I(125)-iothalamate (io) clearance. Animal means of PFR [% of glomerular filtration rate (GFR)] derived from TF/Piothalamate ratios in 12 mice in each of four groups [wild type (WT), Cldn2(-/-), AQP1(-/-), and DKO) were 45.8 ± 0.85 (51 tubules), 35.4 ± 1 (54 tubules; P < 0.01 vs. WT), 36.8 ± 1 (63 tubules; P < 0.05 vs. WT), and 33.9 ± 1.4 (69 tubules; P < 0.01 vs. WT). Kidney and single-nephron GFRs (SNGFR) were significantly reduced in all mutant strains. The direct relationship between PFR and SNGFR was maintained in mutant mice, but the slope of this relationship was reduced in the absence of Cldn2 and/or AQP1. Transtubular osmotic pressure differences were not different between WT and Cldn2(-/-) mice, but markedly increased in DKO. In conclusion, the deletion of Cldn2, AQP1, or of both Cldn2 and AQP1 reduces PFR by 22.7%, 19.6%, and 26%, respectively. Our data are consistent with an up to 25% paracellular contribution to PFR. The reduced osmotic water permeability caused by absence of AQP1 augments luminal hypotonicity. Aided by a fall in filtered load, the capacity of non-AQP1-dependent transcellular reabsorption is sufficient to maintain PFR without AQP1 and claudin-2 at 75% of control.

Keywords: iothalamate; micropuncture; paracellular; transcellular; tubular fluid osmolarity.

Fig. 1.

Urine osmolarity (left) and urine/plasma ratio for iothalamate (U/P_iothalamate) (right) in WT, Cldn2^−/−, AQP1^−/−, and DKO mice. Symbols are animal means, and lines indicate average values. Urine osmolarity was determined in bladder urine aspirated during animal preparation, while U/P_iothalamate was measured in urine collected during the micropuncture period. Significances are given for comparison with WT (ANOVA with Bonferroni post hoc test with P < 0.05 significance level).

Fig. 2.

Fractional reabsorption along the proximal tubule accessible for micropuncture in WT mice and in Cldn2^−/−, AQP1^−/−, and DKO mice. A: tubular fluid/plasma iothalamate ratios (TF/P_iothalamate) measured at the end of the proximal convoluted tubules; symbols represent individual tubules. Solid lines are mean values, and dotted lines indicate 95% confidence intervals. Values in the last column (all WT) represent all measurements in wild-type mice from different experimental groups performed over the past 4 years in this laboratory. Numbers in parentheses are numbers of mice/tubules. B: mean fractional fluid reabsorption expressed as % of GFR in the four experimental groups and in the wild-type control group (All WT). Values are expressed as means ± SD. Significances are indicated for comparisons with WT (ANOVA with Bonferroni post hoc test with P < 0.05 significance level).

Fig. 3.

Glomerular filtration rate (GFR) in single nephrons (SNGFR) and fluid reabsorption rate in WT mice and in Cldn2^−/−, AQP1^−/−, and DKO mice. All WT refers to all measurements in wild-type mice from different experimental groups performed over the past 4 years in this laboratory. A: SNGFR using flow rates at the end of the proximal convoluted tubules; symbols represent individual tubules. Solid lines are mean values, and dotted lines indicate 95% confidence intervals. B: rate of fluid reabsorption in the four experimental groups and in the wild-type control group (All WT). Values are expressed as means ± SD. Significances are indicated for comparisons with WT (ANOVA with Bonferroni post hoc test at P < 0.05 significance level).

Fig. 4.

Relationship between kidney GFR (μl/min) and mean SNGFR for the 48 mice used in this study. Each symbol represents one animal without consideration of genotype. The line shows the linear regression function.

Fig. 5.

Relationship between rates of proximal fluid reabsorption and glomerular filtration rate (SNGFR) in WT mice (A), and in Cldn2^−/− (B), AQP1^−/− (C), and DKO mice (D), as well as for the overall wild-type group (E). F: direct comparison of the linear regression lines is shown: regression lines for the wild-type cohorts and for the three mutant strains fall on top of each other; the ovals indicate normal operating ranges for WT and mutant mice. Solid lines are linear regressions and dotted lines are 95% confidence intervals.

Fig. 6.

TF/P_iothalamate ratios in individual WT, Cldn2^−/−, AQP1^−/−, and DKO mice (labeled 1–12 in each genotype). Bars show means ± SD. Numbers inside the bars are numbers of tubules in each animal. Dotted line indicates the mean for each genotype, and stippled lines reproduce the mean of WT mice. Lines above the bars extend between mice that are significantly different from each other (ANOVA with Bonferroni post hoc test with P < 0.05 or less). Stars inside bars indicate significance vs. mean WT at P < 0.05 or less (ANOVA with Bonferroni post hoc test).

Fig. 7.

Nephron filtration rates (SNGFR) in individual WT, Cldn2^−/−, AQP1^−/−, and DKO mice (labeled 1–12 in each genotype). Bars show means ± SD. Numbers inside the bars are numbers of tubules in each animal. Dotted line indicates the mean for each genotype, and stippled lines reproduce the mean of WT mice. Lines above the bars extend between mice that are significantly different from each other (ANOVA with Bonferroni post hoc test with P < 0.05 or less). Stars inside bars indicate significance vs. mean WT at P < 0.05 or less (ANOVA with Bonferroni post hoc test).

Fig. 8.

TF/P_iothalamate ratios (top), proximal fluid reabsorption in % of GFR (middle), and SNGFR (bottom) in WT, Cldn2^−/−, AQP1^−/−, and DKO mice, as well as in WT mice of various strains (All WT). Each symbol is the average of 2–7 measurements in an individual animal (n = 12 in each genotype). Lines indicate mean values. Significances are indicated for comparisons with WT (ANOVA with Bonferroni post hoc test).

Fig. 9.

Transepithelial osmotic gradients, the differences between plasma and tubular fluid osmolarities, in WT, Cldn2^−/−, and DKO mice. Bars show means, and vertical lines indicate SE. The right columns labeled WT and AQP1^−/− are from an earlier study in our laboratory (42). The factor by which the osmotic difference in AQP1-deficient strains exceeds respective WT values is indicated above the bars.

Fig. 10.

Expression of mRNAs of proximal claudins (left) and aquaporins (AQ) (right) in renal cortical tissue of wild-type mice. Data are expressed as a percentage of claudin-2 and aquaporin1, respectively. Data are means, and vertical lines indicate standard deviations (n = 10).

Fig. 11.

Expression of mRNAs of claudins (upper) and aquaporins (lower) in Cldn2^−/− (n = 10), AQP1^−/− (n = 10), and DKO mice (n = 10). Data are expressed using wild-type expression levels for each claudin or aquaporin as 100% reference. Data are means, and vertical lines are standard deviations. *P < 0.05, **P < 0.01 (compared to WT by t-test).

Fig. 12.

Diagrammatic data summary showing magnitudes and pathways of absolute rates of fluid reabsorption along the proximal convoluted tubule calculated from SNGFR and fractional reabsorption for WT, Cldn2^−/−, DKO, and AQP1^−/− strains. Data for TF/P_osm in AQP1^−/− mice are from our earlier work (41). L, tubular lumen; I, interstitium.