Tubular sodium handling and tubuloglomerular in experimental diabetes mellitus CAROL








Department of Medicine, University of Sydney, Concord Hospital, Concord, New South Wales 2139, Australia

POLLOCK,~AROLA., JAMES R. LAWRENCE,ANDMICHAEL J. FIELD. Tubular sodium handling and tubuloglomerular feedback in experimental diabetes mellitus. Am. J. Physiol. 260 (Renal Fluid Electrolyte Physiol. 29): F946-F952, 1991.-Tubular Na handling and tubuloglomerular feedback (TGF) activity were assessed using micropuncture techniques during the hyperfiltration phase of streptozotocin-induced diabetes mellitus in Sprague-Dawley rats. Three animal groups were studied, designated as having severe diabetes [blood sugar level (BSL) S-25 mmol/l], moderate diabetes (BSL 13-18 mmol/l) and control (BSL < 10 mmol/l). Single-nephron glomerular filtration rate (SNGFR) measured at both late proximal (LP) and early distal (ED) sites was elevated in severe diabetes compared with both other groups. TGF activity, determined as the difference between LP and ED measurements of SNGFR, was significantly increased in severe diabetes (46.4 t 6.6 vs. 30.1 t 6.5 vs. 14.8 t 1.9 nl/min). Tubular Na transport was higher in severe diabetes compared with control, as demonstrated by a decrease in fractional delivery of Na to the LP (42.9 t 3.0 vs. 52.9 t 1.9%), as well as to the ED site (4.5 t 0.4 vs. 12.3 & 0.9%). Administration of phlorizin to severely diabetic animals resulted in a BSL comparable to that observed in moderate diabetes, and whole animal GFR, as well as SNGFR, TGF activity, and tubular Na handling were also similar to those found in moderate diabetes. Studies performed during mannitol infusion demonstrated that osmotic diuresis alone was not associated with the changes in TGF and tubular Na handling observed in the diabetic state. These data suggest that the hyperfiltration occurring in early diabetes is associated with enhanced proximal and loop resorption of Na independent of Na-glucose cotransport and osmotic diuresis. Activation of TGF serves to limit the rise in GFR, which results from factors as yet unrecognized in the diabetic state. hyperfiltration;



sodium transport

MAJOR ADVANCES in the management of insulin-dependent diabetes mellitus, diabetic nephropathy remains a significant cause of morbidity and mortality. One of the earliest manifestations in the development of the nephropathy is an elevation in glomerular filtration rate (GFR) (18) accompanied by renal hypertrophy. As tubuloglomerular feedback (TGF) plays an important role in the normal regulation of GFR and the signal for activation or suppression of this feedback response is NaCl delivery sensed at the macula densa (24), changes in feedback activity or tubular Na transport may result in significant alterations in GFR. As an increased cell Na concentration has been implicated in cell hypertrophy and growth (3l), alterations in Na transport may be DESPITE





involved in the hypertrophy associated with the early hyperfiltration observed in diabetes me11itus. The present experiments were therefore designed to test the hypothesis that TGF and tubular Na handling are altered during the hyperfiltration phase that occurs after experimental induction of diabetes mellitus in rats. Increased Na-glucose cotransport in the proximal tubule has been implicated in the pathogenesis of diabetic hyperfiltration (16, 32), and blockade of this cotransport system with phlorizin has been reported to prevent the renal hypertrophy from occurring (30). Furthermore, increased tubular flow rate in itself may alter tubular Na handling. We therefore additionally studied Na transport and TGF responses during phlorizin administration in diabetic rats and during mannitol infusion in nondiabetic rats to distinguish the effects of glycosuria and osmotic diuresis from those of the “diabetic state.” METHODS

Experiment 1: Diabetes Mellitus Animal preparation. Male Sprague-Dawley rats, 6-8 wk of age at the initiation of experimentation, were studied. Diabetes mellitus was induced by 60 mg/kg streptozotocin (STZ), administered intravenously in citrate buffer (pH 4.2) via the tail vein. Injection of citrate alone served as a control for these experiments. In an initial group of 32 animals, ambulatory GFR was measured, as inulin clearance, for 4 days before and 810 days after the injection of STZ or buffer. This was achieved using a continuous microinfusion of [“Cl inulin delivered via an osmotic minipump (Alzet model 2002) subcutaneously placed under light ether anesthesia (14). Briefly, the minipump reservoir was filled with 200-220 ~1 of [14C]inulin, in a concentration of 1 &i/l0 ~1, which was delivered at a constant rate of 0.4-0.6 pi/h. Inulin was allowed to equilibrate in extracellular fluid for 2436 h, after which the study was started. Animals were placed in metabolic cages, and daily food intake, water consumption, and urine output were monitored. Blood from the tail artery (90 ~1) was sampled daily under light ether anesthesia and was analyzed for inulin concentration and blood sugar. Urinary inulin concentration was also determined, with completeness of urinary inulin collection ensured by daily irrigation of the cage with 20 ml distilled water and subsequent collection for inulin content. Fifty microliters of plasma, 20 ~1 urine, and 20 ~1 irrigating solution were assessed for inulin content

0 1991 the American



Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} ( on July 30, 2018. Copyright © 1991 American Physiological Society. All rights reserved.




with urinary inulin excretion calculated as the sum of urinary and irrigant inulin content. GFR was then calculated as the 24-h inulin clearance. The technique was validated by comparing the GFR assessed by this method with that determined in the same animal by the clearance of intravenously infused inulin at the time of micropuncture (Fig. 1). After being monitored for 4 days in metabolic cages, nonfasted animals were randomly allocated to receive either STZ or buffer, which was administered under light ether anesthesia. After injection of STZ all animals developed diabetes of variable severity within 48 h. Of the 32 animals undergoing continuous GFR monitoring, 11 had severe diabetes as judged by marked hyperglycemia [blood sugar level (BSL) 18-25 mmol/l] with insulin administered in 2 animals on single occasions when ketosis supervened. Fourteen animals had moderate diabetes (BSL 13-18 mmol/l) either spontaneously (n = 6) or with the addition of varying amounts of insulin (n = 8). As no difference in food or water intake, hematocrit, serum Na, Na balance, GFR, or subsequent micropuncture data was observed, these animals were analyzed as a single group. Na balance was assessed in an analogous group of animals identically prepared, except that a minipump was not inserted. Daily Na intake and excretion, as well as plasma Na and hematocrit, were assayed for 4 days before and 8-10 days after the injection of STZ or buffer. At this time, animals in the control and severely diabetic groups were exsanguinated for determination of plasma renin and aldosterone concentrations. Micropuncture studies. Micropuncture studies were performed on animals 8-10 days after the injection of STZ or buffer. The animals were anesthetized with Inaction (Byk), 100 mg/kg ip, and were placed on an operating table with a servo-heating circuit designed to maintain body temperature at 37OC. Jugular vein and carotid artery were cannulated, and a tracheostomy was performed. Fluids were administered via the venous access, whereas blood pressure monitoring and blood sampling were done via the carotid artery. The left kidney was exposed via a flank incision, freed from peritoneal connections, placed in a cup and covered with warmed mineral oil, and the ureter was cannulated. A volume of Ringer solution (2 ml/100 g body wt) containing 100 &i [3H]inulin was administered, and continued at a rate of 2 ml/100 g body wt with 100 &!i [3H]inulin/h. After 45 4 3.5

.5 3 f

zE g


1 2.5 2 1.5

.-E 3



.5 0

C inulin FIG.

1. Validation


of [ 14C]inulin


- Minipump







min equilibration, the animals were included in the micropuncture section of the study if the mean arterial blood pressure was X00 mmHg. The animals were subjected to free-flow micropuncture of the left kidney cortex, while whole kidney data were collected simultaneously for clearance calculations. Urine was collected over 20-min intervals from the ureteric catheter, and arterial blood was collected at the beginning and end of each timed collection from which the whole kidney GFR was calculated. Whole animal GFR was regarded as twice the value of the whole kidney GFR and was reported per 100 g of animal body weight. Proximal tubules were identified morphologically and were confirmed as late proximal by oil droplet disappearance as the last superficial proximal tubular segment. Distal tubules were also identified morphologically and on the basis of oil droplet passage, but in addition the early distal site of puncture was confirmed by subsequent microdissection after injection of tubules with silicone rubber, using Microfil (Canton Biomedical). At the time of micropuncture an oil droplet was placed in and occluded the tubule just distal to the sampling site, and a timed collection of tubular fluid was made into oil-filled micropipettes (8-10 pm diam). Tubular fluid samples were transferred to a calibrated constantbore pipette (0.1 mm ID) for volume estimation, and a known volume was transferred for radioactivity measurement, the remainder being available for measurement of tubular fluid Na concentration. The collected volume of tubular fluid divided by the duration of the collection was a direct measure of tubular flow rate. TGF activity was determined as the difference between the late proximal and early distal estimates of SNGFR, as described and utilized previously by others (27-29). The difference between the mean values for late proximal and early distal measurements of SNGFR in individual animals was calculated as the “feedback activity index.” Results obtained in this way were validated in subsequent studies in which paired collections for SNGFR were made from proximal and distal tubules of the same nephron in a number of severely diabetic and control animals. As no difference was observed in paired vs. unpaired collections, the results are analyzed as a single group. Experiment

2: Phlorizin-Treated

Diabetes Metlitus

These experiments were designed to test the hypothesis that increased Na-glucose cotransport in the proximal tubule in diabetes mellitus is responsible for the observe’d alterations in single nephron function. Seven animals underwent subcutaneous osmotic minipump implantation for measurement of daily inulin clearance as per the above protocol, and after 3-4 days measurement of baseline GFR, received an intravenous injection of streptozotocin 60 mg/kg. Twenty-four hours after STZ administration all had BSLs in the severely diabetic range, and daily subcutaneous injection of phlorizin (400 mg/kg dissolved in 0.4 ml 95% ethanol) was commenced as previously described (30). BSL and inulin clearance were measured daily, and 8-10 days after the administration of STZ the animals were subjected to free-flow micropuncture of the left kidney cortex in an identical experimental protocol to that described above.

Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} ( on July 30, 2018. Copyright © 1991 American Physiological Society. All rights reserved.




Experiment 3: Mannitol- Treated Controls

These experiments assessed the role of an osmotic diuresis and increased tubular flow rate, independent of the diabetic state, in the pathogenesis of the hyperfiltration of diabetes. Six animals without prior intervention were subjected to free-flow micropuncture of the left kidney cortex in an identical experimental protocol to that described above. However, the Ringer solution used for infusion contained 1.5% mannitol, the concentration chosen to achieve a similar tubular flow rate to that of the severely diabetic animals. Analyses

Plasma and urinary Na were measured by flame photometer (model 643; Instrumentation Laboratory); blood glucose was measured by Reflectron glucometer, and results were confirmed by simultaneous assay in 350% of the cases by spectrophotometric measurement of plasma glucose [Enzymatic (Hexokinase) Determination; Sigma Diagnostic Reagents]; tubular Na concentration was measured by an ultramicro adaptation of the flameless atomic absorption spectrophotometer (Video 22 Spectrometer with model 755 Furnace Atomizer; Instrumentation Laboratory). Standards, prepared containing Na concentrations of 20, 40, 80, and 160 mmol/ 1, were assayed both without and with glucose present in concentrations of 0, 5, 10, and 20 mmol/l, and mannitol at 0.5,1, and 2% concentrations. The addition of glucose or mannitol had no significant effect on the standard curve; thus tubular fluid concentrations were measured against nonglucose or mannitol-containing standards. [3H]inulin radioactivity was measured in plasma, tubular fluid, and urine by liquid scintillation counting. Plasma renin activity was assayed at pH 7.4 by radioimmunoassay ( 1261 Plasma Renin Activity Radioimmunoassay Kit; Baxter Clinical Assays GammaCoat), and plasma aldosterone was also measured by radioimmunoassay ( 12’1 Direct Aldosterone Radioimmunoassay; Diagnostic Products Coat-A-Count). calculations

Whole animal and single- kidney glomerular filtration rate were considered equal to the appropriately determined inulin clearance. In micropuncture experiments, SNGFR was calculated as the tubular fluid-to-plasma (TF/P) inulin concentration times u, the rate of flow of tubular fluid (nl/min) at the site punctured. In a minority of instances SNGFR was assessed by measuring the total inulin content in the tubular fluid during a timed collection divided by the plasma inulin concentration. Fractional delivery (percent of filtered load) of Na to the tubular site at which the collection was made was determined as the TF/P Na concentration ratio divided by the TF/P inulin ratio at that site times 100%. The total tubular Na reabsorption up to the early distal site was calculated as the difference between the amount of Na filtered at the glomerulus and the amount of Na remaining in the tubule at the early distal site, i.e., (SNGFR x PNa) - (u x TFN,). Analysis of variance was used to detect differences


between three animal groups, and Scheffe comparisons were used to determine which pairs of groups differed when the F test was significant (P C 0.05). Unpaired t tests were used to detect differences in results when two groups were compared. RESULTS


1: Diabetes Mel&s

Figure 2 illustrates the whole animal GFR, expressed as the mean t SE of daily inulin clearance, both preand postinjection of STZ or buffer. As shown, all animals had a similar baseline GFR. However, the animals with severe diabetes had a significant and sustained rise in GFR within 48-72 h of the administration of STZ (0.96 t 0.08 to 1.22 & 0.05 ml l rein-’ 100 g-l; P < 0.02, mean values pre-STZ vs. mean values post-STZ). In contrast, the animals with moderate diabetes had a stable GFR pre- and postinjection of STZ [l.Ol t 0.01 to 0.97 t 0.02 ml min-’ 100 g-l; P = not significant (NS)], which was similar to that of control animals (0.95 t 0.04 to 0.91 & 0.02 ml min-’ 100 g-l; P = NS). A small positive daily Na balance was observed in all animals, which increased in the severely diabetic animals such that they were in a significantly more positive Na balance 4-5 days after the injection of STZ (0.40 t 0.04 in control vs. 0.61 t 0.12 in moderate diabetes vs. 1.10 & 0.22 mmol Na/day in severely diabetic animals; P c 0.05 in severely diabetic vs. control animals). Although this coincided with the increase in GFR in the severely hyperglycemic group, no correlation was observed between Na balance and GFR for any given day. Plasma renin and aldosterone were similar in the control and severely diabetic animals, being 4.37 t 0.72 vs. 4.37 t 0.59 rig/ml (P = 0.99) and 16.25 t 3.06 vs. 13.00 t 10.38 ng/dl (P = 0.77), respectively. Daily hematocrit was similar in all animals under study, being 0.48 t 0.01 in all cases. Micropuncture data is as detailed in Table 1. As expected, animals with severe diabetes had higher tubular fluid flow rates compared with moderate diabetes and control, a pattern that was further reflected in SNGFR, measured at both late proximal and early distal segments (Fig. 3). Late proximal TF/P inulin ratios were higher in animals with severe diabetes compared with moderate diabetes and control (2.5 t 0.01 vs. 2.1 t 0.01 vs. 1.9 t l





P< 0.05

FIG. 2. Inulin clearance after administration of streptozotocin (STZ) or citrate buffer. IJ, Citrate control; 0, moderate diabetes; l , severe diabetes.

Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} ( on July 30, 2018. Copyright © 1991 American Physiological Society. All rights reserved.





1. Whole animal and micropuncture data

















GFR, ml mid 100 g-’ 0.963-0.06 (6) 0.90t0.06 (9) 1.19to.o5*t BSL, mmol/l 14.5tO.B-t (6) 8.2t0.2 (9) 22.2t0.1*“r Late proximal tubular flow rate, nl/min 16.8-cl.9 (10) 16.3t1.2 (21) 29.2&1.7*-b Early distal tubular flow rate, nl/min 5.9&O. 1 (10) 10.4&0.9*-F 6.220.4 (16) Lat,e proximal SNGFR, nl/min 34.4t2.2 (22) 54.Ok3.9t (16) 83.2t5.0*t Early distal SNGFR, nl/min 21.5t1.2 (25) 27.6t2.2 (16) 46.lt4.7”t Feedback activity index, nl/min 14.8t1.9 (9) 30.1-1-6.5 (6) 46.4&6.6-F Late proximal [Na], mmol/l 137.3t3.3 (17) 133.4k5.7 (14) 114.8+-4.3* t Early distal [Na], mmol/l 55.4-I-5.6t (11) 75.9t4.0 (21) 29.7+2.1*t Late proximal fractional, Na delivery, % 52.9t1.9 (18) 48.5H.6 (12) 42.9t3.0t Early distal fractional, Na delivery, % 12.3t0.9 (22) 8.8&l.lt (11) 4.5zkO.4*t Tubular Na reabsorption to early distal 2.7AO.2 (21) 3.3t0.5 (8) 5.5t0.6* t site, nmol/min Values are means t SE; no. in parentheses refers to n for each result. * Significantly different from citrate control; $ significantly different from severe diabetes. GFR, glomerular filtration nephron GFR. l


n q 100

Late Early

Proximal Distal







Severe Diabetes

FIG. 3. Late proximal and early distal measurements of singlenephron glomerular filtration rate (SNGFR). * P < 0.05 compared with control.

0.01; P < 0.05), whereas at the early distal site TF/P inulin ratios were similar in all three animal groups (5.2 t 0.4 vs. 5.0 t 0.5 vs. 5.1 t 0.4; P = NS). The increased proximal-distal difference in SNGFR (i.e., the feedback activity index) indicated that TGF was significantly more active in the severely diabetic group (46.4 t 6.6 nl/ min) compared with the control animals (14.8 t 1.9 nl/ min; P < 0.05). In the animals with moderate diabetes, the feedback activity index was elevated to a lesser extent (30.1 -I- 6.5 nl/min), and this resulted in the return of the distally measured SNGFR towards normal. This enhanced TGF activity resulted in a percentage reduction from the proximally measured SNGFR of 54.3 t 3.8% in severe diabetes vs. 40.7 & 2.9% in control animals (P < 0.05), with moderate diabetes having a value similar to control of 42.5 t 5.5%. Late proximal tubular Na concentration was significantly lower in the severely diabetic group, due at least in part to the osmotic effect of tubular glucose. At the early distal tubular site, tubular Na concentrations differed significantly among all three groups, with animals with severe diabetes having the lowest Na concentration and those with moderate diabetes having an intermediate value when compared with the control group. The fractional Na delivery at the late proximal tubule in severe diabetes was significantly lower than that in moderately


(11) (11) (11) (31)

+ Phlorizin

221tlO 0.99kU.O3$ 13.8+1.2$ 34.422.3


(6) (6) (6) (17)

+ Mannitol

218t16 0,91+0.06$ 7.9kO.25 32.2-r-2.1 10.8tl.l 59.7+4.3$ 36.1t2.7 20.023.3

(6) (6) (6) (14) (14) (14) (14) (6)




(35) (25) (11)

60.9+2.8$ 30.6&2.7$ 34.9t5.5

(17) (15) (6)







66.0+4.9$ 55.8+3.0$ 9.9*0.6$ 3.7&0.5$

(9) (14) (11) (11)

52.8+4.5$ 54.8*3.3$ 12.4+1.4$ 4.3+0.4$

(15) (15) (14) (11)

(17) (18) (19)

from moderate diabetes; t significant.ly different rate; BSL, blood sugar level; SNGFR, single-

diabetic or control animals, indicating that enhanced Na reabsorption occurred along the proximal tubule. The fractional delivery of Na was further reduced at the early distal site in both the severely diabetic and the moderately diabetic groups when compared with control (Fig. 4), suggesting that Na transport was also increased between the late proximal and early distal tubule, i.e., in the loop of Henle. The calculated total tubular Na reabsorption up to the early distal site was higher in severe diabetes when compared with both moderate diabetes and control values (see Table 1). This enhanced tubular reabsorption of Na was further reflected in a lower fractional excretion of Na in the final urine in severely diabetic animals, as we have previously demonstrated in analogous experiments (22). Experiment 2: Phlorizin-Treated

Diabetes Mellitus

Whole animal GFR assessed by daily inulin clearance administered by osmotic minipump delivery was 1.14 t 0.05 pre- and 1.16 t 0.04 postinduction of diabetes mellitus and phlorizin administration (P = NS). Subsequent micropuncture data is given in Table 1. As the aim of the experiment was to assess the role of proximal Naglucose cotransport in the pathogenesis of the hyperfiltration of diabetes, results are compared with those of the severely diabetic animals. Hematocrit (0.52 t 0.01) and plasma Na (143.0 k 4.7 mmol/l) in phlorizin-treated animals were similar to that observed in severe diabetes. * P

Tubular sodium handling and tubuloglomerular feedback in experimental diabetes mellitus.

Tubular Na handling and tubuloglomerular feedback (TGF) activity were assessed using micropuncture techniques during the hyperfiltration phase of stre...
2MB Sizes 0 Downloads 0 Views