INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

REGULATION OF ACID-BASE balance is a major task of the kidneys. This is because the renal tubule reabsorbs almost all the filtered bicarbonate in exchange for secreted H⁺ and participates in the excretion of nonbicarbonate buffers, such as phosphate and ammonium ion. Several groups have shown that the proximal tubule is the major site of bicarbonate reabsorption (1, 8, 26); however, more distal nephron segments may also participate in the renal regulation of acid-base balance. Among the distal nephron segments, the loop of Henle (LOH) has a very important role, since, under normal conditions, it reabsorbs ~10-15% of the filtered bicarbonate. In vivo perfusion studies of the LOH have demonstrated that most of the bicarbonate reabsorption is mediated by luminal Na⁺/H⁺ exchange, but these studies also have provided evidence for a significant contribution of electrogenic H⁺ secretion to the process of acidification (5). Moreover, we have also proven that the LOH participates in the renal adaptation to acid-base disturbances (6) and that it is the site of action of several hormones (6, 27).

The LOH is a very complex segment and is characterized by at least two peculiar properties: the extreme heterogeneity and the particular anatomical configuration. The LOH is composed of several subsegments that have distinct histological and physiological features. With respect to acid-base transport, in vitro studies have demonstrated that the thick ascending limb (TAL) reabsorbs bicarbonate and may be responsible for the bulk of total LOH bicarbonate transport (11). The LOH is surrounded by tissue with increasing longitudinal concentration gradient (34), resulting from the marked addition of sodium, chloride, and urea contents (15) accomplished through the countercurrent system. As shown by several investigators, medullary cells utilize cellular osmolytes to survive in this hypertonic medium (9); on the other hand, perturbations of extracellular osmolality will affect intracellular pH in a variety of cells, including the TAL cells (16, 31). The present studies have been performed to investigate whether interferences of extracellular osmolality on H⁺ and transport systems do affect transepithelial transport along the LOH. The data collected demonstrate that, following maneuvers that will decrease medullary tonicity, net bicarbonate reabsorption is significantly stimulated along the LOH. Moreover, they reveal that, under identical experimental conditions, chloride transport is also activated.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Preparation of animals. Experiments were done on a total of 40 male (250-350 g body wt) Sprague-Dawley rats grouped in cages at 21°C. The animals received food and drinking water up to the time of study. They were anesthetized intraperitoneally with Inactin (Promonta) using a dose of 120 mg/kg body wt, tracheotomized, placed in the right lateral position on a thermoregulated table (37°C), and prepared for micropuncture as previously described (7). In brief, the right carotid artery was catheterized to record blood pressure and take blood for measurements of hematocrit, arterial pH, and inulin and electrolyte concentrations. The left jugular vein was cannulated with polyethylene PE-50 tubing and used for intravenous infusion via a syringe pump (Braun Apparatus) of different saline solutions according to the protocols detailed below. Urine was drained with PE-10 tubing from the ureter of the left kidney, which had been exposed through a flank incision, freed of perirenal fat, and immobilized in a Lucite chamber with 3% agar in 0.9% saline. Throughout the experiment the kidney was bathed with prewarmed (37°C) paraffin oil.

Experimental protocols. To reduce renal medulla osmolality, the following two approaches were used: 1) intravenous infusion of large dose of hypertonic mannitol solution (mannitol protocol); 2) intravenous infusion of a hypotonic solution at high flow rate (hypotonic protocol). In the first set of experiments (mannitol protocol) the anesthetized animals were infused intravenously during a control period (~120 min) with a modified Ringer-saline solution (125 mM NaCl + 25 mM NaHCO₃) at a rate of 24 µl · min¹ · 100 g¹. Thereafter, an osmotic diuresis was induced by intravenous injection of 18% mannitol solution for 5 min at a rate of 72 µl · min¹ · 100 g body wt¹, followed by a continuous infusion of 9% mannitol solution with 5% BSA at a rate of 18 µl · min¹ · 100 g body wt¹ (Fig. 1). It has been demonstrated that this maneuver induces a significant washout of the renal medulla osmolality (10), but it is also associated with plasma volume expansion, as suggested by the large reduction of blood hematocrit (see below). To minimize volume expansion, a second set of experiments was performed (hypotonic protocol) according to Fig. 2. During the control period, a modified Ringer solution (125 mM NaCl + 25 mM NaHCO₃) was infused intravenously at 10 µl · min¹ · 100 g body wt¹; then a hypotonic solution containing 35% of the control sodium chloride and bicarbonate concentrations plus a small dose of mannitol (36 mM NaCl + 7 mM NaHCO₃+1% mannitol) was infused intravenously at a flow rate 3.5 times higher (35 µl · min¹ · 100 g body wt¹). Following this scheme, the intravenous load of sodium, chloride, and bicarbonate during the two periods remained unaltered, and plasma volumes seem to be less affected, at least based on the hematocrit values (see below). Collection of micropuncture samples were begun 90 min after starting the infusion of hypotonic solution.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Scheme of mannitol protocol.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Scheme of hypotonic solution protocol.

To rule out nonspecific time-dependent changes in loop transport rates over the 2- to 6-h experimental period, we performed a third set of experiments. Following the surgical procedure, the anesthetized animals were infused up to 6 h with a modified Ringer-saline solution as during the control period of protocol 1. Micropuncture samples were collected 120-180 and 240-360 min after the beginning of the infusion. The data obtained from each time period were pooled and compared vs. the data collected during the control period of protocol 1.

Tubule microperfusion. We performed continuous microperfusion of superficial LOHs in vivo to measure bicarbonate, chloride, and fluid transport under conditions of fixed flow rate and bicarbonate delivery. To carry out the micropuncture experiments, a perfusion pipette was inserted into the last surface loop of proximal tubule, and a castor oil block was placed upstream of the perfusion pipette. Microperfusion was started at 20 nl/min with a thermally shielded microperfusion pump (Hampel, Frankfurt, Germany). The perfusion solution was composed of (in mM) 128 NaCl, 16 NaHCO₃, 3.8 KCl, 1 MgCl₂, 0.38 NaH₂PO₄, and 1.62 Na₂HPO₄. Exhaustively dialyzed [methoxy-³H]inulin (50 µCi.ml^-1) was added to the perfusion solution to measure fluid transport. FD & C (0.07%) blue dye was used to color the perfusate. Mannitol was not present in the perfusate. A collection pipette was inserted at the early distal segment, which had been identified by injecting a small droplet of oil in the proximal tubule. A castor oil block, distal to the collection site, prevented contamination by fluid from more distal nephron segments.

Transport data collected during the experimental periods were compared vs. those obtained during the respective control periods, and each animal served as its own control. Data are given per individual loop, as it has been shown that the cortical LOH is a nephron segment of essentially constant length (~6-7 mm) (29).

Whole kidney clearance. In similar groups of animals, glomerular filtration rate (GFR) and urinary bicarbonate excretion were measured. After a priming dose of 20 µCi [methoxy-³H]inulin in 0.5 ml 0.9% saline, the maintenance solution of modified Ringer-bicarbonate delivered [methoxy-³H]inulin at 20 µCi/h iv. After 45-min equilibration, the first of three to five 30- to 60-min urine collections began. Care was taken to measure GFR in the same time periods as for the micropuncture protocol. Carotid artery blood samples were taken at the start and at the end of each clearance period. In the mannitol protocol, renal plasma flow (RPF) was also measured by using the clearance of ¹⁴C-labeled p-aminohippuric acid (PAH) (5 µCi as a bolus, followed by 5 µCi/h as a maintenance infusion).

Analytical methods. Tubule fluid total CO₂ concentration was measured by microcalorimetry (Picapnotherm; WPI, New Haven, CT). To avoid loss of CO₂, all mineral oil used (to bathe kidney surface, to collect tubule fluid samples, and to cover samples for measurements) was equilibrated to cortical carbon dioxide tension (PCO₂) values with a solution containing 100 mM HEPES buffer and 48 mM NaHCO₃, equilibrated with 6.7% CO₂. Each sample analysis was bracketed by running standards of known NaHCO₃ concentration. Tubule fluid Cl concentration was measured by flow-through Cl electrodes. The blood acid-base status of each animal was measured with a blood-gas analyzer (model ABL 300; Radiometer, Copenhagen, Denmark). [Methoxy-³H]inulin and [¹⁴C]PAH radioactivities were measured by a liquid scintillation counter (Perkin-Elmer). Urine concentrations were measured with a carbon dioxide analyzer (model 965, Corning).

Calculations and statistical analysis. GFR, RPF, and fractional electrolyte excretion were calculated, using standard clearance formulas. In the microperfusion experiments, the perfusion rate achieved in vivo was obtained from the rate of fluid collected from the early distal tubule, multiplied by the ratio of inulin concentrations in collected vs. perfused fluids. The perfusion pump was calibrated by timed collections of perfusion fluid delivered directly into counting vials for measurements of [methoxy-³H]inulin concentrations. Net fluid absorption (J_v, nl/min) was the difference between perfusion and collection rates. The volume of the collected fluid was measured with a calibrated capillary. Net bicarbonate (J_HCO3, pmol/min) and chloride (J_Cl, pmol/min) reabsorption were calculated from the amount of ions delivered in the perfusion pipette minus the amount collected in the collection pipette according to the following formula where C_p and C_c are the bicarbonate or chloride concentrations of the perfused and collected fluids, respectively, and V_p and V_c are the perfusion and collection rates, respectively. A sample collection was considered acceptable if V_p value was within 15% of the calibrated microperfusion pump rate.

Statistical analysis was performed by one-way analysis of variance followed by comparison with appropriate control (post hoc) using the least significance difference test. All data are expressed as means ± SE.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Plasma composition, whole kidney function, and loop transport data after infusion of mannitol solution (mannitol protocol). Table 1 summarizes data on packed cell volume, blood pressure, renal function, arterial blood composition, urinary bicarbonate excretion, and urine osmolality of the animals studied before and after mannitol infusion. As expected, the use of large doses of mannitol resulted in significant changes in hematocrit values, a finding pointing to large expansion of the extracellular fluid volume. GFR and RPF were also increased, in agreement with previous reports (3). Although no significant changes in blood pH and bicarbonate concentrations were observed, the infusion of mannitol was associated with a significant increase in urine pH and of fractional bicarbonate excretion. Finally, the infusion of mannitol resulted in a washout of the renal medulla osmotic gradient as demonstrated by the large fall of urine osmolality and by the massive increase in urine flow rate.

Table 2 (top) and Fig. 3 (left) show the micropuncture data during the control period and after the intravenous mannitol infusion. Under conditions of similar perfusion rate and bicarbonate load, there was a significant increase in both the absolute and fractional LOH bicarbonate transport following the washout of the renal medulla gradient by mannitol. It must be emphasized that, in the microperfusion experiments, the perfusate did not contain mannitol.

The time course experiments, performed on four rats, demonstrated that there were no significant changes of LOH bicarbonate reabsorption along the 6-h period; J_HCO3 was 198.61 ± 18.63 pmol/min (n = 7) during the 120-180 min and 189.52 ± 11.33 pmol/min (n = 14) during the last 240-360 min. These values were not significantly different from each other nor from J_HCO3 measured during the control period of protocol 1 (192.94 ± 17.54 pmol/min) (P > 0.05).

Plasma composition, whole kidney function, and loop transport data after infusion of hypotonic solution (hypotonic protocol). Table 3 summarizes data on packed cell volume, blood pressure, GFR, systemic acid-base balance, urine pH, fractional excretion of bicarbonate, urine osmolality, and flow rate of the rats studied before and after the infusion of hypotonic solution (hypotonic protocol). Although this maneuver reduced significantly the urine osmolality and increased fivefold the urine flow rate, it was associated with minimal changes of the plasma volume, at least as estimated from the hematocrit values. An additional and interesting finding was the reduction of GFR (38%) following the infusion of the hypotonic solution (0.98 ± 0.08 vs. 0.61 ± 0.05 ml · min¹ · 100 g body wt¹, respectively; P < 0.01). This reduction was even more apparent when the hemodynamic data were plotted vs. the corresponding urinary osmolality (Fig. 4). A linear and highly significant relationship was found (r = 0.789, P < 0.001). On the other hand, no significant difference was observed on systemic acid-base balance, whereas the urine fractional excretion of bicarbonate almost doubled (from 0.39 ± 0.18% to 0.77 ± 0.28%, P < 0.05) and was associated with a significant urine alkalinization.

View larger version (31K): [in this window] [in a new window]   Fig. 3.   Net bicarbonate reabsorption (JHCO3) along the loop of Henle during mannitol and hypotonic protocols. * P < 0.05 and *** P < 0.001 vs. control (values are means ± SE).

Table 2 (bottom) shows the micropuncture data obtained under these experimental conditions. In the presence of comparable perfusion rate and LOH bicarbonate loads and without mannitol in the perfusate, J_HCO3 was significantly stimulated (Fig. 3, right) during the intravenous infusion of the hypotonic solution (P < 0.001). The same holds for the fractional bicarbonate reabsorption.

To further investigate the effect of medullary osmolality on LOH transepithelial ionic transport, net reabsorption of chloride (J_Cl) was measured according to the experimental conditions detailed in the hypotonic protocol. The corresponding micropuncture data of this set of experiments are reported in Table 4. In the presence of almost identical LOH chloride load, J_Cl was significantly stimulated during the intravenous infusion of the hypotonic solution from 1,538 ± 98 to 1,906 ± 71 pmol/min (P < 0.001). Therefore the intravenous infusion of hypotonic solution increased both bicarbonate and chloride transepithelial net reabsorption along the LOH.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The renal medulla is an important portion of renal parenchyma where several kidney functions are achieved, including the concentration of urine. One of the main features of the medulla is the interstitial accumulation of several ions and urea leading to the formation of the corticomedullary osmotic gradient (15, 34). Acute and/or chronic dissipation of this gradient may alter the properties of the tubular cells, causing alterations of the transepithelial transport processes, including H⁺ and handling. To test this hypothesis, we have measured the net bicarbonate reabsorption along LOHs in conditions of reduced medullary gradient induced by two different maneuvers.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Relationship between glomerular filtration rate (GFR) and urine osmolality in the hypotonic protocol.

Mannitol experiments. One of the most potent tools to reduce medullary osmotic gradient is the intravenous infusion of mannitol solution. Such procedure will induce 1) an increase in RPF, including the medullary and papillary plasma flow; and 2) a reduced tubular reabsorption of several electrolytes (Na⁺, Cl, Mg²⁺, and others), particularly along the ascending limb, related to the reduction in water reabsorption from the thin descending limb of the LOH (20, 23, 25, 35). These two effects will decrease the medullary solute gradient (also in the outer medulla) by preventing the accumulation of both urea and sodium (10), thereby inhibiting urinary concentration. In the present study, most of these findings have been fully confirmed: we have found an increase in RPF and urine flow rate, associated with a decrease in urine osmolality; the GFR was also stimulated, although this effect is not constantly reported (3). With respect to the systemic acid-base balance, we did not find any significant change. This last observation is at variance with some sporadic reports that indicates a reduced blood pH following mannitol infusion. This effect has been attributed to the dilution of bicarbonate, related to the general hemodilution promoted by osmotic abstraction of cell water (33); presumably, the short duration of mannitol infusion in the present experiments accounts for the lack of expansion acidosis. An interesting finding is the significant increase in the urinary bicarbonate excretion paralleled by a corresponding increase in urine pH. Although specific and well-designed micropuncture experiments are lacking, it is possible to explain the bicarbonaturia by the same mechanisms mentioned for the decreased tubular solute transport; the lack of water extraction on the thin descending part of the loop will impede bicarbonate accumulation in the lumen of the LOH, thus inhibiting bicarbonate reabsorption along the ascending part of the loop. In addition, mannitol infusion may lead to a significant bicarbonaturia by the induction of extracellular volume expansion. Indeed, this factor has been shown to be one of the major regulators of proximal tubule bicarbonate reabsorption (8). The bicarbonaturia, found in the clearance experiments, is in seeming contrast to the micropuncture data that show stimulation of bicarbonate reabsorption along the LOH. The discrepancy could be explained by the technical approach used during the micropuncture experiments: single LOH were perfused with an end-like proximal tubule solution without mannitol. Under these conditions, we were able to characterize only the effect of mannitol on the basal lateral site of the LOH cells, i.e., reduced medullary tonicity, while avoiding the intraluminal osmotic effect linked to its characteristic of nonreabsorbable solute. The apparent contradiction between clearance and microperfusion data, obtained without mannitol in the perfusate, is very important, because it demonstrates that, in the presence of a dissipated medullary osmotic gradient and with no additional osmotic substance facing the luminal membrane, the effect of mannitol infusion is an increase in bicarbonate reabsorption along the LOH.

Intravenous infusion of mannitol leads to significant plasma volume expansion, as reflected by the large decrease in packed cell volume (see Table 1). This factor has been demonstrated to alter renal bicarbonate handling (8). Therefore, to support the hypothesis that mannitol action on net bicarbonate reabsorption could be related to alterations of medullary tonicity, we performed a new set of experiments according to a protocol where the plasma volume was less affected.

Hypotonic infusion experiments. In this set of experiments, the intravenous sodium chloride and sodium bicarbonate load remained constant, and the treatment was equivalent to infusion of 25 µl · min¹ · 100 g body wt¹ of pure water. This is very important, because we have previously shown that changes in sodium intake may influence bicarbonate reabsorption along the LOH (6). Similar to hypertonic mannitol infusion, hypotonic saline infusion stimulates bicarbonate reabsorption along the LOH. These results confirm the findings of our previous set of experiments, and they may indicate that the observed alterations of bicarbonate reabsorption were independent from changes in sodium load and/or extracellular volume.

Several points deserve consideration before interpretation of the results. Whereas in the mannitol experiments the osmolality of renal medulla was certainly reduced, this is not necessarily true following the hypotonic protocol. We have not measured the osmolality of different parts of the renal parenchyma under hypotonic infusion. However, it is well known that the osmolality of the renal parenchyma gradually increases from the corticomedullary boundary to the tip of the papilla (34). It has also been demonstrated that water diuresis does not affect mean tissue osmolality of the renal cortex, but it prevents the progressive increase of tissue osmolality along the whole renal medulla (10, 22). The effect is mainly due to decrease in urea content (28) and alteration in electrolyte concentration (21, 22). The changes in medullary tonicity will also influence the outer medulla (21, 22), where the medullary TALs of cortical nephrons are located. Since this segment contributes most to the total TAL bicarbonate absorption, due to its large intraluminal bicarbonate concentration (14), it is conceivable that variations in its bicarbonate transport rate will affect the bicarbonate reabsorption of the whole loop.

Bicarbonate excretion increased during the clearance experiments (see Table 3), contrasting the increase in net bicarbonate reabsorption along the perfused LOH. This seeming discrepancy is best explained by the absence of mannitol in the intraluminal solution used for micropuncture, whereas mannitol was present in the glomerular filtrate of the clearance experiments. The reduced mannitol concentration (1%) in this set of experiments will also explain the modest increase in bicarbonate excretion (2-fold) compared with the previous set, in which bicarbonate excretion increased 15-fold as consequence of the infusion of 9% mannitol solution.

In vivo and in vitro studies (2, 12) have shown that vasopressin, at physiological concentrations, inhibits bicarbonate transport along specific segments of the LOH. In our present experiments, it is very difficult to envisage the plasma concentration of the hormone; it is well established that vasopressin secretion is under the control of specific osmoreceptors that are very sensitive to a number of different variables, with the most important being the osmotic pressure of body water. Sodium has been shown to be a potent stimulator of vasopressin release. However, mannitol also, especially when infused intravenously, is very effective to raise vasopressin plasma level. Therefore, at least in the mannitol protocol, the microperfusion experiments have been performed, most probably, in a condition of high plasma vasopressin level, a situation that according to several reports should inhibit and not stimulate the bicarbonate reabsorption along the LOH. Infusion of hypotonic solution should have inhibited vasopressin release. It is possible, therefore, that in the second set of experiments, stimulation of bicarbonate reabsorption could have been partially due the low level of vasopressin. However, it should be emphasized that there is no experimental evidence indicating that low plasma levels of vasopressin stimulate bicarbonate reabsorption. Moreover, Bichara et al. (2) have shown that under in vivo conditions, at least in thyroparathyroidectomized somatostatin-infused rats, the vasopressin analog 1-desmopressin inhibits bicarbonate but stimulates chloride reabsorption at the level of LOH. Based on these observations, we cannot explain our data on the basis of altered vasopressin action.

The observed changes in bicarbonate transport along the LOH in conditions of reduced medullary osmolality could be explained by several cellular mechanisms. In vitro perfusion studies have clearly shown that hyperosmolality reduces bicarbonate reabsorption in rat medullary TAL (13). Moreover, it has been demonstrated that apical membrane Na⁺/H⁺ exchange (NHE3 isoform) is specifically inhibited by hyperosmolality (19, 31); finally, there is evidence indicating that peritubular hypertonicity blocks basolateral Cl conductance (17, 18). On the basis of these observations, we speculate that the reduced medullary osmolality may have increased bicarbonate and chloride transport along the LOH by activating luminal Na⁺/H⁺ exchange and basolateral chloride conductance. This last effect will facilitate bicarbonate reabsorption, since the chloride conductive pathway allows cellular loss of both chloride and bicarbonate (32). This hypothesis has been partially confirmed by recent in vitro microperfusion study demonstrating that hyposmolality stimulates bicarbonate reabsorption in the rat medullary TAL by activating apical membrane Na⁺/H⁺ exchange (30).

An additional and interesting finding of this study is the close correlation between urinary osmolality and GFR during the hypotonic protocol (Fig. 4), which is similar to recently published data (4) achieved in conscious animals, where chronic alterations in urine concentrating activity was induced either by increasing water intake or by constant intraperitoneal infusion of vasopressin analog. These have also been confirmed in studies on normal humans treated with intense hydration (24). Although, at present, there is not a clear explanation for these findings, it has been hypothesized that these could be related to the action of vasopressin on intrarenal urea recycling. The urea recycling increases its concentration in the lumen of TAL; since urea is an osmotically active solute, its accumulation will reduce the osmotically driven water leakage in this nephron segment. This effect will modify the NaCl concentration of the tubular fluid at the level of the macula densa and therefore alter the tubuloglomerular feedback control of GFR (4). Whatever the mechanism involved, it is clear that these results could have relevant clinical implications: they may indicate that chronic elevation of medullary tonicity, such as during high protein intake and diabetes mellitus, will induce a persistent increase in GFR, whereas a reduction in urine concentrating activity may reduce the progression of chronic renal failure by inhibiting the glomerular hyperfiltration of the surviving nephrons.

In conclusion, the data collected in the present study demonstrate that hypertonic mannitol and hypotonic saline infusion both increase urinary bicarbonate excretion and, at the same time, stimulate net bicarbonate reabsorption along in vivo microperfused LOHs. Since the reduction in medullary tonicity is one factor shared by these two experimental designs, it is likely that alterations in medullary osmolality may be at least partially responsible for the observed increase in electrolytes transport rate along the microperfused LOHs.