S. J. Walter, B. Sampson and D. G. Shirley micropuncture samples were obtainedfrom sites beyond the early distal tubule, information on eventsin the collecting ...
3213
Journal of Physiology (1995), 483.2, pp. 473-479
473
A micropuncture study of renal tubular lithium reabsorption in sodium-depleted rats S. J. Walter, B. Sampson * and D. G. Shirley Departments of Physiology and *Chemical Pathology, Charing Cross and Westminster Medical School, London W6 8RF, UK The marked reduction in fractional lithium excretion (FELl) which accompanies chronic sodium depletion was investigated using free-flow micropuncture in anaesthetized rats which had been maintained in a sodium-depleted state for 8-10 days. 2. Compared with previous values in sodium-replete rats, sodium depletion was associated with small reductions in total and superficial nephron glomerular filtration rate and enhanced fractional reabsorption of water, sodium and lithium in the proximal convoluted tubule. 3. In untreated (sodium-depleted) rats, fractional deliveries of lithium (FDLI) to the late proximal convoluted tubule, early distal tubule and late distal tubule were 0-41 + 0-02, 0-20 + 0-01 and 0-18 + 0-02 (means + S.E.M.), respectively. Fractional lithium excretion (0-08 + 0-01) was significantly lower than late distal FDLi (P < 0-001). 4. Treatment with amiloride did not affect segmental lithium handling up to the late distal tubule. Frusemide had no effect on lithium reabsorption in the proximal convoluted tubule, but early distal FDLi (0 30 + 0-01) was raised compared with the untreated group (P < 0-001). Both diuretics eliminated the difference between late distal FDLi and FELi, respective values being 0-17 + 0-02 and 015 + 0-01 (amiloride-treated rats) and 0'31 + 0-02 and 0 34 + 0-02 (frusemide-treated rats). 5. These data indicate that part of the reduction in FELi in chronic sodium depletion is due to enhanced fractional fluid (and lithium) reabsorption in the proximal convoluted tubule. In addition, however, they provide direct evidence for amiloride-sensitive lithium reabsorption in the collecting ducts. Although frusemide may also inhibit lithium reabsorption at this site, a disproportionate increase in lithium delivery from deep nephrons could partly explain the findings with this diuretic. 1.
The renal clearance of lithium (Cuj) is widely used as a means of assessing proximal tubular function. Its use is based on the assumption that no lithium is reabsorbed beyond the proximal tubule (Thomsen, 1990). Although recent micropuncture studies in rats infused with the loop diuretic frusemide have suggested that a small amount of lithium may additionally be reabsorbed in the loop of Henle (Shirley, Walter & Sampson, 1992; Fransen, Boer, Boer, Dorhout Mees & Koomans, 1993), on balance it seems that C%j may normally provide a reasonable index of end-proximal fluid delivery (Shirley & Walter, 1993). However, when rats are fed a low sodium diet, values for CLi and fractional lithium excretion (FELl) fall to extremely low levels (Thomsen & Leyssac, 1986; Kirchner, 1987; Shirley, Skinner & Walter, 1989; Shalmi, Petersen, Thomsen & Christensen, 1990) and it is widely believed that in this situation a significant amount of filtered lithium is reabsorbed beyond the loop of Henle. This view is supported by the observation that amiloride,
a diuretic whose principal site of action is the collecting tubule (Horisberger & Giebisch, 1987) and which in sodium-replete animals does not affect lithium reabsorption (Shirley et al. 1992), increases Ciin sodium-depleted rats
(Thomsen & Leyssac, 1986; Kirchner, 1987). To date, however, no direct micropuncture evidence of amiloride-sensitive lithium reabsorption in the collecting ducts of sodium-depleted animals has been provided. On theoretical grounds, it is possible that lithium reabsorption in other nephron segments may be enhanced during sodium depletion, and it is possible that, in this setting, amiloride might increase lithium excretion by acting on segments other than the collecting duct. A recent study has attempted to address these points (Fransen, Boer, Boer & Koomans, 1992), but the degree of sodium depletion at the time of micropuncture was mild and consequently FELi was not very different from that in control rats (21 vs. 23 %). Moreover, since no
J. Physiol. 483.2
S. J. Walter, B. Sampson and D. G. Shirley
474
micropuncture samples were obtained from sites beyond the early distal tubule, information on events in the collecting ducts wras limited.
sodium and water excretion wvere balanced by increases in infusion rates; thus sodium and wrater balance in diuretic-treated animals remained similar to that of untreated animals.
In the present study, performed in sodium-depleted rats exhibiting markedly reduced CLi values, we have used micropuncture, combined with measurements of tubular lithium concentrations (Sampson, 1991), to document lithium reabsorption in each accessible nephron segment. We lhave also investigated the nephron site(s) responsible for the lithiuretic effect of amiloride in this situation. Finally, in order to shed light on the possible role of the loop of Henle, we have also performed measurements during infusion of frusemnide.
Frusemide-treated rats (n = 8). The protocol adopted for rats in this group was identical to that for amiloride-treated animals except that: (a) amiloride was replaced by frusemide, infused at a rate of 2-5 mg kg-' h-'; (b) the rate of lithium infusion was increased to 20 #umol h-'; and (c) KCl was included in the second infusate at varying concentrations so that net potassium losses during the period of micropuncture were similar to those measured in untreated animals. In all groups, these infusions were continued for a period of 5 5 h. M\ean arterial blood pressure and renal clearances of [3H]inulin and lithium were measured during the final 4-5 h, when urinary excretion rates were relatively stable. Simultaneously, free-flow micropuncture
METHODS Male Sprague-Dawley rats (initial weight range 260-300 g) received a diet deficient in sodium (4 mmol (kg dry wt)F; Bioteknisk Institut, Kolding, Denmark) ad libitum for 8-10 days before anaesthesia and micropuncture. On the first day of the sodium-deficient diet, each rat was given a single dose of frusemide (40 mg (kg body wt)F'; Hoechst, Hounslow, AMiddlesex, UK) by gav-age so as to establish a marked degree of sodium depletion. Pilot experiments were performed to determine: (a) the minimum sodium infusion consistent with the maintenance of renal function during and after surgery in sodium-depleted rats; and (b) the rates of fluid and sodium excretion during the first 15 min of i.v. infusion of either amiloride (2 mg kg-' h-'; Meerck, Sharp & Dohme, Hoddesdon, Herts, UK) or frusemide (2 5 mg kg-' h-'; Hoechst). The results were used to determine the rates of infusion employed during the subsequent micropuncture experiments described below. On the day of micropuncture, each rat was anaesthetized with Inactin (120 mg kg-'; Byk Gulden, Konstanz, Germany) and prepared surgically as described in previous reports from this laboratory (Wralter, Laycock & Shirley, 1979; Shirley, Zewde & Walter, 1990). In the present study, the initial i.v. infusions consisted of 1 3 ml hV' of 2% glucose solution and 0 5 ml hV' of 0 9% NaCl solution containing LiCl (20 mmol F-'). One hour after the completion of surgery, the rats were given a priming dose of [3H]inulin (50 ,uCi; Amersham International, Aylesbury, Bucks, UK) and then divided into three groups as follows: Untreated rats (n = 9). Two intravenous infusates were used: 2% glucose solution infused at 1-3 ml hV1 and a solution of [3H]inulin and LiCl infused at 0 5 ml h-'. Sodium was not infused in this group. Rates of [3H]inulin and lithium infusions were 50 tCi h-' and 10 yumol h-', respectively. The overall osmolality of the combined infusates was sufficient to prevent haemolysis without the development of glycosuria.
Amiloride-treated rats (n = 9). Rats were infused at 1 3 ml h-' with 0 45% NaCl solution containing [3H]inulin, LiCl and amiloride. Rates of infusion were 50 1tCi h-', 13 /tmol hV' and 2 mg kg-' I', respectively. Rats vere also infused Nwith a solution of NaCl with the iniitial concentration and infusion rate determinined by data obtained in pilot experiments referred to above. U'rine w-as collected and analysed at regular intervals and the sodium concentration and infusion rate of the second i.v. infusate adjusted so that the diuretic-induced increases in
collections of tubular fluid
were
made from three
regions of superficial nephrons: the final loops of the proximal convoluted tubules, and early and late segments of accessible distal tubules (Walter & Shirley, 1986). Two collections wNere made from each region. Analysis of tubular fluid was as described previously (Shirley et al. 1992). Briefly, micropuncture collections Aere deposited under oil, their volumes measured, and duplicate samples taken for determination of [3H]inulin (f4-spectroscopy; model 2000 CA, Canberra Packard, Pangbourne, UK), sodium (helium glow photometry; Aminco, Silver Springs, N.Y., USA) and lithium (furnace atomic absorption spectrometry; model 3030, Perkin Elmer, Beaconsfield, UK). The lithium concentrations of plasma and appropriately diluted urine were measured using an identical technique. Sodium and potassium concentrations in urine and plasma were determined by flame photometry (MIodel 543, Instrumentation Laboratory, Warrington, UK). Calculations The clearances of [3H]inulin (used as a measure of glomerular filtration rate (GFR)) and lithium were calculated by the usual expression:
cx
=
(Ux/Px)VI
where Ux and Px are the urine and plasma concentrations, respectively, of the substance x and V is the urine flow rate. FELi was calculated as CLi/GFR. Appropriate plasma values for clearance measurements and for tubular fluid (TF)/plasma (P) concentration ratios of [3H]inulin (TFIn/P1n) and lithium (TFLi/PLi) were interpolated from the measured values. Single nephron glomerular filtration rate (SNGFR) was calculated, using distal tubular collections only, as (TFIn/Pmn)(VTF), where VTF is the tubular fluid flow rate. Fractional deliveries of sodium and lithium to each puncture site were calculated as (TFNa/PNa)/ (TFIn/P1.) and (TFLi/PLI)/(TFIn/Pmn), respectively. Results were expressed as a single (mean) value per rat. Statistics Values are expressed as means + S.E.M. Comparisons between the three groups of animals were made using one-way analysis of variance. Wlhen marked heterogeneity of variance was present (as assessed by Hartley's Fmax test), the data wvere log transformed before analysis. For micropuncture data fmom different nephron sites, two-way analysis of variance was used. Where a significant inter action between nephron site and treatment group was found, the simple effects of site and ti-eatment Nvere each analysed by one-way analysis of variance.
Student's
unpaired
t test
was
used foi-
coi-nparisons
between
J. Physiol. 483.2
Renal lithium transport in sodium depletion4
475
Table 1. Body weight, clearance values and urinary excretion rates
oi
Body wt
(g) Untreated Amiloride treated Frusemide treated
9 9 8
280+6 288 + 7 282 + 7
Urine flow rate GFR FELi CLi (ml min-') (ml min-) (,u min-') 1P28+0-07 010+001 0.08+0.01 4-8+ 1 0 1-28 + 005 0-19 + 0-02** 0-15 + 001 ** 156 + 28** 1-15 + 005 039 + 002t 034 + 002t 797 + 14'6t
Na excretion
K excretion
rate
rate
(,umol min-')
(,umol min-l)
0 50 + 0-06 0.05+0-01 2-94 + 046** 0 37 + 0.04* 8-61 + 220t 2-85 + 0-27t
Values (means + S.E.M.) are for the left kidney only; ni, number of rats. GFR, glomerular filtration rate; CLi, lithium clearance; FE Li fractional lithium excretion. * P < 0 01, ** P < 0 001 compared with untreated group. t Significantly different from untreated and amiloride groups (P < 0 001).
results in sodium-depleted rats and those obtained in sodiumreplete animals in a previous study. A value of P < 0 05 was considered to be statistically significant.
RESULTS The combination of frusemide administration by gavage and the simultaneous introduction of a diet with an extremely low sodium content was accompanied by a loss of body weight of 8 + 1 g on the first day of treatment, although the rats typically had recovered all of this weight by the day of micropuncture. Evidence of the effectiveness of the sodium-restriction regimen was obtained immediately following anaesthesia: the sodium concentration of urine withdrawn from the bladder ranged from 0 to 1 1 mmol F'.
Overall renal function during the period of micropuncture is shown in Table 1. As expected, sodium restriction resulted (in the untreated group) in a very low rate of sodium excretion and markedly reduced values for Q,j and FELi compared with sodium-replete animals (Shirley et al. 1992). Glomerular filtration rate was slightly lower in sodium-depleted rats (0-46 + 0-02 vs. 0-52 + 0-02 ml min- (100 g body wt)-'; P
