- Email: [email protected]
Ionic requirements of proximal tubular fluid reabsorption: Flow dependence of fluid transport
Kidney International, Vol. 20 (1981), pp. 580—587
Ionic requirements of proximal tubular fluid reabsorption: Flow dependence of fluid transport R. GREEN, R. J. MORIARTY, and GERHARD GIEBIscH Departments of Physiology, University of Manchester, Manchester, England, and Yale University School Connecticut
Ionic requirements of proximal tubular fluid reabsorption: Flow dependence of fluid transport. The effects of changes in luminal flow rate on fluid
absorption in rat renal proximal convoluted tubules were studied by continuous luminal and pentubular microperfusion methods. Luminal flow rate was varied over a range from 5 to 45 nlanin', and the effects of
transepithelial chloride and bicarbonate gradients were tested. Fluid
of Medicine, New Haven,
transport actifde sodium et les gradients transepitheliaux d'anions. Les experiences de perfusion luminale et péritubulaire avec des solutions symCtriques dépourvues de bicarbonate ci ne contenant que du tampon phosphate montrent que même dans ces conditions les mouvements de liquide dépendant d'un transport actif sensible au cyanure augmentent avec Ic debit.
absorption across the proximal convoluted tubule increased with luminal flow rate in the absence of luminal bicarbonate and organic solutes but in the presence of transepithelial chloride and bicarbonate gradients and active sodium transport. Augmenting perfusion rate from 5 to 45 nlmin' resulted in an increase of volume absorption from 0.49 to 3.37 nlmin' per millimeter length of tubule-'. The chloride concentration change in the
An increasing body of evidence suggests that, in the absence from the lumen of such reabsorbable solutes as glucose, amino acids, and bicarbonate, fluid reabsorption along the middle and collected perfusate decreased from 5.9 to 2.6 mEq1iter'mm length terminal portions of superficial convoluted [1—61 and straight tubule-' over the same perfusion range, Thus, tubular chloride concentraproximal tubules [7, 8] depends partially on, and is passively tion rises with perfusion rate such that the steepest transepithelial chloride gradients are maintained at the highest flow rates. Flow dependence driven by transepithelial concentration gradients of anions such as chloride and bicarbonate. These solute gradients arise from continued, albeit at reduced rate, in the absence of active sodium transport (cyanide perfusion) but in the presence of chloride and bicarbonate preferential bicarbonate reabsorption in the early part of the gradients. Flow dependence disappeared in the absence of both active proximal tubule [9, 10]. As a result of such transport activity in sodium transport and transepithelial anion gradients. Luminal and pertitubular perfusion experiments with symmetrical bicarbonate-free solutions the earliest part of the proximal convoluted tubule, there is a higher chloride and a lower bicarbonate concentration in the that contained only phosphate buffer showed that even under those conditions fluid movement driven by cyanide-sensitive active transport remaining tubular fluid. Given the higher chloride than bicarincreased with flow rate. bonate permeability of the epithelium of superficial proximal Nécessités ioniques de Ia reabsorption tubulaire proximale: Dépendance tubules, a lumen-positive electrical potential (PD) develops, thus generating a driving force favoring passive net sodium du transport de liquide vis-à-vis du debit. L'effet des modifications du debit luminar sur l'absorption de liquide par Ic tube contourné proximal de rat a été étudié au cours de perfusions luminales et pérituhulaires. Le debit luminal a vane de 5 a 45 nFmin' et les effets des gradients transCpithéliaux de chlore et de bicarbonate ant ete évalués. L'absorption de liquide a travers Ic tube contourné proximal a augmente avec le debit luminal en l'absence de bicarbonate luminal et de solutCs organiques mais en presence de gradients trans-Cpithéliaux de chiore et de bicar-
efflux from the lumen [10—141. In superficial proximal convolut-
bonate et de transport actif de sodium. L'augmenlation du debit de perfusion de 5 a 45 nl'min' a entrainé une augmentation du debit d'absot-ption de 0,49 a 3,37 nlmin' par millimetre de tubule.-'. La diminution de Ia concentration de chlore dans le perfusat collecté a
tive sodium translocation, are assumed to be through the
evolue de 5,9 a 2,6 mEq1iter' par millimetre de tubule-' en fonction du debit de perfusion. Ainsi Ia concentration tubulaire de chlore augmente
avec le debit de perfusion de idle sorte que les gradients trans-
ed tubules, it has also been suggested that the larger reflection coefficient of sodium bicarbonate compared with that of sodium chloride creates an additional osmotic driving force for net sodium reabsorption by solvent drag [4, 14]. These passive transport modes, that is, electrodiffusion of sodium and convec-
paracellular shunt across the proximal tubular epithelium [4, 141.
The present series of experiments addresses (I) the flow dependence of such ionic gradients along the proximal convo-
luted tubule, and ultimately related to this problem, (2) the dépendance vis a vis du debit se maintient, encore que d'une facon quantitative contributions of active and passive solute transréduite, en l'absence de transport actif de sodium (perfusion de port, respectively, to the flow dependence of fluid transport cyanure) mais en presence de gradients de chiore ci de bicarbonate. La across the proximal convoluted tubule. To this end, we have épithéliaux les plus élevCs sont maintenus aux debits les plus grands. La
dependance vis a vis du debit disparait quand sont a Ia fois absents Ic
Received for publication May 28, 1980 and in revised form January 23, 1981
carried out microperfusion of the rat nephron in vivo. In these experiments, dependence of reabsorption on rate of perfusion, as well as on artificially imposed anion gradients, was tested in the presence and in the absence of active transport. We can
show that a significant fraction of fluid reabsorption in the
0085-2538/81/0020-0580 $01.60
proximal tubule depends on the maintenance of transepithelial
© 1981 by the International Society of Nephrology
chloride and bicarbonate concentration gradients. We also 580
581
F/ow dependence of proximal fluid absorption Table 1. Composition of Ringer perfusates°
Chloride Chloride + cyanide Bicarbonate -F cyanide Chloride-phosphate Ringer Chloride-phosphate + cyanide Bicarbonate-phosphate Ringer
NaCI
NaHCO3
152 150 123 136 134
0 0 30 0 0 30
106
KCI
CaCl2
Na,HPO4
NaH2PO4
5
2 2 2
0
0
5
0.6
7.25
5
0.6
7.25
0.6
7.25
1.55 1.55 1.55
5
5
5
1
I
Urea
NaCN
5 5 5 5 5 5
0 2 2 0 0
a All solutions contained '4C-inulin (50 Ciml I) and 0.1% lissamine green. Values are in mmoles per liter
fluid given is an approximate concentration only. Solutions driven fluid absorption with increasing perfusion rate. Some of were prepared that were hypertonic and were finally diluted to these results are in agreement with those of Schafer et al match the osmolality of a plasma sample obtained from the tail obtained in vitro in the straight descending portion of the rabbit vein of the anesthetized, operated animal. This ensured that although there were differences in plasma osmolality of the experproximal tubule [151. imental animals (ranging from 285 to 303 mOsmkg water) the Methods osmolality of the tubular perfusate matched that of each rat. Series 1. To avoid the complicating presence of bicarbonate All experiments were performed on male Sprague-Dawley
observed in our experiments a marked increase of actively
rats weighing 150 to 200 g, anesthetised with 100 mgkg'
in the tubular lumen (at least in the initial fluid), we used a
i.p. mactin ([5-ethyl-5-( I '-methyl-propyl)1-2-thiobarbiturate) and placed on a thermostatically heated table to keep body tempera-
bicarbonate-free perfusate with a buffering capacity supplied by 2 mM phosphate. This solution was gassed with 100% oxygen, and the final pH was 6.8. Tubules were perfused at 5, 1:5, 30, and 45 nlmin± The length of the perfused segments (expressed
ture at 38° C. The operative procedure has been described previously [31 so it is only reported briefly here. Catheters were
inserted into the left jugular vein for infusions, and the right in mm) was, at5 nlmin, 1.99 0.38 mm; at 15 nlmin, 2.17 carotid artery to monitor systemic blood pressure; a tracheosto- 0.21 mm; at 30 n1min, 1.70 0.25 mm; at 45 nlmin. 1.65 my was performed. The left kidney was exposed via a flank 0.17 mm. Series 2. Inhibition of active transport processes with cyanide incision, placed in a plastic (Lucite) kidney cup and covered with liquid paraffin heated to 37° C; the left ureter was catheter- leaves only the passive components of transport. The experiized. The transit time of lissamine green was determined, and ments in series 1 were repeated using a perfusate in which 2 mtvi any animal whose blood pressure fell below 100 mm Hg, or cyanide was substituted for 2 m chloride. The solution was which had a proximal transit time of more than 13 sec, was gassed with 100% oxygen, and the final pH was 7.6. The discarded. Every animal received 0.99k saline, I ml, as a alkaline pH is due to the low buffer capacity of the perfusion priming dose during surgery followed by a continuous infusion fluid and the strongly alkalinizing effect of cyanide. Perfusions were performed at 5, 15, 30, and 45 nFmin. The length of the at 20 lmin. Randomly selected proximal convoluted tubules were per- perfused segments (expressed in mm) was, at 5 nlmin', 1 77 fused with a 1-lampel-pump according to the technique previ- 0.15 mm; at 15 nlmin ', 1.69 0.16 mm; at 30 nlmin, 1.65 ously described [3, 161 with the exception that in these experi- 20mm; at45 n1min, 1.55 0.11 mm. Series 3. This series of experiments was carried out to test the ments the rate of infusion was not constant but varied, being either 5, 15, 30, or 45 nlmin'. The majority of tubules were effects of the inhibition of active transport under conditions in perfused at two separate rates, and the sequence of perfusion which the role of a transepithelial chloride gradient is minirates was altered in a random fashion. After collection at the mized. Thus, to inhibit active transport processes, cyanide was first rate, at least 3 mm (and usually 5) were allowed for the substituted for 2 m chloride in bicarbonate Ringer so as to pump to equilibrate at the new speed. Fluid was collected from prevent the rise in luminal chloride along the proximal convothe perfused tubule over a timed interval with a distal oil block luted tubule and to provide no passive driving force. The solution was gassed with 5% carbon dioxide and 95% oxygen. in place. Two series of experiments (series 4 and 5) used simultaneous and the final pH was 7.7 to 7.8. Again, the disproportionately perfusion of both tubules and capillaries, as previously de- alkaline pH of this soluton is due to the presence of cyanide. In scribed [17]. It was technically very difficult in these experi- this series, perfusions were performed at 15 and 45 nlrnin ments to perfuse at more than one perfusion rate so paired data The length of the perfused segments (expressed in mm) was, at were not obtained. At the end of the experiments, perfused 15 nlmin, 1.90 0.23 mm; at 45 nlmin, 1.44 0.13 mm. Additional experiments were carried out in an attempt to tubules were filled with silicone rubber (Canton Biomedical Inc., Boulder, Colorado), the kidneys were excised, allowed to modify transepithelial anion gradients by changing the peril ubustand overnight in deionized water at 4° C, and subsequently lar fluid environment. Three series of experiments were carried digested in 10% sodium hydroxide. The filled tubules were then dissected out and their length measured using a camera lucida attachment on a Leitz TS steromicroscope. In all, five series of experiments were performed with differ-
out in which the buffering power of both tubular and peritubular perfusates was increased with phosphate. Series 4. The tubular perfusate was chloride-phosphate Ringer, which was gassed with 100% oxygen and had a pH of 7.4,
ent tubular perfusates (see Table I). The composition of the
whereas the peritubular perfusate was bicarbonate phosphate
582
Green ci a!
Ringer with 100 g1iter' of albumin, which was gassed with •—• Chloride Ringer
carbon dioxide in 95% oxygen and also had a final pH of 7.4 (see Table 1). This gave a "normal" chloride concentration gradient and was used as a control series. Perfusion was performed at IS
x— — —x Chloride Ringer + 2 m cyanide
and 45 nlmin. The length of the perfused segments (expressed in mm) was, at 15 nlmin', 1.65 0.22 mm; at 45 nlmin, 1.42 0.09 mm. Series 5. These experiments were designed to eliminate the chloride gradient across the tubule normally created by preferential bicarbonate reabsorption, bicarbonate being omitted from both perfusion solutions. The buffering in these perfusates was provided by phosphate, the tubular perfusate being chloride phosphate Ringer and the capillary perfusate chloride phosphate Ringer with 100 g-liter albumin. Both solutions were gassed with 100% oxygen and had a final pH of 7.4. Perfusion of the lumen
0
was carried out at 15, 30, and 45 nFmin. The length of the
.0
perfused segments (expressed in mm) was, at 15 nlmin, 1.31 0.2 mm; at 30 nlmin, 1.43 31 mm; at 45 nlmin, 1.40 0.15
a,
3.0 B
-
Bicarbonate Ringer + 2 mta cyanide
B B
2.0 a
0 0 0.
0
+
Co
T
1.0
T—
mm.
Series 6. This was the same as series 5 but cyanide was added to demonstrate the effects of inhibiting active transport. Perfu-
sion of the lumen was carried out at 30 nl'min'. The length of the perfused segments (expressed in mm) was 1.20 0.12 mm. Analyses. The volume of each collected sample was measured in a calibrated capillary, and aliquots were taken, togeth-
er with aliquots of the perfusate, for counting in a liquid scintillation spectrometer (Intertechnique SL 30, France). Aliquots of the remaining fluid were used to measure chloride by an electrometric titration technique first described by Ramsay, Brown, and Croghan [18]. The concentration of chloride was measured in aliquots of plasma, obtained at the end of the
15
30
45
Perfusion rate, nImin'
Fig. 1. Relationship between reabsorptive fluid flux and tubular perfu-
sion rate in three series of experiments. Data are expressed as mean values SEM. The lines drawn for each series of experiments are least square regression lines.
experiment by centrifugation of a small sample (200 p.1) of blood
from a tail vein, and in perfusion fluid. Thereafter net fluid absorption was calculated as
V = V (1 — In0/In) where V0 is the initial perfusion rate and In0 and In are the inulin concentrations in the perfusate and collected fluid, respectively. This was factored by the length of the perfused segment to give the rate per millimeter length of tubule. The total amount of inulin collected was calculated for each
relationship between the perfusion rate and the net fluid absorp-
tion was observed (Fig. I and Table 2). Whereas only 0.49 nlmin' of fluid was reabsorbed at a perfusion rate of 5 nlmin, 3.37 nl-min' was absorbed when the perfusion rate was increased to 45 nlmin.
The magnitude of the transepithelial chloride concentration difference is not precisely known, but making reasonable asperfusion, and if it was not 90 to 110% of the amount perfused, sumptions for plasma water and Donnan factor corrections, we the sample was discarded, Overall, the recollection of inulin can estimate a pertibular chloride concentration of some 114 was 99.8 0.2% and in no series was it significantly less than mM [19]. Thus, under the present experimental conditions, a 100%. About 10% of samples had to be discarded. maximal effective chloride concentration difference operative Results are presented as means SEM for each infusion. across the proximal tubular epithelium would be of the order of Preliminary analysis indicated that there was no significant some 45 mEq/liter. difference between the paired and the unpaired data, so all were Cyanide, in the presence of the imposed gradient, inhibits included, and the analysis performed as unpaired data. Compar- active transport processes and allows the expression of only isons were performed using Student's (test. passive forces. Yet there was still a significant linear correlation between perfusion rate and fluid reabsorption, which presumably reflects these passive forces and the way they are affected Results by altering the perfusion rate (Table 2, Fig. 1). It is instructive There were no significant differences in blood pressure, to note that this can account for only about 30% of the plasma chloride concentration, mean length of tubule perfused, reabsorption, however—a figure which agrees with previously and the percentage of inulin re-collected between any of the published results [1, 31. The slope of the regression line with series. cyanide present was significantly different from that when Chloride perfusate. With a fluid having a chloride concentra- cyanide was absent. Thus, there not only continues to be flowtion gradient from lumen to capillary of some 40 mmoles1iter, dependent fluid reabsorption but, importantly, the difference in there is a passive driving force (the chloride and bicarbonate fluid absorption in the presence of a transepithelial chloride gradients) to aid sodium and fluid reabsorption. A significant gradient, with and without transport inhibitor, increased mark-
583
Flow dependence of proximal fluid absorption
Table 2. Net fluid reabsorption' Net fluid reabsorptiona
nImin 'nim length t,ib,ele
At 5 nImin
a) Tubular Perfusion Tubular perfusion Series 1: Chloride Ringer Series 2: Chloride Series 3:
2.013
3.375
0.050
(15) 0.413 0.060
0.201 (16) 0.692 0.180
(Il)
(8)
(5) Bicarbonate
At45nlniin
1.423
(22)
0.201
At 30 nImin
0.042
0.492
+ cyanide
Ringer
At 15 nl'min
+ cyanide
0.145
Ringer
0.103
(9)
—
0.164
0.210 (15) 0.989 0.180 0.430
0.082
(13)
capillary perfusionb Chloride phosphate; bicarbonate
(14)
Tubular and Series 4:
1.070
phosphate + albumin
0.129
2.924
Series 5: Chloride phosphate; chloride phosphate + albumin Series 6: Chloride phosphate; chloride phosphate + cyanide + albumin
0.557
0.330
(14)
(14)
0.106
1.021
(12)
0.081
(6)
0.042
0.095
1.396
0.24
(12) —
(6)
a Numbers in parentheses represent number of perfusions. b Chloride
phosphate Ringer's solution was the tubular perfusate. The solution listed next was the capillary perfusate.
Table 3. Perfusion rate dependence of fluid transport across proximal convoluted rat tubulea
significant alteration in chloride concentration along the tubule.
Jv V
At 5
nlmin'
At 15 nImin
047 0.077
+
0.20
0.05
0.27
—
1.42 0.41 2.01
0.10 0.06 0.20
1.011
0.69 3.37 0.99
0.18
1.32
O.2l 0.26
2.38
—
+ At
45 nlmin'
which a chloride phosphate Ringer perfusion fluid was present in the tubules and bicarbonate-phosphate Ringer used to per-
—
+
At 30 nImin
Perfusion of both tubules and capillaries. The situation in
nImi,, 'mm
Cyanide
—
+
Measurements of the chloride concentration in the collected perfusate showed that at neither rate the perfusion was there a
a \ is perfusion flow rate; J, fluid transport; change in proximal tubular fluid transport. Minus and plus signs under Cyanide heading denote presence or absence of cyanide.
edly with perfusion rate. If it is accepted that the fraction of fluid reabsorption that is subject to cyanide inhibition represents active transport (see below), this observation clearly shows that fluid transport related to active transport increases with flow rate. This is also shown in Table 3, in which data on the perfusion rate dependence of fluid transport in the presence or absence of cyanide are compared. The increase in the active
transport component is apparent from the very dramatic increase of the rate of tubular reabsorption, AJ from 0.27 nI
min'mm' at a perfusion rate of 5 nlmin to a value of 2.38 nlmin'mm at a perfusion rate of 45 nlmin'. Such a flow dependent increase in active solute and fluid transport was also observed by Imai, Seldin, and Kokko [20] and by Schafer et al [15] in the isolated rabbit proximal tubule. That all fluid reabsorption in the absence of anion gradients is
indeed dependent on metabolic energy is indicated by the
fuse peritubular capillaries (series 4) is analogous to that in series 1 where a chloride gradient exists with bicarbonate present in the peritubular capillaries. It is noteworthy that the net fluid fluxes at perfusion rates of 15 and 45 nlmin are similar to those of the tubular perfusion experiments of series I (chloride Ringer in lumen, blood perfusion of peritubular capil-
laries), in spite of the double perfusion and other changes in composition of the perfusate (see Table 2), although in the double perfusion experiments slightly lower net reabsorptive fluxes are observed. When the chloride-phosphate Ringer in the peritubular capillaries and in the tubules had the same ionic composition (series 5), there was no significant chloride gradient that could drive passive reabsorption, and this is reflected by the lack of change in the chloride concentration of the collected tubular perfusate (see below). Figure 2 summarizes relevant data. Yet under these circumstances, without preferential bicarbonate reabsorption to produce transepithelial anion gradients, there was still a highly significant (P < 0.005) stimulation of fluid reabsorption at the higher perfusion rate, although such augmentation of net fluid transport is less than in the presence of anion gradients. Figure 2 summarizes these results and also includes. for comparison, data that were obtained when a significant chloride (and bicarbonate gradient) was imposed across the epithelium (chloride-phosphate Ringer in tubule, bicarbonatephosphate Ringer in capillary). It should be emphasized that in this last series of experiments, absence of bicarbonate in the peritubular fluid could by itself have an effect upon transepithelial fluid movement. The absence of bicarbonate from the peritubular fluid has been shown to reduce proximal tubular sodium and water transport
results with bicarbonate and cyanide perfusate. Although only two perfusion rates were used, there was no significant reabsorption of fluid with either perfusion rates (Table 2, Fig. I). Hence, under these experimental conditions, there is complete loss of the flow-dependency of transepithelial fluid transport. [1, 21]. This may account for the fact that the reduction of fluid
584
Green ci a!
reabsorption observed in series 5 (no chloride gradient, peritubular bicarbonate gradient present) is greater (48%) than that to be expected if 30% were due to passive reabsorption driven by the anion gradient. it is noteworthy that, despite the absence of both an anion gradient and of peritubular bicarbonate, there is still dependence of fluid reabsorption on perfusion rate, Again, the addition of cyanide to the tubular perfusion fluid completely abolished such transepithelial fluid transport. A summary of the changes in chloride concentrations in the different perfusion experiments is presented in Table 4. In the experiments using a bicarbonate-free, high chloride perfusion fluid (series I), there was a change in chloride concentration in collected fluid when compared with the original perfusate at all
the perfusion rates. This was most marked at the lowest perfusion rate and gradually diminished as the perfusion rate was increased. At the highest perfusion rate, the fall in chloride concentration was statistically significantly less than that at 5 nlmin. This reflects a time-dependent dissipation of the chloride gradient because the loss of chloride at lower perfusion rates results in a greater fall of the luminal chloride concentration'. The chloride concentration changes in the experiments using chloride Ringer plus cyanide are similar to those in which
•• Chloride phosphate Ringer in tubule; bicarbonate phosphate Ringer in capillary
X———* Chloride phosphate Ringer in tubule and capillary
3.0 -
A Chloride phosphate Ringer in tubule and capillary
+2 m cyanide
=
2.0 -
0 x0
I
0.
0
'-—I T
1.0 T .1.
bicarbonate-free chloride was used. Again, chloride loss from the
lumen results in a significantly larger drop in luminal chloride 15 30 45 concentration (P < 0.05) at a perfusion rate of 5 nlmin comPerfusion rate, nImin pared with 45 nlmin'. Inhibition of active transport processes Fig. 2. Relationship between reabsorptive fluid flux and tubular perfushould lead to cessation of hydrogen ion secretion and thereby sion rate in three series of experiments in which both tubules and prevent maintenance of the steady-state luminal bicarbonate perituhular capillaries were perfused. Data are expressed as in Fig. I. concentration. As a consequence, this would he expected to enhance the change in luminal chloride concentration. Our data show a slight trend in this direction but the difference of the perfused with plasma ultrafiltrate. Radtke et a! [25] showed that chloride concentration changes is not significant. That we have in rats a threefold increase in perfusion rate only increased fluid not observed a significant difference in our experiments could be reabsorption by 19% when the diameter was held constant, and due to the quite small differences in the absolute chloride concen- that increasing the diameter made no difference to the amount tration and that possibly not all of the chloride concentration reabsorbed. Similarly, Morel and Murayama [261 reported no changes are related to reciprocal alterations in luminal bicarbon- change of fluid reabsorption in a series of paired experiments ate. where rat proximal tubules were perfused at two different Discussion
Over the past few years, there has been considerable controversy about the relationship between the flow rate along renal
proximal tubules, the rate of reabsorption from the same tubules, and its relationship to glomerulotubular balance. Some
authors have described a linear relationship whereas others
perfusion rates; Morgan and Berliner [27] also observed no change in proximal tubular fluid absorption with an increase in perfusion rate.
in contrast to these studies there have been reports of a definite relationship between reabsorption and flow rate, and these are exemplified by the studies of Wiederholt et a1 1281, Bartoli et a! [29], Häberle et a! [30] and Senekjian et al [31]. Imai, Seldin, and Kokko, working with isolated rabbit tubules, showed a relationship only at flow rates less than 11 nlmin
have been unable to demonstrate one. In this respect, Burg and Orloff [24] reported no significant effect on fluid reabsorption of increasing the perfusion rate from [20]. Although some of these authors have suggested that such a 8 to 19 nl-min in isolated rabbit proximal convoluted tubules relationship contributes to glomerulotubular balance, they did not claim it was solely responsible for such a balance. Bartoli et a!, for instance, showed that doubling the perfusion rate only 'Although the perfusion fluids are initially bicarbonate-free, the flowincreased reabsorption by 60% [29]. dependent changes in chloride concentration signal entry of bicarbonate The current results in proximal convoluted tubules in rats and during the experiments. Because pH measurements were not carried out on the effluent perfusate, the extent of luminal acidification those by Schafer et a! 1151 on proximal straight tubules in achieved at the different flow rates was not assessed. Hence it is rabbits indicate that there is indeed, particularly under experiimpossible to determine whether the fall in chloride was completely mental conditions in which transepithelial anion gradients are matched by reciprocal bicarbonate concentration changes. Assuming maintained, a highly significant correlation between the rate of peritubular Pco2 values in the range of 40 to 65 mm Hg 122, 231, the perfusion and that of reabsorption. We have previously demonrange of bicarbonate concentrations resulting from transepithelial anion strated, in perfused proximal rat tubules in vivo that some 20% exchange is consistent with the maintenance of significant pH gradients across the proximal convoluted epithelium. to 30% of sodium reabsorption is passive and dependent on a
585
Flow dependence of proi,nal fluid ah,orptio,z Table 4. Changes in chloride concentration of perfusate
Chloride concentration changes, rn,nolesliter 'rn/n length tubule
At 5
Chloride Ringer Chloride Ringer + cyanide
nI-mm
At 15 nImin
—5.9
II
- 5.3
—6.7
1.1
—4.9
2.1 1.0
At 30 nI-mm
--3.9 --4.6
1.8 1.2
'
At 45 nI-mm —2.6
0.9
—3.5 + 1.0
transepithelial chloride and bicarbonate concentration gradient [1, 3]. The present results confirm that this is applicable at each perfusion rate, that is, the proportion of reabsorption dependent on such passive forces is constant. Thus, when active transport is blocked by cyanide and a chloride gradient is present, the
[4, 5, 7, 8, 11, 14, 33]. Thus, perfusion solutions in which chloride replaces bicarbonate, even though they have the same
amount reabsorbed is about 3O of that reabsorbed when
Jacobson [121, and Berry and Rector [101 have pointed out that
cyanide is absent (see Fig. 1). Recent experiments by Chantrelle and Rector, using deletion of potassium from the peritubular fluid to inhibit active sodium transport, have also shown that in the presence of similar anion gradients, only sonic 3OY
this type of osmotic disequilibrium between luminal and peritu-
of fluid reabsorption remains intact [321, that is, that two thirds are due to active transport.
This conclusion is further supported by data from experiments where both tubules and capillaries were perfused. When there was a chloride gradient between lumen and capillaries (series 4), there was increased reabsorption when compared with experiments (series 5) in which there was no gradient. Several factors are operating in this situation, however, and as well as the abolition of the chloride gradient, there is also the removal of the stimulating effect on fluid reabsorption of peritubular bicarbonate, which is considerable [1, 21]. A striking observation concerns the fact that active reabsorption of sodium is also dependent on the luminal flow rate. Two lines of evidence support this view. The first is based on the demonstration that, in the presence of anion gradients, the cyanide-inhibitable fraction of transepithelial fluid transport increases with luminal flow rate. Thus, when passive reabsorption (series 2) is subtracted from total reabsorption (series 1), there is still significant variation of reabsorption with flow rate.
The second line of evidence for flow-dependence of active
transport is given by experiments in which there was no chloride gradient (series 5). In the absence of a chloride gradient, without glucose or other organic substances in the perfusate and in the absence of bicarbonate, the stimulating effect of increasing flow rate upon reabsorption still exists. That this is an active, metabolically driven process is confirmed by its total abolition by cyanide.
The mechanisms responsible for the dependence of both active and passive fluid absorption on luminal flow rate need further consideration. With respect to the flow dependence of passive fluid transport, it is suggested that the tendency of increased perfusion rate to maintain the transepithelial chloride
and bicarbonate gradients along the nephron is important. Whereas at lower flow rates dissipation of the transepithelial anion gradients is favored because of passive chloride exit from and passive influx of bicarbonate into the tubular lumen, such events are less effective to modify anion gradients at the higher flow rates. It has been suggested that passive fluid absorption under these conditions is due to the relatively larger reflection
coefficient of bicarbonate compared to chloride and the high osmotic water permeability of the proximal convoluted tubule
osmolarity as the peritubular environment, are effectively hypo-
tonic with respect to the lower chloride peritubular fluid. Andreoli and Schafer et al [7, 8, 15, 33], Barratt et al [5], bular fluid may constitute an effective driving force for fluid
absorption. As these anion gradients are more effectively maintained at higher flow rates, passive fluid transport is stimulated pan passu. In addition, Schafer et a! [15] iave
suggested as a second mechanism to account for flow-dependent stimulation of fluid transport that osmotic equilibration is also minimized as tubular fluid delivery is enhanced. Considering the possible mechanisms responsible for the dependence of an active solute (and fluid) transport component on perfusion rate, several possibilities must be considered, but the situation is far from clear. As we have demonstrated in this study, a flow-dependent increase in the active component of solute (and fluid) transport occurred both in the presence as well as in the absence of transepithelial anion gradients. Stimulation of active solute (sodium) transport in the presence of high luminal chloride may be due to stimulation of sodium entry into proximal tubule cells across brushborder membranes by the favorable chloride gradient. As a consequence of this event, active sodium transport across the peritubular membrane would also be augmented. Suggestive evidence for coupled sodium chloride transport across the luminal cell membrane of proximal tubule cells, either by neutral sodium chloride cotransport [34],
or by a double exchange mechanism, one for sodium and hydrogen, and one for chloride and hydroxyl [11, 35] has recently become available. These mechanisms, in addition to other co- and counter-transport mechanisms involving sodium translocation across the luminal brushborder membrane of proximal tubule cells [II, 36—38], suggest an important role for anion concentration gradients in modulating sodium entry into proximal tubule cells. Because transepithelial chloride gradients are maintained more effectively at higher axial flow rates, it
is possible that this constitutes an effective mechanism of stimulating active sodium transport in the presence of flowdependent anion gradients. An additional aspect of the dependence of active solute (fluid) transport concerns this phenomenon in the absence of transepithelial anion gradients. Originally, Gertz et al [39] suggested that the relationship was with diameter of the tubule, but this is not now thought to be so [24, 25]. Other possibilities that were considered are some sort of mechanism akin to a flow reactor [39, 401, and the intriguing possibility that the layer of fluid trapped between microvilli could act as an unstirred layer, the depth of which is altered by the flow rate [41]. Although this
latter mechanism was applied to passive transport, similar
586
Green et a!
arguments could be applied for the passive entry of sodium into the cell prior to its active extrusion across the peritubular cell membrane.
It has recently been suggested that active solute (sodium) absorption across the proximal tubular epithelium results directly in the generation of small osmolarity differences between
References 1. GREEN R, GIEBISCH G: Ionic requirements of proximal tubular
sodium transport: I. Bicarbonate and chloride. Am J Physiol 229:1205—1215, 1975
2. GREEN R, GLEBI5cH G: Ionic requirements of proximal tubular sodium transport: II. Hydrogen ion. Am J Physiol 229:1216—1226, 1975
the luminal and peritubular solutions [42, 43]. Thus, fluid absorption across the proximal tubular epithelium may be driven, in addition to the above mentioned differences in effective osmotic pressures due to anion gradients, by the additional driving force of osmotic pressure differences be-
3. GREEN R, BISHOP JVH, GIEBISCH 0: Ionic requirements of proximal tubular sodium transport: III. Selective luminal anion substitution. Am J Physiol 236:F268—F277, 1979 4. NEUMANN KH, RECTOR FC JR: Mechanism of NaCI and water reabsorption in the proximal convoluted tubule of rat kidney. J C/in Invest 58:1110—1118, 1976
tween the lumen and the peritubular fluid [42, 43]. Experimental
5. BARRATT U, RECTOR FC, KOKKO JP, SELDIN DW: Factors involving the transepithelial potential difference across the proximal tubule of the rat kidney. J C/in invest 53:454—460, 1974
evidence in support of small but significant osmotic pressure differences across proximal convoluted rat tubules has recently been reported [16]. The dissipation of such an osmotic disequilibrium would depend, in principle, on two opposing events. On the one hand, active salt (sodium) absorption, in the presence of a given water permeability, generates and maintains osmotic disequilibrium. The flow-dependent increase in such active sodium transport would assure maintenance of luminal hypotonicity with increased flow along the tubule. On the other hand, passive osmotic equilibration would tend towards dissipation, in a flow-dependent manner, of such an osmotic disequilibrium. Given the flow dependence of active transport, osmotic disequi-
librium, that is, luminal hypotonicity and transepithelial fluid absorption could be more effectively maintained with increase
in axial volume delivery if the time course of establishing hypotonicity and its dissipation were different. Clearly the observed increase in active solute (sodium) transport with flow rate, although poorly understood as to its cellular mechanisms,
is an important factor to accentuate both the generation and maintenance of osmotic disequilibrium over a wide range of
6. RECTOR FC JR, MARTINEZ-MALDONADO, BRUNNER FP, SELDIN
DW: Evidence for passive reabsorption of NaCI in proximal tubule of rat kidney (abst). iC/in Invest 45:1060, 1966 7. SCHAFER JA, PATLAK CS, ANDREOLI TE: A component of fluid absorption linked to passive ion flows in the superficial pars recta. J Gen Physio/ 66:445—471, 1975 8. SCHAFER JA, PATLAK CS, ANDREOLI TE: Fluid absorption and active and passive ion flows in the rabbit superficial pars recta, Am J Physio/ 233:F154, 1977 9. MALNIC G, STEINMETZ PR: Transport process in urinary acidilication. Kidney mt 9:172—188, 1976 10. BERRY CA, RECTOR FC JR: Active and passive sodium transport in the proximal tubule. Mineral Electrolyte Metab 4:149—160, 1980 11. BERRY CA, WARNOCK DG, RECTOR FC: Ion selectivity and proximal salt reabsorption. Am J Physiol 235:F234—F245, 1978 12. JACOBSON HR: Characteristics of volume reabsorption in rabbit superficial and juxtamedullary proximal convoluted tubules: J C/in Invest 63:410—418, 1979 13. SEELY JF, CHIRIT0 E: Studies of the electrical potential difference in rat proximal tubule. Am J Physio/ 229:72—80, 1975 14. FROMTER E, RUMRICH G, ULLRICH KJ: Phenomenologic descrip-
tion of Nat, CU' and HC13 absorption from proximal tubules of the rat kidney. Pfluegers Arch 343:189—220, 1973
tubular flow rates. The relevance of these findings to glomerulotubular balance
15. SCHAFER JA, TROUTMAN SL, WATKINS ML, ANDREOLI TE: Flow
is difficult to assess under normal in vivo conditions. Even under conditions described here (series 1), a ninefold increase
16. BISHOP JHV, GREEN R, THOMAS S: Effect of glucose on water and
in perfusion rate only resulted in a sixfold increase in reabsorp-
tion rate. This figure of 65% glomerulotubular balance is comparable to that reported by Bartoli et al [29] and implies that other factors, such as peritubular protein concentration [17, 44],
or humoral factors [30], may be important. Both active and passive transport components may be involved in the flowdependent stimulation of proximal tubular sodium transport.
dependence of fluid transport in the isolated superficial pars recta. Kidney Int, this issue
sodium reabsorption in the proximal convoluted tubule of rat kidney. J Physiol (Lond) 275:481—493, 1978 17. GREEN R, WINDHAGER EE, GIEBIscH G: Protein oncotic pressure effects on proximal tubular fluid movement in the rat. Am J Physiol 226:265—276, 1974 18. RAMSAY JA, BROWN RHJ, CROGHAN PC: Electrometric titration of chloride in small volumes. J Exp Biol 32:822—829, 1955 19. ATHERTON J: Comparison of chloride concentration and osmolality
in proximal tubular fluid, peritubular capillary plasma and systemic
plasma in the rat. J Physio/ (Lond) 273:765—773, 1977 Besides the flow-dependent increase in the active component of solute (sodium) and fluid reabsorption, the flow dependence of 20. 1MM M, SELEMN DW, KOKKO JP: Effect of perfusion rate on the fluxes of water, sodium, chloride and urea across the proximal the passive component only applies once the chloride gradient convoluted tubule. Kidney mt 11:18—27, 1977
has been generated. Indeed, flow dependence of the active component, if it occurs also in the first part of the proximal
21. ULLRICH KJ, RADTKE HW, RUMRICH G: The role of bicarbonate
and other buffers on isotonic fluid absorption in the proximal convolution of the rat kidney. Pfluegers Arch 330:149—161, 1971 tubule, might be a means of ensuring that the chloride gradient was established at approximately the same distance along the 22. DUBOSE TD, PUCACCO LR, LUCCI MS, CARTER RW: Micropuncture determination of pH, pCO2 and total CO2 concentration in tubule at each flow rate. Alteration of this equilibrium point accessible structures of the rat renal cortex. J C/in Invest 64:476— could of itself alter glomerulotubular balance. 482, 1979 Acknowledgments Some of the results have been presented in abstract form in C/in Res 27:497A, 1979.
23. SOHTELL M: pCO2 of the proximal tubular fluid and the afferent arteriolar blood in the rat kidney. Acta Physiol Scand 105:137—145, 1979
24. BURG M, ORLOFF J: Control of fluid absorption in the renal proximal tubule. J C/in Invest 47:2016—2024, 1968
This research was supported by NIH grant AM 1743305 and by 25. RADTKE HW, RUMRICH 0, KLOSS 5, ULLRLCH KJ: Influence of
Medical Research Council Grant G973/2531B.
Reprint requests to Dr. G. Giebisch, Department of Physiology, Yale University School of Medicine, 333 Cedar Street, New Haven, Connecticut 06510, USA
luminal diameter and flow velocity on the isotonic fluid absorption and 36Cl permeability of the proximal convolution of the rat kidney. Pfluegers Arch 324:288—296, 1971 26. MOREL FF, MURAYAMA Y: Simultaneous measurement of unidi-
rectional and net sodium fluxes in microperfused rat proximal tubules. Pfluegers Arch 320:1—23, 1970
587
Flow dependence of proximal fluid absorption 27. MORGAN T, BERLINER RW: In vivo perfusion of proximal tubules of the rat: Glomerotubular balance. Am J Physiol 271:992—997, 1969 28. WIEDERHOLT M, HIERHOLZERK, WINDHAGER EE, GIEBISCH 0:
Microperfusion study of fluid reabsorption in proximal tubules of rat kidneys. Am J Physiol 213:809—818, 1967 29. BARTOLI E, CONGER JD, EARLEY LE: Effect of intraluminal flow on proximal tubular reabsorption. J C/in invest 52:843—849 30. HABERLE DA, SHIIGAI T: Flow-dependent volume reabsorption in the proximal convolution of the rat kidney: The role of glomerularborn tubular fluid for the maintenance of glomerulotubular balance,
in Newer Aspects of Renal Function, edited by VOGEL HG, ULLRICH KJ, Amsterdam-Oxford, Excerpta Medica, 1978, pp 198-206
31. SENEKJIAN JO, KNIGHT TF, SANSON SC, WEINMAN ED: Effect of
flow rate and the extracellular fluid volume on proximal urate and water absorption. Kidney mt 17:155—161, 1980 32. CHANTRELLE B, RECTOR FC JR: Active and passive components of
volume reabsorption in rat superficial proximal convoluted tubule (abst). C/in Res 28:44lA, 1980 33. ANDREOLI TE: Salt and water transport processes in the proximal straight tubule, in Renal Function, edited by GIEBISCU 0, PURCELL EF, New York. Josiah Macy Jr Foundation, 1978, pp 186—2 13
34. SPRING KR, KIMURA 0: Chloride reabsorption by renal proximal tubules of Necturus. J Me,nhr Biol 38:233—254, 1978 35. Lucci MS, WARNOCK DG: Effect of anion-transport inhibitors on NaCl reabsorption in the rat superficial proximal convoluted tubule. J C/in Invest 68:570—579, 1979
36. WINDUAGER EE, GIEBI5CH
0: Proximal sodium and fluid trans-
port. Kidney mt 9:121—133, 37.
1976 The renal handling of
BURG MB: sodium chloride, in The Kidney, edited by BRENNER BM, RECTOR FC Ja, Philadelphia, London, Toronto, Saunders, 1976, pp 272—3 17
38. WINDHAGER EE: Sodium chloride transport, in Membrane Tran-
0, TosTEsoN DC, Ussmc HH, Berlin-Heidelberg-New York, Springer-Verlag, 1978, vol 4A, pp
port in Biology, edited by GIEBIsCH 145—2 13
GERTZ KH, MANGOS JA, BRAUN 0, PAGEL HD: On glomerulartubular balance in the rat kidney. Pfluegers Arch 285:360—372, 1965 40. KELMAN RB: A theoretical note on exponential flow in the proxi39.
mal part of the mammalian nephron. Bull
Math Biophys 24:303,
1962
41. RICHARDSON 1W, LICKO V, BARTOLI E: The nature of passive flows through tightly folded membranes: The influence of nhicrostructure. J Membr Biol 11:293—308, 1973 42. ANDREOLI TE, SCHAFER JA: Effective luminal hypotonicity: the driving force for isotonic proximal tubular fluid absorption. 4m J Physiol 233:F89—F96, 1979
43. ANDREOLI TE, SCHAFER JA: Volume absorption in the pars recta
III. Luminal hypotonicity as a driving force for isotonic volume absorption. Am J Physiol F349—F355, 1978
44. WINDI-IAGER EE, LEWY JE, SPITZER A: Intrarenal control of proximal tubular reabsorption of sodium and water. Nephron 6:247—259, 1969
Ionic requirements of proximal tubular fluid reabsorption: Flow dependence of fluid transport R. GREEN, R. J. MORIARTY, and GERHARD GIEBIscH Departments of Physiology, University of Manchester, Manchester, England, and Yale University School Connecticut
Ionic requirements of proximal tubular fluid reabsorption: Flow dependence of fluid transport. The effects of changes in luminal flow rate on fluid
absorption in rat renal proximal convoluted tubules were studied by continuous luminal and pentubular microperfusion methods. Luminal flow rate was varied over a range from 5 to 45 nlanin', and the effects of
transepithelial chloride and bicarbonate gradients were tested. Fluid
of Medicine, New Haven,
transport actifde sodium et les gradients transepitheliaux d'anions. Les experiences de perfusion luminale et péritubulaire avec des solutions symCtriques dépourvues de bicarbonate ci ne contenant que du tampon phosphate montrent que même dans ces conditions les mouvements de liquide dépendant d'un transport actif sensible au cyanure augmentent avec Ic debit.
absorption across the proximal convoluted tubule increased with luminal flow rate in the absence of luminal bicarbonate and organic solutes but in the presence of transepithelial chloride and bicarbonate gradients and active sodium transport. Augmenting perfusion rate from 5 to 45 nlmin' resulted in an increase of volume absorption from 0.49 to 3.37 nlmin' per millimeter length of tubule-'. The chloride concentration change in the
An increasing body of evidence suggests that, in the absence from the lumen of such reabsorbable solutes as glucose, amino acids, and bicarbonate, fluid reabsorption along the middle and collected perfusate decreased from 5.9 to 2.6 mEq1iter'mm length terminal portions of superficial convoluted [1—61 and straight tubule-' over the same perfusion range, Thus, tubular chloride concentraproximal tubules [7, 8] depends partially on, and is passively tion rises with perfusion rate such that the steepest transepithelial chloride gradients are maintained at the highest flow rates. Flow dependence driven by transepithelial concentration gradients of anions such as chloride and bicarbonate. These solute gradients arise from continued, albeit at reduced rate, in the absence of active sodium transport (cyanide perfusion) but in the presence of chloride and bicarbonate preferential bicarbonate reabsorption in the early part of the gradients. Flow dependence disappeared in the absence of both active proximal tubule [9, 10]. As a result of such transport activity in sodium transport and transepithelial anion gradients. Luminal and pertitubular perfusion experiments with symmetrical bicarbonate-free solutions the earliest part of the proximal convoluted tubule, there is a higher chloride and a lower bicarbonate concentration in the that contained only phosphate buffer showed that even under those conditions fluid movement driven by cyanide-sensitive active transport remaining tubular fluid. Given the higher chloride than bicarincreased with flow rate. bonate permeability of the epithelium of superficial proximal Nécessités ioniques de Ia reabsorption tubulaire proximale: Dépendance tubules, a lumen-positive electrical potential (PD) develops, thus generating a driving force favoring passive net sodium du transport de liquide vis-à-vis du debit. L'effet des modifications du debit luminar sur l'absorption de liquide par Ic tube contourné proximal de rat a été étudié au cours de perfusions luminales et pérituhulaires. Le debit luminal a vane de 5 a 45 nFmin' et les effets des gradients transCpithéliaux de chlore et de bicarbonate ant ete évalués. L'absorption de liquide a travers Ic tube contourné proximal a augmente avec le debit luminal en l'absence de bicarbonate luminal et de solutCs organiques mais en presence de gradients trans-Cpithéliaux de chiore et de bicar-
efflux from the lumen [10—141. In superficial proximal convolut-
bonate et de transport actif de sodium. L'augmenlation du debit de perfusion de 5 a 45 nl'min' a entrainé une augmentation du debit d'absot-ption de 0,49 a 3,37 nlmin' par millimetre de tubule.-'. La diminution de Ia concentration de chlore dans le perfusat collecté a
tive sodium translocation, are assumed to be through the
evolue de 5,9 a 2,6 mEq1iter' par millimetre de tubule-' en fonction du debit de perfusion. Ainsi Ia concentration tubulaire de chlore augmente
avec le debit de perfusion de idle sorte que les gradients trans-
ed tubules, it has also been suggested that the larger reflection coefficient of sodium bicarbonate compared with that of sodium chloride creates an additional osmotic driving force for net sodium reabsorption by solvent drag [4, 14]. These passive transport modes, that is, electrodiffusion of sodium and convec-
paracellular shunt across the proximal tubular epithelium [4, 141.
The present series of experiments addresses (I) the flow dependence of such ionic gradients along the proximal convo-
luted tubule, and ultimately related to this problem, (2) the dépendance vis a vis du debit se maintient, encore que d'une facon quantitative contributions of active and passive solute transréduite, en l'absence de transport actif de sodium (perfusion de port, respectively, to the flow dependence of fluid transport cyanure) mais en presence de gradients de chiore ci de bicarbonate. La across the proximal convoluted tubule. To this end, we have épithéliaux les plus élevCs sont maintenus aux debits les plus grands. La
dependance vis a vis du debit disparait quand sont a Ia fois absents Ic
Received for publication May 28, 1980 and in revised form January 23, 1981
carried out microperfusion of the rat nephron in vivo. In these experiments, dependence of reabsorption on rate of perfusion, as well as on artificially imposed anion gradients, was tested in the presence and in the absence of active transport. We can
show that a significant fraction of fluid reabsorption in the
0085-2538/81/0020-0580 $01.60
proximal tubule depends on the maintenance of transepithelial
© 1981 by the International Society of Nephrology
chloride and bicarbonate concentration gradients. We also 580
581
F/ow dependence of proximal fluid absorption Table 1. Composition of Ringer perfusates°
Chloride Chloride + cyanide Bicarbonate -F cyanide Chloride-phosphate Ringer Chloride-phosphate + cyanide Bicarbonate-phosphate Ringer
NaCI
NaHCO3
152 150 123 136 134
0 0 30 0 0 30
106
KCI
CaCl2
Na,HPO4
NaH2PO4
5
2 2 2
0
0
5
0.6
7.25
5
0.6
7.25
0.6
7.25
1.55 1.55 1.55
5
5
5
1
I
Urea
NaCN
5 5 5 5 5 5
0 2 2 0 0
a All solutions contained '4C-inulin (50 Ciml I) and 0.1% lissamine green. Values are in mmoles per liter
fluid given is an approximate concentration only. Solutions driven fluid absorption with increasing perfusion rate. Some of were prepared that were hypertonic and were finally diluted to these results are in agreement with those of Schafer et al match the osmolality of a plasma sample obtained from the tail obtained in vitro in the straight descending portion of the rabbit vein of the anesthetized, operated animal. This ensured that although there were differences in plasma osmolality of the experproximal tubule [151. imental animals (ranging from 285 to 303 mOsmkg water) the Methods osmolality of the tubular perfusate matched that of each rat. Series 1. To avoid the complicating presence of bicarbonate All experiments were performed on male Sprague-Dawley
observed in our experiments a marked increase of actively
rats weighing 150 to 200 g, anesthetised with 100 mgkg'
in the tubular lumen (at least in the initial fluid), we used a
i.p. mactin ([5-ethyl-5-( I '-methyl-propyl)1-2-thiobarbiturate) and placed on a thermostatically heated table to keep body tempera-
bicarbonate-free perfusate with a buffering capacity supplied by 2 mM phosphate. This solution was gassed with 100% oxygen, and the final pH was 6.8. Tubules were perfused at 5, 1:5, 30, and 45 nlmin± The length of the perfused segments (expressed
ture at 38° C. The operative procedure has been described previously [31 so it is only reported briefly here. Catheters were
inserted into the left jugular vein for infusions, and the right in mm) was, at5 nlmin, 1.99 0.38 mm; at 15 nlmin, 2.17 carotid artery to monitor systemic blood pressure; a tracheosto- 0.21 mm; at 30 n1min, 1.70 0.25 mm; at 45 nlmin. 1.65 my was performed. The left kidney was exposed via a flank 0.17 mm. Series 2. Inhibition of active transport processes with cyanide incision, placed in a plastic (Lucite) kidney cup and covered with liquid paraffin heated to 37° C; the left ureter was catheter- leaves only the passive components of transport. The experiized. The transit time of lissamine green was determined, and ments in series 1 were repeated using a perfusate in which 2 mtvi any animal whose blood pressure fell below 100 mm Hg, or cyanide was substituted for 2 m chloride. The solution was which had a proximal transit time of more than 13 sec, was gassed with 100% oxygen, and the final pH was 7.6. The discarded. Every animal received 0.99k saline, I ml, as a alkaline pH is due to the low buffer capacity of the perfusion priming dose during surgery followed by a continuous infusion fluid and the strongly alkalinizing effect of cyanide. Perfusions were performed at 5, 15, 30, and 45 nFmin. The length of the at 20 lmin. Randomly selected proximal convoluted tubules were per- perfused segments (expressed in mm) was, at 5 nlmin', 1 77 fused with a 1-lampel-pump according to the technique previ- 0.15 mm; at 15 nlmin ', 1.69 0.16 mm; at 30 nlmin, 1.65 ously described [3, 161 with the exception that in these experi- 20mm; at45 n1min, 1.55 0.11 mm. Series 3. This series of experiments was carried out to test the ments the rate of infusion was not constant but varied, being either 5, 15, 30, or 45 nlmin'. The majority of tubules were effects of the inhibition of active transport under conditions in perfused at two separate rates, and the sequence of perfusion which the role of a transepithelial chloride gradient is minirates was altered in a random fashion. After collection at the mized. Thus, to inhibit active transport processes, cyanide was first rate, at least 3 mm (and usually 5) were allowed for the substituted for 2 m chloride in bicarbonate Ringer so as to pump to equilibrate at the new speed. Fluid was collected from prevent the rise in luminal chloride along the proximal convothe perfused tubule over a timed interval with a distal oil block luted tubule and to provide no passive driving force. The solution was gassed with 5% carbon dioxide and 95% oxygen. in place. Two series of experiments (series 4 and 5) used simultaneous and the final pH was 7.7 to 7.8. Again, the disproportionately perfusion of both tubules and capillaries, as previously de- alkaline pH of this soluton is due to the presence of cyanide. In scribed [17]. It was technically very difficult in these experi- this series, perfusions were performed at 15 and 45 nlrnin ments to perfuse at more than one perfusion rate so paired data The length of the perfused segments (expressed in mm) was, at were not obtained. At the end of the experiments, perfused 15 nlmin, 1.90 0.23 mm; at 45 nlmin, 1.44 0.13 mm. Additional experiments were carried out in an attempt to tubules were filled with silicone rubber (Canton Biomedical Inc., Boulder, Colorado), the kidneys were excised, allowed to modify transepithelial anion gradients by changing the peril ubustand overnight in deionized water at 4° C, and subsequently lar fluid environment. Three series of experiments were carried digested in 10% sodium hydroxide. The filled tubules were then dissected out and their length measured using a camera lucida attachment on a Leitz TS steromicroscope. In all, five series of experiments were performed with differ-
out in which the buffering power of both tubular and peritubular perfusates was increased with phosphate. Series 4. The tubular perfusate was chloride-phosphate Ringer, which was gassed with 100% oxygen and had a pH of 7.4,
ent tubular perfusates (see Table I). The composition of the
whereas the peritubular perfusate was bicarbonate phosphate
582
Green ci a!
Ringer with 100 g1iter' of albumin, which was gassed with •—• Chloride Ringer
carbon dioxide in 95% oxygen and also had a final pH of 7.4 (see Table 1). This gave a "normal" chloride concentration gradient and was used as a control series. Perfusion was performed at IS
x— — —x Chloride Ringer + 2 m cyanide
and 45 nlmin. The length of the perfused segments (expressed in mm) was, at 15 nlmin', 1.65 0.22 mm; at 45 nlmin, 1.42 0.09 mm. Series 5. These experiments were designed to eliminate the chloride gradient across the tubule normally created by preferential bicarbonate reabsorption, bicarbonate being omitted from both perfusion solutions. The buffering in these perfusates was provided by phosphate, the tubular perfusate being chloride phosphate Ringer and the capillary perfusate chloride phosphate Ringer with 100 g-liter albumin. Both solutions were gassed with 100% oxygen and had a final pH of 7.4. Perfusion of the lumen
0
was carried out at 15, 30, and 45 nFmin. The length of the
.0
perfused segments (expressed in mm) was, at 15 nlmin, 1.31 0.2 mm; at 30 nlmin, 1.43 31 mm; at 45 nlmin, 1.40 0.15
a,
3.0 B
-
Bicarbonate Ringer + 2 mta cyanide
B B
2.0 a
0 0 0.
0
+
Co
T
1.0
T—
mm.
Series 6. This was the same as series 5 but cyanide was added to demonstrate the effects of inhibiting active transport. Perfu-
sion of the lumen was carried out at 30 nl'min'. The length of the perfused segments (expressed in mm) was 1.20 0.12 mm. Analyses. The volume of each collected sample was measured in a calibrated capillary, and aliquots were taken, togeth-
er with aliquots of the perfusate, for counting in a liquid scintillation spectrometer (Intertechnique SL 30, France). Aliquots of the remaining fluid were used to measure chloride by an electrometric titration technique first described by Ramsay, Brown, and Croghan [18]. The concentration of chloride was measured in aliquots of plasma, obtained at the end of the
15
30
45
Perfusion rate, nImin'
Fig. 1. Relationship between reabsorptive fluid flux and tubular perfu-
sion rate in three series of experiments. Data are expressed as mean values SEM. The lines drawn for each series of experiments are least square regression lines.
experiment by centrifugation of a small sample (200 p.1) of blood
from a tail vein, and in perfusion fluid. Thereafter net fluid absorption was calculated as
V = V (1 — In0/In) where V0 is the initial perfusion rate and In0 and In are the inulin concentrations in the perfusate and collected fluid, respectively. This was factored by the length of the perfused segment to give the rate per millimeter length of tubule. The total amount of inulin collected was calculated for each
relationship between the perfusion rate and the net fluid absorp-
tion was observed (Fig. I and Table 2). Whereas only 0.49 nlmin' of fluid was reabsorbed at a perfusion rate of 5 nlmin, 3.37 nl-min' was absorbed when the perfusion rate was increased to 45 nlmin.
The magnitude of the transepithelial chloride concentration difference is not precisely known, but making reasonable asperfusion, and if it was not 90 to 110% of the amount perfused, sumptions for plasma water and Donnan factor corrections, we the sample was discarded, Overall, the recollection of inulin can estimate a pertibular chloride concentration of some 114 was 99.8 0.2% and in no series was it significantly less than mM [19]. Thus, under the present experimental conditions, a 100%. About 10% of samples had to be discarded. maximal effective chloride concentration difference operative Results are presented as means SEM for each infusion. across the proximal tubular epithelium would be of the order of Preliminary analysis indicated that there was no significant some 45 mEq/liter. difference between the paired and the unpaired data, so all were Cyanide, in the presence of the imposed gradient, inhibits included, and the analysis performed as unpaired data. Compar- active transport processes and allows the expression of only isons were performed using Student's (test. passive forces. Yet there was still a significant linear correlation between perfusion rate and fluid reabsorption, which presumably reflects these passive forces and the way they are affected Results by altering the perfusion rate (Table 2, Fig. 1). It is instructive There were no significant differences in blood pressure, to note that this can account for only about 30% of the plasma chloride concentration, mean length of tubule perfused, reabsorption, however—a figure which agrees with previously and the percentage of inulin re-collected between any of the published results [1, 31. The slope of the regression line with series. cyanide present was significantly different from that when Chloride perfusate. With a fluid having a chloride concentra- cyanide was absent. Thus, there not only continues to be flowtion gradient from lumen to capillary of some 40 mmoles1iter, dependent fluid reabsorption but, importantly, the difference in there is a passive driving force (the chloride and bicarbonate fluid absorption in the presence of a transepithelial chloride gradients) to aid sodium and fluid reabsorption. A significant gradient, with and without transport inhibitor, increased mark-
583
Flow dependence of proximal fluid absorption
Table 2. Net fluid reabsorption' Net fluid reabsorptiona
nImin 'nim length t,ib,ele
At 5 nImin
a) Tubular Perfusion Tubular perfusion Series 1: Chloride Ringer Series 2: Chloride Series 3:
2.013
3.375
0.050
(15) 0.413 0.060
0.201 (16) 0.692 0.180
(Il)
(8)
(5) Bicarbonate
At45nlniin
1.423
(22)
0.201
At 30 nImin
0.042
0.492
+ cyanide
Ringer
At 15 nl'min
+ cyanide
0.145
Ringer
0.103
(9)
—
0.164
0.210 (15) 0.989 0.180 0.430
0.082
(13)
capillary perfusionb Chloride phosphate; bicarbonate
(14)
Tubular and Series 4:
1.070
phosphate + albumin
0.129
2.924
Series 5: Chloride phosphate; chloride phosphate + albumin Series 6: Chloride phosphate; chloride phosphate + cyanide + albumin
0.557
0.330
(14)
(14)
0.106
1.021
(12)
0.081
(6)
0.042
0.095
1.396
0.24
(12) —
(6)
a Numbers in parentheses represent number of perfusions. b Chloride
phosphate Ringer's solution was the tubular perfusate. The solution listed next was the capillary perfusate.
Table 3. Perfusion rate dependence of fluid transport across proximal convoluted rat tubulea
significant alteration in chloride concentration along the tubule.
Jv V
At 5
nlmin'
At 15 nImin
047 0.077
+
0.20
0.05
0.27
—
1.42 0.41 2.01
0.10 0.06 0.20
1.011
0.69 3.37 0.99
0.18
1.32
O.2l 0.26
2.38
—
+ At
45 nlmin'
which a chloride phosphate Ringer perfusion fluid was present in the tubules and bicarbonate-phosphate Ringer used to per-
—
+
At 30 nImin
Perfusion of both tubules and capillaries. The situation in
nImi,, 'mm
Cyanide
—
+
Measurements of the chloride concentration in the collected perfusate showed that at neither rate the perfusion was there a
a \ is perfusion flow rate; J, fluid transport; change in proximal tubular fluid transport. Minus and plus signs under Cyanide heading denote presence or absence of cyanide.
edly with perfusion rate. If it is accepted that the fraction of fluid reabsorption that is subject to cyanide inhibition represents active transport (see below), this observation clearly shows that fluid transport related to active transport increases with flow rate. This is also shown in Table 3, in which data on the perfusion rate dependence of fluid transport in the presence or absence of cyanide are compared. The increase in the active
transport component is apparent from the very dramatic increase of the rate of tubular reabsorption, AJ from 0.27 nI
min'mm' at a perfusion rate of 5 nlmin to a value of 2.38 nlmin'mm at a perfusion rate of 45 nlmin'. Such a flow dependent increase in active solute and fluid transport was also observed by Imai, Seldin, and Kokko [20] and by Schafer et al [15] in the isolated rabbit proximal tubule. That all fluid reabsorption in the absence of anion gradients is
indeed dependent on metabolic energy is indicated by the
fuse peritubular capillaries (series 4) is analogous to that in series 1 where a chloride gradient exists with bicarbonate present in the peritubular capillaries. It is noteworthy that the net fluid fluxes at perfusion rates of 15 and 45 nlmin are similar to those of the tubular perfusion experiments of series I (chloride Ringer in lumen, blood perfusion of peritubular capil-
laries), in spite of the double perfusion and other changes in composition of the perfusate (see Table 2), although in the double perfusion experiments slightly lower net reabsorptive fluxes are observed. When the chloride-phosphate Ringer in the peritubular capillaries and in the tubules had the same ionic composition (series 5), there was no significant chloride gradient that could drive passive reabsorption, and this is reflected by the lack of change in the chloride concentration of the collected tubular perfusate (see below). Figure 2 summarizes relevant data. Yet under these circumstances, without preferential bicarbonate reabsorption to produce transepithelial anion gradients, there was still a highly significant (P < 0.005) stimulation of fluid reabsorption at the higher perfusion rate, although such augmentation of net fluid transport is less than in the presence of anion gradients. Figure 2 summarizes these results and also includes. for comparison, data that were obtained when a significant chloride (and bicarbonate gradient) was imposed across the epithelium (chloride-phosphate Ringer in tubule, bicarbonatephosphate Ringer in capillary). It should be emphasized that in this last series of experiments, absence of bicarbonate in the peritubular fluid could by itself have an effect upon transepithelial fluid movement. The absence of bicarbonate from the peritubular fluid has been shown to reduce proximal tubular sodium and water transport
results with bicarbonate and cyanide perfusate. Although only two perfusion rates were used, there was no significant reabsorption of fluid with either perfusion rates (Table 2, Fig. I). Hence, under these experimental conditions, there is complete loss of the flow-dependency of transepithelial fluid transport. [1, 21]. This may account for the fact that the reduction of fluid
584
Green ci a!
reabsorption observed in series 5 (no chloride gradient, peritubular bicarbonate gradient present) is greater (48%) than that to be expected if 30% were due to passive reabsorption driven by the anion gradient. it is noteworthy that, despite the absence of both an anion gradient and of peritubular bicarbonate, there is still dependence of fluid reabsorption on perfusion rate, Again, the addition of cyanide to the tubular perfusion fluid completely abolished such transepithelial fluid transport. A summary of the changes in chloride concentrations in the different perfusion experiments is presented in Table 4. In the experiments using a bicarbonate-free, high chloride perfusion fluid (series I), there was a change in chloride concentration in collected fluid when compared with the original perfusate at all
the perfusion rates. This was most marked at the lowest perfusion rate and gradually diminished as the perfusion rate was increased. At the highest perfusion rate, the fall in chloride concentration was statistically significantly less than that at 5 nlmin. This reflects a time-dependent dissipation of the chloride gradient because the loss of chloride at lower perfusion rates results in a greater fall of the luminal chloride concentration'. The chloride concentration changes in the experiments using chloride Ringer plus cyanide are similar to those in which
•• Chloride phosphate Ringer in tubule; bicarbonate phosphate Ringer in capillary
X———* Chloride phosphate Ringer in tubule and capillary
3.0 -
A Chloride phosphate Ringer in tubule and capillary
+2 m cyanide
=
2.0 -
0 x0
I
0.
0
'-—I T
1.0 T .1.
bicarbonate-free chloride was used. Again, chloride loss from the
lumen results in a significantly larger drop in luminal chloride 15 30 45 concentration (P < 0.05) at a perfusion rate of 5 nlmin comPerfusion rate, nImin pared with 45 nlmin'. Inhibition of active transport processes Fig. 2. Relationship between reabsorptive fluid flux and tubular perfushould lead to cessation of hydrogen ion secretion and thereby sion rate in three series of experiments in which both tubules and prevent maintenance of the steady-state luminal bicarbonate perituhular capillaries were perfused. Data are expressed as in Fig. I. concentration. As a consequence, this would he expected to enhance the change in luminal chloride concentration. Our data show a slight trend in this direction but the difference of the perfused with plasma ultrafiltrate. Radtke et a! [25] showed that chloride concentration changes is not significant. That we have in rats a threefold increase in perfusion rate only increased fluid not observed a significant difference in our experiments could be reabsorption by 19% when the diameter was held constant, and due to the quite small differences in the absolute chloride concen- that increasing the diameter made no difference to the amount tration and that possibly not all of the chloride concentration reabsorbed. Similarly, Morel and Murayama [261 reported no changes are related to reciprocal alterations in luminal bicarbon- change of fluid reabsorption in a series of paired experiments ate. where rat proximal tubules were perfused at two different Discussion
Over the past few years, there has been considerable controversy about the relationship between the flow rate along renal
proximal tubules, the rate of reabsorption from the same tubules, and its relationship to glomerulotubular balance. Some
authors have described a linear relationship whereas others
perfusion rates; Morgan and Berliner [27] also observed no change in proximal tubular fluid absorption with an increase in perfusion rate.
in contrast to these studies there have been reports of a definite relationship between reabsorption and flow rate, and these are exemplified by the studies of Wiederholt et a1 1281, Bartoli et a! [29], Häberle et a! [30] and Senekjian et al [31]. Imai, Seldin, and Kokko, working with isolated rabbit tubules, showed a relationship only at flow rates less than 11 nlmin
have been unable to demonstrate one. In this respect, Burg and Orloff [24] reported no significant effect on fluid reabsorption of increasing the perfusion rate from [20]. Although some of these authors have suggested that such a 8 to 19 nl-min in isolated rabbit proximal convoluted tubules relationship contributes to glomerulotubular balance, they did not claim it was solely responsible for such a balance. Bartoli et a!, for instance, showed that doubling the perfusion rate only 'Although the perfusion fluids are initially bicarbonate-free, the flowincreased reabsorption by 60% [29]. dependent changes in chloride concentration signal entry of bicarbonate The current results in proximal convoluted tubules in rats and during the experiments. Because pH measurements were not carried out on the effluent perfusate, the extent of luminal acidification those by Schafer et a! 1151 on proximal straight tubules in achieved at the different flow rates was not assessed. Hence it is rabbits indicate that there is indeed, particularly under experiimpossible to determine whether the fall in chloride was completely mental conditions in which transepithelial anion gradients are matched by reciprocal bicarbonate concentration changes. Assuming maintained, a highly significant correlation between the rate of peritubular Pco2 values in the range of 40 to 65 mm Hg 122, 231, the perfusion and that of reabsorption. We have previously demonrange of bicarbonate concentrations resulting from transepithelial anion strated, in perfused proximal rat tubules in vivo that some 20% exchange is consistent with the maintenance of significant pH gradients across the proximal convoluted epithelium. to 30% of sodium reabsorption is passive and dependent on a
585
Flow dependence of proi,nal fluid ah,orptio,z Table 4. Changes in chloride concentration of perfusate
Chloride concentration changes, rn,nolesliter 'rn/n length tubule
At 5
Chloride Ringer Chloride Ringer + cyanide
nI-mm
At 15 nImin
—5.9
II
- 5.3
—6.7
1.1
—4.9
2.1 1.0
At 30 nI-mm
--3.9 --4.6
1.8 1.2
'
At 45 nI-mm —2.6
0.9
—3.5 + 1.0
transepithelial chloride and bicarbonate concentration gradient [1, 3]. The present results confirm that this is applicable at each perfusion rate, that is, the proportion of reabsorption dependent on such passive forces is constant. Thus, when active transport is blocked by cyanide and a chloride gradient is present, the
[4, 5, 7, 8, 11, 14, 33]. Thus, perfusion solutions in which chloride replaces bicarbonate, even though they have the same
amount reabsorbed is about 3O of that reabsorbed when
Jacobson [121, and Berry and Rector [101 have pointed out that
cyanide is absent (see Fig. 1). Recent experiments by Chantrelle and Rector, using deletion of potassium from the peritubular fluid to inhibit active sodium transport, have also shown that in the presence of similar anion gradients, only sonic 3OY
this type of osmotic disequilibrium between luminal and peritu-
of fluid reabsorption remains intact [321, that is, that two thirds are due to active transport.
This conclusion is further supported by data from experiments where both tubules and capillaries were perfused. When there was a chloride gradient between lumen and capillaries (series 4), there was increased reabsorption when compared with experiments (series 5) in which there was no gradient. Several factors are operating in this situation, however, and as well as the abolition of the chloride gradient, there is also the removal of the stimulating effect on fluid reabsorption of peritubular bicarbonate, which is considerable [1, 21]. A striking observation concerns the fact that active reabsorption of sodium is also dependent on the luminal flow rate. Two lines of evidence support this view. The first is based on the demonstration that, in the presence of anion gradients, the cyanide-inhibitable fraction of transepithelial fluid transport increases with luminal flow rate. Thus, when passive reabsorption (series 2) is subtracted from total reabsorption (series 1), there is still significant variation of reabsorption with flow rate.
The second line of evidence for flow-dependence of active
transport is given by experiments in which there was no chloride gradient (series 5). In the absence of a chloride gradient, without glucose or other organic substances in the perfusate and in the absence of bicarbonate, the stimulating effect of increasing flow rate upon reabsorption still exists. That this is an active, metabolically driven process is confirmed by its total abolition by cyanide.
The mechanisms responsible for the dependence of both active and passive fluid absorption on luminal flow rate need further consideration. With respect to the flow dependence of passive fluid transport, it is suggested that the tendency of increased perfusion rate to maintain the transepithelial chloride
and bicarbonate gradients along the nephron is important. Whereas at lower flow rates dissipation of the transepithelial anion gradients is favored because of passive chloride exit from and passive influx of bicarbonate into the tubular lumen, such events are less effective to modify anion gradients at the higher flow rates. It has been suggested that passive fluid absorption under these conditions is due to the relatively larger reflection
coefficient of bicarbonate compared to chloride and the high osmotic water permeability of the proximal convoluted tubule
osmolarity as the peritubular environment, are effectively hypo-
tonic with respect to the lower chloride peritubular fluid. Andreoli and Schafer et al [7, 8, 15, 33], Barratt et al [5], bular fluid may constitute an effective driving force for fluid
absorption. As these anion gradients are more effectively maintained at higher flow rates, passive fluid transport is stimulated pan passu. In addition, Schafer et a! [15] iave
suggested as a second mechanism to account for flow-dependent stimulation of fluid transport that osmotic equilibration is also minimized as tubular fluid delivery is enhanced. Considering the possible mechanisms responsible for the dependence of an active solute (and fluid) transport component on perfusion rate, several possibilities must be considered, but the situation is far from clear. As we have demonstrated in this study, a flow-dependent increase in the active component of solute (and fluid) transport occurred both in the presence as well as in the absence of transepithelial anion gradients. Stimulation of active solute (sodium) transport in the presence of high luminal chloride may be due to stimulation of sodium entry into proximal tubule cells across brushborder membranes by the favorable chloride gradient. As a consequence of this event, active sodium transport across the peritubular membrane would also be augmented. Suggestive evidence for coupled sodium chloride transport across the luminal cell membrane of proximal tubule cells, either by neutral sodium chloride cotransport [34],
or by a double exchange mechanism, one for sodium and hydrogen, and one for chloride and hydroxyl [11, 35] has recently become available. These mechanisms, in addition to other co- and counter-transport mechanisms involving sodium translocation across the luminal brushborder membrane of proximal tubule cells [II, 36—38], suggest an important role for anion concentration gradients in modulating sodium entry into proximal tubule cells. Because transepithelial chloride gradients are maintained more effectively at higher axial flow rates, it
is possible that this constitutes an effective mechanism of stimulating active sodium transport in the presence of flowdependent anion gradients. An additional aspect of the dependence of active solute (fluid) transport concerns this phenomenon in the absence of transepithelial anion gradients. Originally, Gertz et al [39] suggested that the relationship was with diameter of the tubule, but this is not now thought to be so [24, 25]. Other possibilities that were considered are some sort of mechanism akin to a flow reactor [39, 401, and the intriguing possibility that the layer of fluid trapped between microvilli could act as an unstirred layer, the depth of which is altered by the flow rate [41]. Although this
latter mechanism was applied to passive transport, similar
586
Green et a!
arguments could be applied for the passive entry of sodium into the cell prior to its active extrusion across the peritubular cell membrane.
It has recently been suggested that active solute (sodium) absorption across the proximal tubular epithelium results directly in the generation of small osmolarity differences between
References 1. GREEN R, GIEBISCH G: Ionic requirements of proximal tubular
sodium transport: I. Bicarbonate and chloride. Am J Physiol 229:1205—1215, 1975
2. GREEN R, GLEBI5cH G: Ionic requirements of proximal tubular sodium transport: II. Hydrogen ion. Am J Physiol 229:1216—1226, 1975
the luminal and peritubular solutions [42, 43]. Thus, fluid absorption across the proximal tubular epithelium may be driven, in addition to the above mentioned differences in effective osmotic pressures due to anion gradients, by the additional driving force of osmotic pressure differences be-
3. GREEN R, BISHOP JVH, GIEBISCH 0: Ionic requirements of proximal tubular sodium transport: III. Selective luminal anion substitution. Am J Physiol 236:F268—F277, 1979 4. NEUMANN KH, RECTOR FC JR: Mechanism of NaCI and water reabsorption in the proximal convoluted tubule of rat kidney. J C/in Invest 58:1110—1118, 1976
tween the lumen and the peritubular fluid [42, 43]. Experimental
5. BARRATT U, RECTOR FC, KOKKO JP, SELDIN DW: Factors involving the transepithelial potential difference across the proximal tubule of the rat kidney. J C/in invest 53:454—460, 1974
evidence in support of small but significant osmotic pressure differences across proximal convoluted rat tubules has recently been reported [16]. The dissipation of such an osmotic disequilibrium would depend, in principle, on two opposing events. On the one hand, active salt (sodium) absorption, in the presence of a given water permeability, generates and maintains osmotic disequilibrium. The flow-dependent increase in such active sodium transport would assure maintenance of luminal hypotonicity with increased flow along the tubule. On the other hand, passive osmotic equilibration would tend towards dissipation, in a flow-dependent manner, of such an osmotic disequilibrium. Given the flow dependence of active transport, osmotic disequi-
librium, that is, luminal hypotonicity and transepithelial fluid absorption could be more effectively maintained with increase
in axial volume delivery if the time course of establishing hypotonicity and its dissipation were different. Clearly the observed increase in active solute (sodium) transport with flow rate, although poorly understood as to its cellular mechanisms,
is an important factor to accentuate both the generation and maintenance of osmotic disequilibrium over a wide range of
6. RECTOR FC JR, MARTINEZ-MALDONADO, BRUNNER FP, SELDIN
DW: Evidence for passive reabsorption of NaCI in proximal tubule of rat kidney (abst). iC/in Invest 45:1060, 1966 7. SCHAFER JA, PATLAK CS, ANDREOLI TE: A component of fluid absorption linked to passive ion flows in the superficial pars recta. J Gen Physio/ 66:445—471, 1975 8. SCHAFER JA, PATLAK CS, ANDREOLI TE: Fluid absorption and active and passive ion flows in the rabbit superficial pars recta, Am J Physio/ 233:F154, 1977 9. MALNIC G, STEINMETZ PR: Transport process in urinary acidilication. Kidney mt 9:172—188, 1976 10. BERRY CA, RECTOR FC JR: Active and passive sodium transport in the proximal tubule. Mineral Electrolyte Metab 4:149—160, 1980 11. BERRY CA, WARNOCK DG, RECTOR FC: Ion selectivity and proximal salt reabsorption. Am J Physiol 235:F234—F245, 1978 12. JACOBSON HR: Characteristics of volume reabsorption in rabbit superficial and juxtamedullary proximal convoluted tubules: J C/in Invest 63:410—418, 1979 13. SEELY JF, CHIRIT0 E: Studies of the electrical potential difference in rat proximal tubule. Am J Physio/ 229:72—80, 1975 14. FROMTER E, RUMRICH G, ULLRICH KJ: Phenomenologic descrip-
tion of Nat, CU' and HC13 absorption from proximal tubules of the rat kidney. Pfluegers Arch 343:189—220, 1973
tubular flow rates. The relevance of these findings to glomerulotubular balance
15. SCHAFER JA, TROUTMAN SL, WATKINS ML, ANDREOLI TE: Flow
is difficult to assess under normal in vivo conditions. Even under conditions described here (series 1), a ninefold increase
16. BISHOP JHV, GREEN R, THOMAS S: Effect of glucose on water and
in perfusion rate only resulted in a sixfold increase in reabsorp-
tion rate. This figure of 65% glomerulotubular balance is comparable to that reported by Bartoli et al [29] and implies that other factors, such as peritubular protein concentration [17, 44],
or humoral factors [30], may be important. Both active and passive transport components may be involved in the flowdependent stimulation of proximal tubular sodium transport.
dependence of fluid transport in the isolated superficial pars recta. Kidney Int, this issue
sodium reabsorption in the proximal convoluted tubule of rat kidney. J Physiol (Lond) 275:481—493, 1978 17. GREEN R, WINDHAGER EE, GIEBIscH G: Protein oncotic pressure effects on proximal tubular fluid movement in the rat. Am J Physiol 226:265—276, 1974 18. RAMSAY JA, BROWN RHJ, CROGHAN PC: Electrometric titration of chloride in small volumes. J Exp Biol 32:822—829, 1955 19. ATHERTON J: Comparison of chloride concentration and osmolality
in proximal tubular fluid, peritubular capillary plasma and systemic
plasma in the rat. J Physio/ (Lond) 273:765—773, 1977 Besides the flow-dependent increase in the active component of solute (sodium) and fluid reabsorption, the flow dependence of 20. 1MM M, SELEMN DW, KOKKO JP: Effect of perfusion rate on the fluxes of water, sodium, chloride and urea across the proximal the passive component only applies once the chloride gradient convoluted tubule. Kidney mt 11:18—27, 1977
has been generated. Indeed, flow dependence of the active component, if it occurs also in the first part of the proximal
21. ULLRICH KJ, RADTKE HW, RUMRICH G: The role of bicarbonate
and other buffers on isotonic fluid absorption in the proximal convolution of the rat kidney. Pfluegers Arch 330:149—161, 1971 tubule, might be a means of ensuring that the chloride gradient was established at approximately the same distance along the 22. DUBOSE TD, PUCACCO LR, LUCCI MS, CARTER RW: Micropuncture determination of pH, pCO2 and total CO2 concentration in tubule at each flow rate. Alteration of this equilibrium point accessible structures of the rat renal cortex. J C/in Invest 64:476— could of itself alter glomerulotubular balance. 482, 1979 Acknowledgments Some of the results have been presented in abstract form in C/in Res 27:497A, 1979.
23. SOHTELL M: pCO2 of the proximal tubular fluid and the afferent arteriolar blood in the rat kidney. Acta Physiol Scand 105:137—145, 1979
24. BURG M, ORLOFF J: Control of fluid absorption in the renal proximal tubule. J C/in Invest 47:2016—2024, 1968
This research was supported by NIH grant AM 1743305 and by 25. RADTKE HW, RUMRICH 0, KLOSS 5, ULLRLCH KJ: Influence of
Medical Research Council Grant G973/2531B.
Reprint requests to Dr. G. Giebisch, Department of Physiology, Yale University School of Medicine, 333 Cedar Street, New Haven, Connecticut 06510, USA
luminal diameter and flow velocity on the isotonic fluid absorption and 36Cl permeability of the proximal convolution of the rat kidney. Pfluegers Arch 324:288—296, 1971 26. MOREL FF, MURAYAMA Y: Simultaneous measurement of unidi-
rectional and net sodium fluxes in microperfused rat proximal tubules. Pfluegers Arch 320:1—23, 1970
587
Flow dependence of proximal fluid absorption 27. MORGAN T, BERLINER RW: In vivo perfusion of proximal tubules of the rat: Glomerotubular balance. Am J Physiol 271:992—997, 1969 28. WIEDERHOLT M, HIERHOLZERK, WINDHAGER EE, GIEBISCH 0:
Microperfusion study of fluid reabsorption in proximal tubules of rat kidneys. Am J Physiol 213:809—818, 1967 29. BARTOLI E, CONGER JD, EARLEY LE: Effect of intraluminal flow on proximal tubular reabsorption. J C/in invest 52:843—849 30. HABERLE DA, SHIIGAI T: Flow-dependent volume reabsorption in the proximal convolution of the rat kidney: The role of glomerularborn tubular fluid for the maintenance of glomerulotubular balance,
in Newer Aspects of Renal Function, edited by VOGEL HG, ULLRICH KJ, Amsterdam-Oxford, Excerpta Medica, 1978, pp 198-206
31. SENEKJIAN JO, KNIGHT TF, SANSON SC, WEINMAN ED: Effect of
flow rate and the extracellular fluid volume on proximal urate and water absorption. Kidney mt 17:155—161, 1980 32. CHANTRELLE B, RECTOR FC JR: Active and passive components of
volume reabsorption in rat superficial proximal convoluted tubule (abst). C/in Res 28:44lA, 1980 33. ANDREOLI TE: Salt and water transport processes in the proximal straight tubule, in Renal Function, edited by GIEBISCU 0, PURCELL EF, New York. Josiah Macy Jr Foundation, 1978, pp 186—2 13
34. SPRING KR, KIMURA 0: Chloride reabsorption by renal proximal tubules of Necturus. J Me,nhr Biol 38:233—254, 1978 35. Lucci MS, WARNOCK DG: Effect of anion-transport inhibitors on NaCl reabsorption in the rat superficial proximal convoluted tubule. J C/in Invest 68:570—579, 1979
36. WINDUAGER EE, GIEBI5CH
0: Proximal sodium and fluid trans-
port. Kidney mt 9:121—133, 37.
1976 The renal handling of
BURG MB: sodium chloride, in The Kidney, edited by BRENNER BM, RECTOR FC Ja, Philadelphia, London, Toronto, Saunders, 1976, pp 272—3 17
38. WINDHAGER EE: Sodium chloride transport, in Membrane Tran-
0, TosTEsoN DC, Ussmc HH, Berlin-Heidelberg-New York, Springer-Verlag, 1978, vol 4A, pp
port in Biology, edited by GIEBIsCH 145—2 13
GERTZ KH, MANGOS JA, BRAUN 0, PAGEL HD: On glomerulartubular balance in the rat kidney. Pfluegers Arch 285:360—372, 1965 40. KELMAN RB: A theoretical note on exponential flow in the proxi39.
mal part of the mammalian nephron. Bull
Math Biophys 24:303,
1962
41. RICHARDSON 1W, LICKO V, BARTOLI E: The nature of passive flows through tightly folded membranes: The influence of nhicrostructure. J Membr Biol 11:293—308, 1973 42. ANDREOLI TE, SCHAFER JA: Effective luminal hypotonicity: the driving force for isotonic proximal tubular fluid absorption. 4m J Physiol 233:F89—F96, 1979
43. ANDREOLI TE, SCHAFER JA: Volume absorption in the pars recta
III. Luminal hypotonicity as a driving force for isotonic volume absorption. Am J Physiol F349—F355, 1978
44. WINDI-IAGER EE, LEWY JE, SPITZER A: Intrarenal control of proximal tubular reabsorption of sodium and water. Nephron 6:247—259, 1969