Drugs 41 (Suppl. 3): 1-13, 1991 0012-6667/91/0300-0001/$6.50/0 © Adis International Limited All rights reserved. DRSUP16788
How Do Loop Diuretics Act? M. Wittner,l A. Di Stefano,l P. Wangemann 2 and R. Greger3 Institut de Recherche Fondamentale, Service de Biologie Cellulaire, Commisariat a l'Energie Atomique, Centre d'Etudes Nucleaires de Saclay, Gif-sur-Yvette, France 2 Boys' Town National Institute for Communication Disorders in Children, Omaha, Nebraska, USA 3 Physiologisches Institut, Universitat Freiburg, Freiburg, Federal Republic of Germany
Summary
In the thick ascending limb of the loop of Henle, NaCI reabsorption is mediated by a Na+j2Cl-jK+ cotransport system, present in the luminal membrane of this nephron segment. Loop diuretics such as furosemide (frusemide), piretanide, bumetanide and torasemide bind reversibly to this carrier protein, thus reducing or abolishing NaCl reabsorption. This leads to a decrease in interstitial hypertonicity and thus to a reduced water reabsorption. In nephron segments other than the thick ascending limb, loop diuretics have no quantitative importance with respect to their saluretic and diuretic activities. Loop diuretics also reduce Ca++ and Mg++ reabsorption in the thick ascending limb in a way which is still not clear. Furthermore, these drugs increase the urinary K+ excretion by enhancing distal tubular K+ secretion and reducing K+ reabsorption in the loop of Henle. Finally, by reduction of active NaCl transport, loop diuretics drastically reduce the substrate requirement and oxygkn dependence of the thick ascending limb cells. This renders these cells, which are chara.cterised by high transport rates and only limited substrate reserves, less vulnerable in acute renal failure.
Of the various classes of diuretics, the most powerful act in the thick ascending limb of the loop of Henle (Burg & Green 1973; Burg et al. 1973; Greger 1981a; Imai 1977; Schlatter et al. 1983). This review focuses on the cellular mechanisms of action of loop diuretics in the thick ascending limb of the loop of Henle. More specifically, the following issues will be discussed: the site of action of diuretics in the nephron; the protection against ischaemia and/or reduction in metabolic substrates of the thick ascending limb by loop diuretics; the structure-activity relationships for loop diuretics.
1. NaCI Reabsorption in the Nephron and Site of Action of Diuretics Figure 1 shows where NaCI and water are reabsorbed in the nephron. The numbers indicate the fractional delivery of NaCl and water to the different nephron segments as percentages of the filtered load. 60 to 70% of the filtered NaCI and water are reabsorbed iso-psmotically in the proximal tubule: In the loop of Henle, some 30% of NaCI and 10% of water are Feabsorbed. The reabsorption of NaCI occurs mainly in the thick ascending limb of
2
Fig. 1. Tubular reabsorption of NaCl and water. The numbers indicate the fractional delivery of NaCl and water as percentages of the filtered load.
the loop of Henle. This nephron segment dilutes the lumen fluid and generates a hypertonic interstitium, because it is able to reabsorb NaCI and simultaneously prevent a corresponding water reabsorption by its water impermeability. In antidiuresis, this process is responsible for the concentration of urine (withdrawal of water from the collecting ducts into the hypertonic interstitium). In water diuresis, the collecting ducts are water impermeable, and the diluted fluid, produced by the thick ascending limb, is excreted in the urine. Along the distal convoluted tubule, a further 8% of the filtered NaCI and 10 to 20% of the filtered water are reabsorbed. In the collecting duct, reabsorption of Na+ and Cl- continues, so that less than I% of the filtered load appears in the urine. Water reabsorption depends on the water permeability of the luminal cell membrane of the collecting duct and is regulated by antidiuretic hormone. The rate at which s~dium is reabsorbed is also hormonally regulated by! mineralocorticoids (aldosterone) and in some species by the antidiuretic hormone as well (Schla~ter & Schaefer 1987). NaCI reabsorption can be blocked by diu'retics at different sites in the nephron. In the proximal tubule, NaCI reabsorption can be inhibited by car-
Drugs 41 (Suppl. 3) 1991
bonic anhydrase inhibitors (e.g. ethoxzolamide, acetazolamide). The induced saluresis and diuresis, however, are not very pronounced, since some bicarbonate reabsorption occurs independently of the presence of carbonic anhydrase. Furthermore, the diuresis induced by carbonic anhydrase inhibitors is mostly compensated for by an increase in NaCI reabsorption in more distal nephron segments. In the thick ascending limb of the loop of Henle, NaCl reabsorption can be blocked effectively by the socalled loop diuretics [e.g. furosemide (frusemide), bumetanide, piretanide, torasemide]. In the early distal tubule, thiazide diuretics are effective. They seem to interfere with an electro-neutral NaCI cotransport system, present in the luminal cell membrane. Since the transport rate of this nephron (see above) is quite low, the diuretic effect of thiazides is small compared with that of loop diuretics. Diuretics, such as amiloride and triamterene, act in the collecting tubule by blocking luminal Na+ channels. The diuretic effect of these substances is also low, since only a small percentage of the filtered NaCI is reabsorbed in this nephron segment. These substances are, however, of importance since they are K+ sparing. Inhibition of luminal Na+ channels by amiloride or triamterene leads to an inhibition of K+ secretion. This K+ sparing effect may be of advantage if K+ secretion were increased by a saluresis, induced more proximal to the collecting duct. 1.1 Effects of Loop Diuretics of the Furosemide Type in Nephron Segments Other Than the Thick Ascending Limb
Figure 2 shows the cellular models for NaCl reabsorption in the proximal tubule, in the thick ascending limb and in the collecting duct. In all 3 segments, NaCl reabsorption is energised by the (Na+ /K+) -ATPase, localised in the basolateral membrane. In the thick ascending limb, NaCI reabsorption is mediated by a Na+ /2Cl-/K+ cotransport system present in the apical cell membrane. The (Na+/K+)-ATPase reduces the cytosolic Na+ concentration and thus provides the driving force for the operation of the Na+/2Cl-/K+
3
How Do Loop Diuretics Act?
carrier. The K+ taken up by the carrier recycles via a K+ conductance present in the luminal membrane. Cl- leaves the cell via Cl- channels and a KCI cotransport system present in the basolateral membrane (Greger 1985). The positive voltage of the lumen, due to the polarisation of the cell, drives paracellular Na+ reabsorption. The Na+/2Cl-/K+ cotransport system can be reversibly blocked by loop diuretics of the furosemide type, which bind to one of the Cl- binding sites of the carrier (Greger & Schlatter 1983). In the proximal tubule, 30% of the filtered Na+ is reabsorbed together with bicarbonate through the cell (transcellular), whereas 70% of the deliveredNa+ and 100% of Cl- are reabsorbed by means of the paracellular shunt pathway. Transcellular and paracellular reabsorption of the Na+, together with the reabsorption of Cl- and water, represent 60 to 70% of the filtered load (Fromter et al. 1973). Effects of furosemide in the proximal tubule are of minor importance and relate to the fact that loop diuretics act as rather weak inhibitors of carbonic anhydrase. The mechanisms by which NaCl is reabsorbed in the early distal tubule are presently being studied. NaCI reabsorption in this nephron segment is mediated by a carrier system, which is sensitive to thiazide
CELL
1.2 The Consequences of Inhibition of NaCI Reabsorption by Loop Diuretics Firstly, NaCl net transport in the thick ascending limb, comprising 20 to 30% of the filtered load, is considerable. Inhibition of this net transport by loop diuretics abolishes the hypertonicity of the interstitium, and so inhibits the water reabsorption in the collecting duct. The inhibition of NaCI reabsorption in the loop of Henle cannot be compensated for by more distal nephron segments, thus a pronounced saluresis and diuresis is induced. Secondly, loop diuretics produce a kaliuresis which has the following 2 causes: the small net reabsorption Cortical collecting tubule
Thick ascending limb
Proximal tubule
LUMEN
diuretics but insensitive to furosemide (Greger & Velasquez 1987). In the collecting duct, Na+ reabsorption and K+ secretion occur via luminal Na+ and K+ channels (Palmer & Frindt 1986), as there is no furosemide-sensitive transport system. It has, however, been claimed that furosemide acts in the papillary collecting duct and on the epithelium lining of the renal papilla (Sands et al. 1986; Sonnenberg 1978). It is very likely that none of these effects is of quantitative importance, with respect to the saluretic and diuretic activities of these drugs.
BLOOD
LUMEN
CELL
BLOOD
G Na+ Gluc
LUMEN
BLOOD
CELL
G Na+ K
Na+
Fig. 2. Mechanisms of NaCl reabsorption in the proximal tubule, the thick ascending limb of the loop of Henle and the collecting tubule. - = ionic channels; . - = primary active pump; 0- = carrier system.
Drugs 41 (Suppl. 3) 1991
4
cells, is abolished; Na+ reabsorption and K+ secretion in the collecting duct depend on the luminal NaCl concentration and the luminal flow rate. Both are increased after application of loop diuretics, which are thus responsible for increases in Na+ reabsorption and K+ secretion.
;:- 200 I
E
E 150 I
C
"::> E •
;: 0 0;
z
E Q.
100
50
o PRE FUR POST
PRE FUR
POST
;:- 1.00 I
E
~ 0.75
I
C
x ::>
'E 0.50 •
;:0
- E ~ S 0.25
o (6)
PRE FUR
POST
PRE FUR POST
+3 ;:I
E E
" ~ 0;
c
0
'j"
c
E
(5
+ E :.:: S
-3
(7)
PRE
FUR
POST
Fig. 3. Effect of furosemide (FUR) [10- 4 mOI/L, lumen) on transepithelial ion net fluxes ofNa+, CI-, K+, Ca++ and Mg++ (pmol/minjmm), measured by electron probe analysis in isolated perfused cortical thick ascending limbs of the mouse nephron. PRE = before furosemide; POST = after furosemide. Furosemide reversibly inhibits Na+, CI-, K+, Ca++ and Mg++ reabsorptions. Number of tubules in brackets. Mean values ± SEM. * = p < 0.05 (data from Wittner et al. 1989).
of K+ in the thick ascending limb is abolished by loop diuretics, since the Na+ /2CI-/K+ carrier is blocked, and the lumen-positive transepithelial voltage maintaining the passive reabsorption ofK+ across the cell, and more importantIybetween the
2. Cellular Mechanisms of Action of Loop Diuretics Figure 3 shows the effects of furosemide on the transepithelial ion net fluxes ofNa+, CI-, K+, Ca++ and Mg++, measured in isolated perfused cortical thick ascending limbs of the mouse nephron by electron probe analysis (Wittner et al. 1989). Furosemide abolished not only the transepithelial Na+ and CI- net fluxes, but also inhibited Ca++ and Mg++ reabsorptions and the small net reabsorption of K+. The inhibitory effect offurosemide on electrolyte transport was completely reversible after removal of the drug from the lumen perfusate. Recent observations from our laboratory indicate that Ca++ and Mg++ are reabsorbed by paracellular and transcellular pathways. The paracellular component is driven by the transepithelial voltage (Wittner et al. 1989). Inhibition of Ca++ and Mg++ transport by furosemide could therefore be explained by the fact that furosemide, by abolishing the transepithelial voltage, abolishes the driving force for Ca++ and Mg++ reabsorption. Figure 4 shows the effects of furosemide on the transepithelial potential difference (PO te ) and the voltage across the basolateral membrane (PObl) of an isolated, perfused, cortical thick ascending limb segment. Furosemide added to the lumen perfusate at a concentration of 5 X 10- 5 molfL leads, with a time constant of less than half a second, to a reduction of PO te . Simultaneously, PObl hyperpolarises, and the transepithelial resistance and fractional resistance of the basolateral cell membrane increase (increase in the size of the voltage transients across the entire epithelium and the basolateral cell membrane). The magnitude of the voltage transients, obtained by the periodic injection of current pulses into the tubular lumen, is proportional to the respective resistances. The in-
5
How Do Loop Diuretics Act?
,.
-
>
.
§.
r5
0..
-"~
~
('
-5
(
,
0
+5 - 20
I"
./' • 1£
--
lS C 0..
I
- ...y.
I -
> §.
-
-
f- ~
I
r
I
-40 -60
I
("
-
, -f -
"\. f""- ~
~
-80
':
\
-100
n ~ r,-.\; 1-0.. .,
o
5
10
Time (sec)
Fig. 4. Effect offurosemide (5 X 10- 5 mol/L. lumen) on transepithelial potential difference (PDtcl and on the potential difference across the basolateral membrane (PDbl). At 5-second intervals. short current pulses are injected into the lumen to obtain a measurement of the transepithelial resistance and the fractional resistance across the basolateral membrane. The collapse of PD tc . the hyperpolarisation of PD tc . the increase in the transepithelial resistance (increase in height of the transepithelial voltage transient) and the increase in the fractional resistance across the basolateral membrane are observed (data from Greger & Schlatter 1983).
crease in the basolateral membrane resistance after application of furosemide is due to a decrease in cytosolic CI- activity. This decrease in activity has the following 2 causes: the Na+/2CI-/K+ carrier is blocked, and therefore no more CI- enters the cell; CI- leaves the cell via CI- channels and the KCI cotransport system of the basolateral membrane until the cytosolic chloride concentration is in Nernst equilibrium. For a membrane voltage of -90mV, this would correspond to a cytosolic CIactivity of 5 mmol/L. Very similar values have been measured with CI- selective electrodes (Greger et al. 1983). Table I shows that there was a decrease in cytosolic CI- activity from 19 to 7 mmol/L in the presence of furosemide, which suggests a passive distribution of CI- across the basolateral membrane in the presence of this loop diuretic. The final status after application of furosemide is shown by the model in figure 5. Furosemide 'locks' the Na+/2C\-/K+ carrier. Thus, the entry of Na+, CIand K+ into the cell ceases instantaneously. The cell CI-, which is above equilibrium concentrations under control conditions, declines towards passive equilibrium, because the entry step for this ion is blocked, and outflux across the basolateral membrane via Cl- channels and the KCl cotransport system continues until CI- reaches this equilibrium (Greger et al. 1983). Cell Na+ also declines. In the Amphiuma diluting segment, a decrease from 9 to 4 mmoljL has been reported (Oberleithner et al. 1983). This decrease in Na+ concentration is
Table I. Estimates of intracellular CI- activity in isolated perfused cortical thick ascending limb of the rabbit nephron. In the presence of furosemide, CI- is passively distributed (data from Oberleithner et al. 1983) PDbl (mV)
PDcl(mV)
ECI(mV)
ACI(mmol/L)
Pretreatment control
-74
- 27
+ 47
-'> 19
Furosemide a
- 88
-16
+ 72
-'> 7
Post-treatment control
-74
- 26
+ 48
-'>18
a
3 X 10- 5 mol/L added to lumen perfusate, mean of 29 observations.
Abbreviations and symbols: PDbl = electrical potential difference across the basolateral membrane; PDclby CI- selective microelectrodes; ECI- = chemical potential for CI-; ACI- = intracellular CI- activity.
= potential difference read
Drugs 41 (Suppl. 3) 1991
6
lumen
I
Cell
+90mV
I
3. Protection of the Thick Ascending Limb Against Ischaemia and/or Reduction in Metabolic Substrates by Loop Diuretics
Furosemide
. . - - - - - - Na· -
I
Perilubule space
-90mV
J
-----+
Fig. 5. The thick ascending limb cell in the presence of furosemide. Net transport is reduced to zero, and the cell potential approaches the K+ equilibrium potential. K+ and Cl- are passively distributed across both cell membranes. The (Na+/K+) -ATPase activity ceases as a result of blocked Na+ uptake.
due to the pumping of Na+ out of the cell by the (Na+ /K+) -ATPase until a minimum concentration of Na+ is reached. Then the (Na+/K+) -ATPase comes to a standstill and no longer consumes adenosine triphosphate (ATP). Cell K+ does not change after furosemide application (Greger et al. 1984a). After inhibition of the Na+/2Cl-/K+ carrier, K+ leaks out of the cell across the luminal membrane via its conductive pathway and along its electrochemical gradient. On the other hand, K+ is pumped into the cell as long as the (Na+/K+)-ATPase is operating. These opposing effects appear to balance each other, so that the intracellular K+ activity remains fairly constant. The luminal and basolateral cell membranes hyperpolarise to the K+ equilibrium potential. Cland K+ are passively distributed across the 2 cell membranes. The ion net fluxes across the cell membranes and across the shunt pathway therefore become zero. In addition, the transports of Ca++ and Mg++ are inhibited.
The cells of the thick ascending limb are particularly rich in mitochondria, and the density of the (Na+/K+) -ATPase, localised in the basolateral membrane, is high (Garg et al. 1981). NaCi reabsorption in the cortical and medullary parts of this nephron segment are energised by aerobic metabolism. In the event of oxygen shortage, the amount of ATP produced solely by glycolysis is not sufficient to maintain active NaCI transport (Wittner et al. 1984). Substrates are taken up from the peritubular side via simple (butyrate, acetate) or facilitated (sugars and small anions such as lactate or pyruvate) diffusion. From the lumen side, only butyrate can enter the cell via non-ionic diffusion (Wittner et al. 1984). Figure 6 shows the close link between active NaCI transport in the thick ascending limb and substrate availability. At time zero, all substrates were removed from both sides of the epithelium. The equivalent short circuit current, which corresponds stoichiometrically to the rate of active NaCI reabsorption (Greger 198Ia), collapsed within 10 minutes. This inhibition was reversible if substrate removal was only transient (10 min300
if 200 E ()
§.
ID·Glucose
I +
25
~ -100 ,
a
I
Iii
5
Time (mm)
Fig. 8. Effect of substrate removal on the basolateral membrane potential (PObl) in an isolated perfused cortical thick ascending limb segment of the rabbit nephron. The substrate (5 mmol/L O-glucose) was removed and furosemide (5 x lQ-5 moljL) was simultaneously added to the lumen perfusate. This caused a hyperpolarisation of PObl. After removal of furosemide, PObl depolarised. Readmission of glucose led to a hyperpolarisation of PObl back to control values. Substrate removal is devoid of effect on the cell as long as furosemide is present (data from Greger et al. 1984c).
Drugs 41 (Suppl. 3) 1991
8
PDbl depolarised with a delay of only 3 minutes. Readmission of glucose led to a hyperpolarisation of PDbl back to the control value (Greger et al. 1984c). This shows that loop diuretics of the furosemide type, by abolishing active NaCI transport, cause the cells to enter a resting state, which renders them almost independent of metabolic fuels. Figure 9 shows the protective character of furosemide against inhibition of the (Na+/K+) ATPase by ouabain or phloretin. It is evident that neither ouabain nor phloretin has any effect on the cell potential, as long as furosemide is present. Both substances lead to a rapid depolarisation of the cell after removal of furosemide (Greger & Schlatter 1983). It seems, therefore, that the effects of substrate removal and/or direct inhibition of the (Na+/ K+) -ATPase can be prevented by blocking the Na+ / 2Cl-/K+ cotransporter with furosemide. These data emphasise how loop diuretics such as furosemide can be used to protect the thick ascending limb against the hazards of substrate shortage and hypoxia.
4. Torasemide, a New Loop Diuretic Figure 10 shows the effect of a new loop diuretic, torasemide, on NaCI reabsorption in the thick ascending limb. The structural formula of to-
rasemide (fig. II), which is a powerful diuretic (Lesne et al. 1985), suggests that it may interfere with the Na+/2CI+/K+ cotransporter. Like furosemide and related compounds, torasemide contains an anionic group and a secondary amine. The nitrogen of the pyridine, in meta position to the anionic group, should correspond to the sulphonamide group offurosemide and related compounds (Schlatter et al. 1983). As evident from the doseresponse curves, obtained when torasemide was added either to the lumen or to the bath perfusate (fig. 10), this substance acts mainly from the luminal side by interfering with the Na+/2Cl-/K+ carrier. Half maximal inhibition of active NaCI reabsorption, measured as the equivalent short circuit current (lsd, was achieved at 3 X 10- 7 mol/L (Wittner et al. 1986). Figure II shows that this value is comparable to that found for powerful diuretics such as furosemide derivatives and bumetanide (Schlatter et al. 1983). The torasemide-mediated inhibition of the equivalent short circuit current, which is rapidly reversible, is due to a fall in the transepithelial lumen positive voltage and a moderate increase in the transepithelial resistance. These effects and the hyperpolarisation of the basolateral membrane potential in the presence of luminal torasemide (Wittner et al. 1986) are qualitatively and Ouabain Furosemide
o - 20
;;-
- 40
S - 60 21
o
0..
- 80
I
o
I
0.5
I
3.0
Time (min)
Fog. 9. Inhibition of the (Na+jK+) -ATPase in the presence and absence of furosemide (5 X 10- 5 moljL, lumen) ina cortical
thick ascending limb cell. Furosemide leads to a hyperpolarisation of the potential difference across the basolateral membrane (PDbl) during which phloretin (5 X 10- 6 moljL, bath) has no effect on PDbl. As soon as furosemide is removed, phloretin depolarises PDbl. Upon removal of phloretin, the potential difference of the cell recovers, and furosemide again leads to a small hyperpolarisation. Ouabain, like phloretin, is then without effect on PDbl. Only after removal of furosemide can ouabain lead to its typical depolarisation (data from Greger & Schlatter 1983).
How Do Loop Diuretics
9
Act~
300 (10)
crE
200
"
100
.J!'
o
\ (6\
~~
(11)
u
(4)
\,---1 '\~3) I",' "
(5)
(5)
'I__ -- __(2) ,
~.~I----,,------,,------,,----~,
10- 7
10- 6
10-5
10-4
Torasemide (mol' L- ')
Fig. 10. Dose response curves for torasemide added to the luminal (X---X) and to the peritubular perfusates (0-0 ). The equivalent short circuit current (I se) is plotted versus the torasemide concentration. A half maximal effect of torasemide is achieved at 3 x 10- 7 moljL in the luminal perfusate and at 3 x lQ-5 mol/L in the peritubular perfusate. Numbers in brackets refer to the number of tubules tested. Mean values ± SEM (data from Wittner et al. 1986).
even quantitatively similar to those obtained for other diuretics of the furosemide type (Schlatter et al. 1983). Like these loop diuretics, torasemide may interfere by means of its negatively charged sulphonyl urea moiety with one of the Cl- binding sites of the N a +/2CI-/K + transport protein. The blocked uptakes of Na+, K+ and Cl- lead to a fall in cytosolic Cl- activity. This leads to a hyperpolarisation of the cell towards the K+ equilibrium potential (Schlatter et al. 1983) and to a collapse of the transcellular current flow. Torasemide, unlike furosemide, acts also from the peri tubular side (fig. 10), probably by interfering with the Cl- channels of the basolateral membrane. When compared with the effects exerted from the luminal side, far higher doses are required to inhibit the equivalent short circuit current from the peritubular side. A concentration of 3 X 10- 5 mol/L is needed to inhibit the equivalent short circuit current by 50% (Wittner et al. 1986). The discovery that the effects of luminal and peritubular torasemide are qualitatively similar and that torasemide, unlike furosemide, is far more lipophilic (Wittner et al. 1986), suggests that the peri tubular effects of torasemide are caused by a diffusion
across the cell and an interference with the luminal Na+ /2CI -/K+ cotransporter. Nevertheless, the following findings seem to disprove this hypothesis: torasemide acts equally rapidly from both sides of the cell, whereas furosemide and related compounds at very high concentrations (see below) act quickly from the luminal side, but only slowly from the peri tubular side; with furosemide, peritubular effects are only seen if the peritubular concentration exceeds 10- 3 mol/L. These effects are only slowly reversible; as shown by the patch-clamp recording given in figure 12, torasemide, unlike furosemide but like other Cl- channel blockers (Oi Stefano et al. 1985; Wangemann et al. 1986), inhibits Cl- channels in excised membrane patches of the luminal membrane of the shark rectal gland (Greger & Gbgelein 1985). These Cl- channels are similar if not identical to those present in the basolateral cell membrane of the thick ascending limb (Oi Stefano et al. 1985; Greger & Gbgelein 1985). The interference of torasemide with CI- channels leads to an inhibition of the conductive exit of this ion from the cell and thus reduces the transcellular current and the rate of Cl- reabsorption.
5. Structure-Activity Relationship Figure II shows several well-known loop diuretics and their respective IC50 values. These values correspond to the luminal concentration needed to inhibit 50% of NaCI reabsorption. In 2 previous studies (Schlatter et al. 1983; Wittner et al. 1987) the structure-activity relationships of analogues of furosemide and torasemide have been examined. The following are requirements for reversible interaction with the Na+/2Cl-/K+ carrier: an anionic group such as a sulphonate, carboxylate, tetrazolate or sulphonyl urea is mandatory; a secondary (e.g. furosemide, torasemide) or a tertiary (e.g. piretanide) amine is needed in ortho position or meta position to the anionic group; this amino group must link the anionic moiety to an apolar residue; in meta position to the amino group a sulphonamide group (e.g. furosemide, piretanide) or a pyridino ring (e.g. torasemide) is required. As was previously shown (Greger, in press; Gre-
Drugs 41 (Suppl. 3) 1991
10
BTS 39542 Xi pam ide
Torasemide
L-Ozolinone
Indacrinone CH3 0
CI
@C@r o :
:
CI
:
O-CH2-COO: Muzolimine .
Furosemide
I 10- 8
,, , I 10- 7
t
t
I 10-- 6
,
C'if~-\. CI J.8.I N..( . . NH2
, I 10- 5
CH3 0
I 10-4
,
:.... ..... .. :.......~ »
10-4
I
10- 3
IC50 (mmol· L- 1) Fig. 11. Loop diuretics and their· respective 1C50 values. The IC50 value is the luminal concentration at which the substance
inhibits NaCI reabsorption half maximally. Data were obtained on isolated perfused cortical thick ascending limbs of the rabbit nephron (data from Schlatter et al. 1983; Wittner et al. 1987).
ger et al. 1987a,b), loop diuretics have structural similarities to Cl- channel blockers and probably also to blockers of the Cl- /HC03- exchanger of the red blood cell membrane (Cousin & Motais 1982). Figure 13 shows some Cl- channel blockers and the concentrations needed to inhibit NaCI reabsorption half maximally. One of the most effective Cl- channel blockers is 5-nitro-2-(3-phenylpropylamino)-benzoate, which inhibits the equivalent short circuit current half maximally at 8 x 10- 8 mol/L (Wangemann et al. 1986). Torasemide, with its chemical structure placing it between loop diuretics and Cl- channel blockers, has an affinity for Cl-channels similar to that of diphenylamine2-carboxylate, the first compound shown to specifically inhibit Cl- channels in the thick ascending limb (Di Stefano et al. 1985). Triflocine and niflu-
mate, which are inefficient on the Cl- channel and on the Na+/2CI-/K+ carrier, are known to inhibit the band 3 protein in red blood cells (Cousin & Motais 1982). 5.1 The Common Features of, and the Differences between, Loop Diuretics and Cl- Channel Blockers All these compounds, most of which are carboxylates, are weak acids with pK-values ,., 4 to 6. All compounds carry an amino nitrogen at a certain distance from the carboxylate group. The ortho-positioning and meta-positioning between the 2 groups leads to active compounds, whereas greater distances lead to inactive ones. Chloride channel blockers mayor may not have a substituent on the
II
How Do Loop Diuretics Act?
right hand side ring (figs 11, 13); a chloride substitution in para position or a nitro substitution in meta position to the carboxylate group. This moiety carries a partial negative charge. The blockers of the Na+/2Cl-/K+ carrier possess an electronegative sulphonamide group in the same position. Finally, both classes of substances possess an apolar moiety (phenyl moiety, furfuryl moiety, etc.). For blockers of the Na+/2Cl-/K+ carrier, the restrictions on this moiety are not very marked, for example, in the case of furosemide, a Cl-substitution can replace the apolar moiety. For chloride channel blockers, the apolar residue is subject to very narrow spatial constraints; it must be a ring structure, and this ring must be separated by a well-defined distance from the amino group. In general, the blockers of chloride channels are more hydrophobic than the blockers of the N a +/2Cl-/K + carrier. Finally, loop diuretics can be 'converted' into Cl- channel blockers and vice versa by small modifications of the molecule (Greger et al. 1987b). For
example, removal of chloride in the furosemide molecule reduces the inhibitory effect on the Na+/ 2Cl-/K+ carrier. Further replacement of the sulphonamide group by an amino group abolishes any inhibitory activity. Replacement now of the amino group by a nitro group leads to an inhibitor of Clchannels, with no inhibitory effect on the Na+ /2Cl-/ K+ carrier. Comparable observations have been made in an extensive study of torasemide analogues; replacement of the pyridino nitrogen by a N02-Substituted carbon atom abolished the ability to block the Na+/2Cl-/K+ carrier, but did not affect the inhibitory potency for Cl- channels (Wittner et al. 1987). The structural similarities between loop diuretics and Cl- channel blockers and the fact that minor structural modifications convert a loop diuretic into a Cl- channel blocker and vice versa suggest that the different membrane transport proteins share in common some portion of the molecule.
NaCI/NaCI, -24mV Torasemide 1 mmol· L-l, 0-8 sec
9-16 sec
17-24 sec
Control. 0-8 sec
Control, 9-16 sec
2pA
L 800 msec
Fig. 12. Patch-clamp recording of an excised luminal chloride channel of the rectal gland of the shark. Torasemide
(10- 4 mol/L) reversibly blocks chloride channel activity (data from Greger & G6gelein 1985).
Drugs 41 (Suppl. 3) 1991
12
Flufenamate
•
:
H(D)
Torasemide
~Nr@
N02~COO-
CH3)j-j(
~-.lQ\
.~c:.!l- ~ : s coo133B
DPC
Triflocin
oj~-@ @COO-
:
d ~\Q;
F3
N~COO-
coo-
OO© ,
10- 8
• I
10-7
, I
10- 6
I
10- 5
••
L....• » I
10-4
10-4
I
10-3
1C50 (mmol' L-1) Fig. 13. Chloride channel blockers and their respective 1C50 values. i.e. the basolateral concentrations needed for half-maximal
inhibition of active NaCI reabsorption in the thick ascending limb of the loop of Henle. Data were obtained on isolated perfused cortical thick ascending limbs of the rabbit nephron. Torasemide is also a CI- channel blocker with a similar inhibitory potency to diphenylamine-2-carboxylate (OPC) [data from Oi Stefano et al. 1985; Wangemann et al. 1986; Wittner el al. 1986].
6. Conclusions The pharmacodynamics of loop diuretics are very similar. Loop diuretics inhibit the Na+/2CI-/ K+ carrier. Some are also Cl- channel blockers, but their affinity is rather poor. For torasemide, the affinity for the Na+/2Cl-/K+ carrier is IOO-fold higher than that for the Cl- channel. The inhibition of the Na+/2CI-/K+ carrier will therefore be the dominant effect of this substance if given in vivo. The pharmacokinetics of loop diuretics are variable, as some are cleared by metabolism, whereas others are metabolised to a lesser extent.
Some have a high lipid solubility with consequently slow and prolonged action, and others are predominantly water soluble. Some of the possible pharmacological applications of these substances other than for their diuretic effect are prevention of glial cell swelling, prevention of diarrhoea, relaxation of smooth muscle, etc. The possible use of loop diuretics for the treatment of brain oedema, however, would only be feasible with drugs of sufficient lipid solubility. To produce any of these additional effects, it would be necessary to concentrate the drug within the organ in question.
How Do Loop Diuretics Act?
References Brezis M, Rosen S, Silva P, Epstein FH. Renal ischemia: a new perspective. Kidney International 26: 375-383, 1984 Burg MB, Green N. Effect of mersalyl on the thick ascending limb of Henle's loop. Kidney International 4: 245-251, 1973 Burg MB, Stoner L, Cardinal J, Green N. Furosemide effect on isolated perfused tubules. American Journal of Physiology 225 (I): 119-124, 1973 Cousin JL, Motais R. Inhibition of anion transport in the red blood cell by anionic amphiphilic compounds. II. Chemical properties of the flufenamate-binding site on the band 3 protein. Biochimica et Biophysica Acta 687: 156-164, 1982 Oi Stefano A, Wittner M, Schlatter E, Lang HJ, Englert H, et al. Oiphenylamine-2-carboxylate, a blocker of the Cl- conductive pathway in Cl- transporting epithelia. Pfliigers Archiv 405 (Suppl. I): S95-S 100, 1985 Friimter E, Rumrich G, Ullrich KJ. Phenomenologic description of Na+, Cl- and HC03 absorption from proximal tubules of the rat kidney. Pfliigers Archiv 343: 189-220, 1973 Garg LC, Knepper MA, Burg MB, et al. Mineralocorticoid effects on Na-K-A TPase in individual nephron segments. American Journal of Physiology 240: F536-F544, 1981 Greger R. Coupled transport of Na+ and Cl- in the thick ascending limb of Henle's loop of rabbit nephron. Scandinavian Audiology 14 (Suppl.): 1-15, 1981a Greger R. Chloride reabsorption in the rabbit cortical thick ascending limb of the loop of Henle. A sodium dependent process. Pfliigers Archiv 390: 38-43, 1981 b Greger R. Ion transport mechanisms in thick ascending limb of Henle's loop of mammalian nephron. Physiological Reviews 65: 760-797, 1985 Greger R. Pathophysiologie der renalen Ischamie. Zeitschrift fUr Kardiologie 76 (Suppl. 4): 81-86,1987 Greger R. Chloride channel blockers. In Fleischer B, Fleischer S (Eds) Methods in enzymology, Vol. 5, Academy Press, Orlando, in press Greger R, Giigelein H. Patch clamp analysis of the ionic con. ductances in the rectal gland of squalus acanthias. Bulletin of the Mont Oesert Island Biological Lab 25: 116-119, 1985 Greger R, Lang HJ, Englert HC, Wangemann P. Blockers of the Na+j2Cl-jK+ carrier and of chloride channels in the thick ascending limb of the loop of Henle. In Puschett (Ed.) Oiuretics II. Chemistry, pharmacology and clinical applications, pp. 131137, Elsevier, Amsterdam, 1987a Greger R, Oberleithner H, Schlatter E, Cassola AC, Weidtke C. Chloride activity in cells of isolated perfused cortical thick ascending limbs of rabbit kidney. Pfliigers Archiv 399: 29-34, 1983 Greger R, Schlatter E. Cellular mechanisms of the action of loop diuretics on the thick ascending limb of Henle's loop. Klinische Wochenschrift 61: 1019-1027, 1983 Greger R, Schlatter E, Wittner M. Cellular mechanisms of action of furosemide-like diuretics in the thick ascending limb of the loop of Henle. In Puschett J (Ed.) Oiuretics I. Chemistry, pharmacology and clinical applications, pp. 215-221, Elsevier, Amsterdam, 1984c Greger R, Velasquez H. The cortical thick ascending limb and early distal convoluted tubule in the urinary concentrating mechanism. Kidney International 31: 590-596, 1987 Greger R, Wangemann P. Wittner M, Oi Stefano A, Lang HJ, et al. Blockers of active transport in the thick ascending limb of
13
the loop of Henle. In Adreucci, Oal Canton (Eds) Oiuretics: basic, pharmacological and clinical aspects, pp. 33-38, 1987b Greger R, Weidtke C, Schlatter E, Wittner M, Gebler B. Potassium activity in cells of isolated perfused cortical thick ascending limbs of rabbit kidney. Pfliigers Archiv 401: 52-57, 1984a Greger R, Wittner M, Schlatter E, Oi Stefano A. Na+j2Cl-jK+ co-transport in the thick ascending limb of Henle's loop and mechanism of action ofloop diuretics. In Hoshi T (Ed.) Coupled transport in nephron. Miura Foundation, Tokyo, 1984b Imai M. Effect of bumetanide and furosemide on the thick ascending limb of Henle's loop of rabbits and rats perfused in vitro. European Journal of Pharmacology 41: 409-416, 1977 Lesne M, Clerckx-Braun F, Cuvelier R, Van Ypersele de Strihou CH. Comparison between diuretic effects and pharmacokinetic parameters of torasemide and furosemide in human volunteers. Naunyn-Schmiedeberg's Archives of Pharmacology 330: 125, 1985 Oberleithner H, Lang F, Greger R, Wang W, Giebisch G. Effect of luminal potassium on cellular sodium activity in the early distal tubule of Amphiuma kidney. Pfliigers Archiv 396: 3440, 1983 Palmer LG, Frindt G. Amiloride-sensitive Na channels from the apical membrane of the rat cortical collecting tubule. Proceedings of the National Academy of Science USA 83: 2767, 1986 Sands JM, Knepper MA, Spring KR. Na-K-Cl cotransport in apical membrane of rabbit renal papillary surface epithelium. American Journal of Physiology 251: F475-F484, 1986 Schlatter E, Greger R, Weidtke C. Effect of 'high ceiling' diuretics on active salt transport in the cortical thick ascending limb of Henle's loop of rabbit kidney: correlations of chemical structure and inhibitory potency. Pfliigers Archiv 396: 210-217, 1983 Schlatter E, Schaefer JA. Electrophysiological studies in principal cells of rat cortical collecting tubules. Pfliigers Archiv 409: 8192, 1987 Sonnenberg H. Effects of furosemide, acetazolamide and mannitol on medullary collecting duct function in the rat kidney. Pfliigers Archiv 373: 113-123, 1978 Wangemann P, Wittner M, Oi Stefano A, Englert He, Lang HJ, et al. Cl- channel blockers in the thick ascending limb of the loop of Henle. Structure-activity relationship. Pfliigers Archiv 407 (Suppl. 2): 128-141, 1986 Wittner M, Oi Stefano A, de Rouffignac C, Roinel N. Basal Ca++ and Mg++ reabsorption in the mouse cTAL is driven by the transepithelial potential difference. 31 sl International Congress of Physiological Sciences, Finland, 1989 Wittner M, Oi Stefano A, Schlatter E, Oelarge J, Greger R. Torasemide inhibits NaCI reabsorption in the thick ascending limb of the loop of Henle. Pfliigers Archiv 407: 611-614,1986 Wittner M, Oi Stefano A, Wangemann P, Oelarge J, Greger R. Analogues of torasemide-structure-function relationships - experiments in the thick ascending limbs of the loop of Henle of rabbit nephron. Pfliigers Archiv 408: 54-62, 1987 Wittner M, Weidtke C, Schlatter E, Oi Stefano A, Greger R. Substract utilization in the isolated perfused cortical thick ascending limb of rabbit nephron. Pfliigers Archiv 402, 52-62, 1984
Author's address: Or M. Wittner, Service de Biologie Cellulajre, Oepartement de Biologie CEA-CEN Saclay, 91191 Gif-sur-Yvette, Cedex, France.
How Do Loop Diuretics Act? M. Wittner,l A. Di Stefano,l P. Wangemann 2 and R. Greger3 Institut de Recherche Fondamentale, Service de Biologie Cellulaire, Commisariat a l'Energie Atomique, Centre d'Etudes Nucleaires de Saclay, Gif-sur-Yvette, France 2 Boys' Town National Institute for Communication Disorders in Children, Omaha, Nebraska, USA 3 Physiologisches Institut, Universitat Freiburg, Freiburg, Federal Republic of Germany
Summary
In the thick ascending limb of the loop of Henle, NaCI reabsorption is mediated by a Na+j2Cl-jK+ cotransport system, present in the luminal membrane of this nephron segment. Loop diuretics such as furosemide (frusemide), piretanide, bumetanide and torasemide bind reversibly to this carrier protein, thus reducing or abolishing NaCl reabsorption. This leads to a decrease in interstitial hypertonicity and thus to a reduced water reabsorption. In nephron segments other than the thick ascending limb, loop diuretics have no quantitative importance with respect to their saluretic and diuretic activities. Loop diuretics also reduce Ca++ and Mg++ reabsorption in the thick ascending limb in a way which is still not clear. Furthermore, these drugs increase the urinary K+ excretion by enhancing distal tubular K+ secretion and reducing K+ reabsorption in the loop of Henle. Finally, by reduction of active NaCl transport, loop diuretics drastically reduce the substrate requirement and oxygkn dependence of the thick ascending limb cells. This renders these cells, which are chara.cterised by high transport rates and only limited substrate reserves, less vulnerable in acute renal failure.
Of the various classes of diuretics, the most powerful act in the thick ascending limb of the loop of Henle (Burg & Green 1973; Burg et al. 1973; Greger 1981a; Imai 1977; Schlatter et al. 1983). This review focuses on the cellular mechanisms of action of loop diuretics in the thick ascending limb of the loop of Henle. More specifically, the following issues will be discussed: the site of action of diuretics in the nephron; the protection against ischaemia and/or reduction in metabolic substrates of the thick ascending limb by loop diuretics; the structure-activity relationships for loop diuretics.
1. NaCI Reabsorption in the Nephron and Site of Action of Diuretics Figure 1 shows where NaCI and water are reabsorbed in the nephron. The numbers indicate the fractional delivery of NaCl and water to the different nephron segments as percentages of the filtered load. 60 to 70% of the filtered NaCI and water are reabsorbed iso-psmotically in the proximal tubule: In the loop of Henle, some 30% of NaCI and 10% of water are Feabsorbed. The reabsorption of NaCI occurs mainly in the thick ascending limb of
2
Fig. 1. Tubular reabsorption of NaCl and water. The numbers indicate the fractional delivery of NaCl and water as percentages of the filtered load.
the loop of Henle. This nephron segment dilutes the lumen fluid and generates a hypertonic interstitium, because it is able to reabsorb NaCI and simultaneously prevent a corresponding water reabsorption by its water impermeability. In antidiuresis, this process is responsible for the concentration of urine (withdrawal of water from the collecting ducts into the hypertonic interstitium). In water diuresis, the collecting ducts are water impermeable, and the diluted fluid, produced by the thick ascending limb, is excreted in the urine. Along the distal convoluted tubule, a further 8% of the filtered NaCI and 10 to 20% of the filtered water are reabsorbed. In the collecting duct, reabsorption of Na+ and Cl- continues, so that less than I% of the filtered load appears in the urine. Water reabsorption depends on the water permeability of the luminal cell membrane of the collecting duct and is regulated by antidiuretic hormone. The rate at which s~dium is reabsorbed is also hormonally regulated by! mineralocorticoids (aldosterone) and in some species by the antidiuretic hormone as well (Schla~ter & Schaefer 1987). NaCI reabsorption can be blocked by diu'retics at different sites in the nephron. In the proximal tubule, NaCI reabsorption can be inhibited by car-
Drugs 41 (Suppl. 3) 1991
bonic anhydrase inhibitors (e.g. ethoxzolamide, acetazolamide). The induced saluresis and diuresis, however, are not very pronounced, since some bicarbonate reabsorption occurs independently of the presence of carbonic anhydrase. Furthermore, the diuresis induced by carbonic anhydrase inhibitors is mostly compensated for by an increase in NaCI reabsorption in more distal nephron segments. In the thick ascending limb of the loop of Henle, NaCl reabsorption can be blocked effectively by the socalled loop diuretics [e.g. furosemide (frusemide), bumetanide, piretanide, torasemide]. In the early distal tubule, thiazide diuretics are effective. They seem to interfere with an electro-neutral NaCI cotransport system, present in the luminal cell membrane. Since the transport rate of this nephron (see above) is quite low, the diuretic effect of thiazides is small compared with that of loop diuretics. Diuretics, such as amiloride and triamterene, act in the collecting tubule by blocking luminal Na+ channels. The diuretic effect of these substances is also low, since only a small percentage of the filtered NaCI is reabsorbed in this nephron segment. These substances are, however, of importance since they are K+ sparing. Inhibition of luminal Na+ channels by amiloride or triamterene leads to an inhibition of K+ secretion. This K+ sparing effect may be of advantage if K+ secretion were increased by a saluresis, induced more proximal to the collecting duct. 1.1 Effects of Loop Diuretics of the Furosemide Type in Nephron Segments Other Than the Thick Ascending Limb
Figure 2 shows the cellular models for NaCl reabsorption in the proximal tubule, in the thick ascending limb and in the collecting duct. In all 3 segments, NaCl reabsorption is energised by the (Na+ /K+) -ATPase, localised in the basolateral membrane. In the thick ascending limb, NaCI reabsorption is mediated by a Na+ /2Cl-/K+ cotransport system present in the apical cell membrane. The (Na+/K+)-ATPase reduces the cytosolic Na+ concentration and thus provides the driving force for the operation of the Na+/2Cl-/K+
3
How Do Loop Diuretics Act?
carrier. The K+ taken up by the carrier recycles via a K+ conductance present in the luminal membrane. Cl- leaves the cell via Cl- channels and a KCI cotransport system present in the basolateral membrane (Greger 1985). The positive voltage of the lumen, due to the polarisation of the cell, drives paracellular Na+ reabsorption. The Na+/2Cl-/K+ cotransport system can be reversibly blocked by loop diuretics of the furosemide type, which bind to one of the Cl- binding sites of the carrier (Greger & Schlatter 1983). In the proximal tubule, 30% of the filtered Na+ is reabsorbed together with bicarbonate through the cell (transcellular), whereas 70% of the deliveredNa+ and 100% of Cl- are reabsorbed by means of the paracellular shunt pathway. Transcellular and paracellular reabsorption of the Na+, together with the reabsorption of Cl- and water, represent 60 to 70% of the filtered load (Fromter et al. 1973). Effects of furosemide in the proximal tubule are of minor importance and relate to the fact that loop diuretics act as rather weak inhibitors of carbonic anhydrase. The mechanisms by which NaCl is reabsorbed in the early distal tubule are presently being studied. NaCI reabsorption in this nephron segment is mediated by a carrier system, which is sensitive to thiazide
CELL
1.2 The Consequences of Inhibition of NaCI Reabsorption by Loop Diuretics Firstly, NaCl net transport in the thick ascending limb, comprising 20 to 30% of the filtered load, is considerable. Inhibition of this net transport by loop diuretics abolishes the hypertonicity of the interstitium, and so inhibits the water reabsorption in the collecting duct. The inhibition of NaCI reabsorption in the loop of Henle cannot be compensated for by more distal nephron segments, thus a pronounced saluresis and diuresis is induced. Secondly, loop diuretics produce a kaliuresis which has the following 2 causes: the small net reabsorption Cortical collecting tubule
Thick ascending limb
Proximal tubule
LUMEN
diuretics but insensitive to furosemide (Greger & Velasquez 1987). In the collecting duct, Na+ reabsorption and K+ secretion occur via luminal Na+ and K+ channels (Palmer & Frindt 1986), as there is no furosemide-sensitive transport system. It has, however, been claimed that furosemide acts in the papillary collecting duct and on the epithelium lining of the renal papilla (Sands et al. 1986; Sonnenberg 1978). It is very likely that none of these effects is of quantitative importance, with respect to the saluretic and diuretic activities of these drugs.
BLOOD
LUMEN
CELL
BLOOD
G Na+ Gluc
LUMEN
BLOOD
CELL
G Na+ K
Na+
Fig. 2. Mechanisms of NaCl reabsorption in the proximal tubule, the thick ascending limb of the loop of Henle and the collecting tubule. - = ionic channels; . - = primary active pump; 0- = carrier system.
Drugs 41 (Suppl. 3) 1991
4
cells, is abolished; Na+ reabsorption and K+ secretion in the collecting duct depend on the luminal NaCl concentration and the luminal flow rate. Both are increased after application of loop diuretics, which are thus responsible for increases in Na+ reabsorption and K+ secretion.
;:- 200 I
E
E 150 I
C
"::> E •
;: 0 0;
z
E Q.
100
50
o PRE FUR POST
PRE FUR
POST
;:- 1.00 I
E
~ 0.75
I
C
x ::>
'E 0.50 •
;:0
- E ~ S 0.25
o (6)
PRE FUR
POST
PRE FUR POST
+3 ;:I
E E
" ~ 0;
c
0
'j"
c
E
(5
+ E :.:: S
-3
(7)
PRE
FUR
POST
Fig. 3. Effect of furosemide (FUR) [10- 4 mOI/L, lumen) on transepithelial ion net fluxes ofNa+, CI-, K+, Ca++ and Mg++ (pmol/minjmm), measured by electron probe analysis in isolated perfused cortical thick ascending limbs of the mouse nephron. PRE = before furosemide; POST = after furosemide. Furosemide reversibly inhibits Na+, CI-, K+, Ca++ and Mg++ reabsorptions. Number of tubules in brackets. Mean values ± SEM. * = p < 0.05 (data from Wittner et al. 1989).
of K+ in the thick ascending limb is abolished by loop diuretics, since the Na+ /2CI-/K+ carrier is blocked, and the lumen-positive transepithelial voltage maintaining the passive reabsorption ofK+ across the cell, and more importantIybetween the
2. Cellular Mechanisms of Action of Loop Diuretics Figure 3 shows the effects of furosemide on the transepithelial ion net fluxes ofNa+, CI-, K+, Ca++ and Mg++, measured in isolated perfused cortical thick ascending limbs of the mouse nephron by electron probe analysis (Wittner et al. 1989). Furosemide abolished not only the transepithelial Na+ and CI- net fluxes, but also inhibited Ca++ and Mg++ reabsorptions and the small net reabsorption of K+. The inhibitory effect offurosemide on electrolyte transport was completely reversible after removal of the drug from the lumen perfusate. Recent observations from our laboratory indicate that Ca++ and Mg++ are reabsorbed by paracellular and transcellular pathways. The paracellular component is driven by the transepithelial voltage (Wittner et al. 1989). Inhibition of Ca++ and Mg++ transport by furosemide could therefore be explained by the fact that furosemide, by abolishing the transepithelial voltage, abolishes the driving force for Ca++ and Mg++ reabsorption. Figure 4 shows the effects of furosemide on the transepithelial potential difference (PO te ) and the voltage across the basolateral membrane (PObl) of an isolated, perfused, cortical thick ascending limb segment. Furosemide added to the lumen perfusate at a concentration of 5 X 10- 5 molfL leads, with a time constant of less than half a second, to a reduction of PO te . Simultaneously, PObl hyperpolarises, and the transepithelial resistance and fractional resistance of the basolateral cell membrane increase (increase in the size of the voltage transients across the entire epithelium and the basolateral cell membrane). The magnitude of the voltage transients, obtained by the periodic injection of current pulses into the tubular lumen, is proportional to the respective resistances. The in-
5
How Do Loop Diuretics Act?
,.
-
>
.
§.
r5
0..
-"~
~
('
-5
(
,
0
+5 - 20
I"
./' • 1£
--
lS C 0..
I
- ...y.
I -
> §.
-
-
f- ~
I
r
I
-40 -60
I
("
-
, -f -
"\. f""- ~
~
-80
':
\
-100
n ~ r,-.\; 1-0.. .,
o
5
10
Time (sec)
Fig. 4. Effect offurosemide (5 X 10- 5 mol/L. lumen) on transepithelial potential difference (PDtcl and on the potential difference across the basolateral membrane (PDbl). At 5-second intervals. short current pulses are injected into the lumen to obtain a measurement of the transepithelial resistance and the fractional resistance across the basolateral membrane. The collapse of PD tc . the hyperpolarisation of PD tc . the increase in the transepithelial resistance (increase in height of the transepithelial voltage transient) and the increase in the fractional resistance across the basolateral membrane are observed (data from Greger & Schlatter 1983).
crease in the basolateral membrane resistance after application of furosemide is due to a decrease in cytosolic CI- activity. This decrease in activity has the following 2 causes: the Na+/2CI-/K+ carrier is blocked, and therefore no more CI- enters the cell; CI- leaves the cell via CI- channels and the KCI cotransport system of the basolateral membrane until the cytosolic chloride concentration is in Nernst equilibrium. For a membrane voltage of -90mV, this would correspond to a cytosolic CIactivity of 5 mmol/L. Very similar values have been measured with CI- selective electrodes (Greger et al. 1983). Table I shows that there was a decrease in cytosolic CI- activity from 19 to 7 mmol/L in the presence of furosemide, which suggests a passive distribution of CI- across the basolateral membrane in the presence of this loop diuretic. The final status after application of furosemide is shown by the model in figure 5. Furosemide 'locks' the Na+/2C\-/K+ carrier. Thus, the entry of Na+, CIand K+ into the cell ceases instantaneously. The cell CI-, which is above equilibrium concentrations under control conditions, declines towards passive equilibrium, because the entry step for this ion is blocked, and outflux across the basolateral membrane via Cl- channels and the KCl cotransport system continues until CI- reaches this equilibrium (Greger et al. 1983). Cell Na+ also declines. In the Amphiuma diluting segment, a decrease from 9 to 4 mmoljL has been reported (Oberleithner et al. 1983). This decrease in Na+ concentration is
Table I. Estimates of intracellular CI- activity in isolated perfused cortical thick ascending limb of the rabbit nephron. In the presence of furosemide, CI- is passively distributed (data from Oberleithner et al. 1983) PDbl (mV)
PDcl(mV)
ECI(mV)
ACI(mmol/L)
Pretreatment control
-74
- 27
+ 47
-'> 19
Furosemide a
- 88
-16
+ 72
-'> 7
Post-treatment control
-74
- 26
+ 48
-'>18
a
3 X 10- 5 mol/L added to lumen perfusate, mean of 29 observations.
Abbreviations and symbols: PDbl = electrical potential difference across the basolateral membrane; PDclby CI- selective microelectrodes; ECI- = chemical potential for CI-; ACI- = intracellular CI- activity.
= potential difference read
Drugs 41 (Suppl. 3) 1991
6
lumen
I
Cell
+90mV
I
3. Protection of the Thick Ascending Limb Against Ischaemia and/or Reduction in Metabolic Substrates by Loop Diuretics
Furosemide
. . - - - - - - Na· -
I
Perilubule space
-90mV
J
-----+
Fig. 5. The thick ascending limb cell in the presence of furosemide. Net transport is reduced to zero, and the cell potential approaches the K+ equilibrium potential. K+ and Cl- are passively distributed across both cell membranes. The (Na+/K+) -ATPase activity ceases as a result of blocked Na+ uptake.
due to the pumping of Na+ out of the cell by the (Na+ /K+) -ATPase until a minimum concentration of Na+ is reached. Then the (Na+/K+) -ATPase comes to a standstill and no longer consumes adenosine triphosphate (ATP). Cell K+ does not change after furosemide application (Greger et al. 1984a). After inhibition of the Na+/2Cl-/K+ carrier, K+ leaks out of the cell across the luminal membrane via its conductive pathway and along its electrochemical gradient. On the other hand, K+ is pumped into the cell as long as the (Na+/K+)-ATPase is operating. These opposing effects appear to balance each other, so that the intracellular K+ activity remains fairly constant. The luminal and basolateral cell membranes hyperpolarise to the K+ equilibrium potential. Cland K+ are passively distributed across the 2 cell membranes. The ion net fluxes across the cell membranes and across the shunt pathway therefore become zero. In addition, the transports of Ca++ and Mg++ are inhibited.
The cells of the thick ascending limb are particularly rich in mitochondria, and the density of the (Na+/K+) -ATPase, localised in the basolateral membrane, is high (Garg et al. 1981). NaCi reabsorption in the cortical and medullary parts of this nephron segment are energised by aerobic metabolism. In the event of oxygen shortage, the amount of ATP produced solely by glycolysis is not sufficient to maintain active NaCI transport (Wittner et al. 1984). Substrates are taken up from the peritubular side via simple (butyrate, acetate) or facilitated (sugars and small anions such as lactate or pyruvate) diffusion. From the lumen side, only butyrate can enter the cell via non-ionic diffusion (Wittner et al. 1984). Figure 6 shows the close link between active NaCI transport in the thick ascending limb and substrate availability. At time zero, all substrates were removed from both sides of the epithelium. The equivalent short circuit current, which corresponds stoichiometrically to the rate of active NaCI reabsorption (Greger 198Ia), collapsed within 10 minutes. This inhibition was reversible if substrate removal was only transient (10 min300
if 200 E ()
§.
ID·Glucose
I +
25
~ -100 ,
a
I
Iii
5
Time (mm)
Fig. 8. Effect of substrate removal on the basolateral membrane potential (PObl) in an isolated perfused cortical thick ascending limb segment of the rabbit nephron. The substrate (5 mmol/L O-glucose) was removed and furosemide (5 x lQ-5 moljL) was simultaneously added to the lumen perfusate. This caused a hyperpolarisation of PObl. After removal of furosemide, PObl depolarised. Readmission of glucose led to a hyperpolarisation of PObl back to control values. Substrate removal is devoid of effect on the cell as long as furosemide is present (data from Greger et al. 1984c).
Drugs 41 (Suppl. 3) 1991
8
PDbl depolarised with a delay of only 3 minutes. Readmission of glucose led to a hyperpolarisation of PDbl back to the control value (Greger et al. 1984c). This shows that loop diuretics of the furosemide type, by abolishing active NaCI transport, cause the cells to enter a resting state, which renders them almost independent of metabolic fuels. Figure 9 shows the protective character of furosemide against inhibition of the (Na+/K+) ATPase by ouabain or phloretin. It is evident that neither ouabain nor phloretin has any effect on the cell potential, as long as furosemide is present. Both substances lead to a rapid depolarisation of the cell after removal of furosemide (Greger & Schlatter 1983). It seems, therefore, that the effects of substrate removal and/or direct inhibition of the (Na+/ K+) -ATPase can be prevented by blocking the Na+ / 2Cl-/K+ cotransporter with furosemide. These data emphasise how loop diuretics such as furosemide can be used to protect the thick ascending limb against the hazards of substrate shortage and hypoxia.
4. Torasemide, a New Loop Diuretic Figure 10 shows the effect of a new loop diuretic, torasemide, on NaCI reabsorption in the thick ascending limb. The structural formula of to-
rasemide (fig. II), which is a powerful diuretic (Lesne et al. 1985), suggests that it may interfere with the Na+/2CI+/K+ cotransporter. Like furosemide and related compounds, torasemide contains an anionic group and a secondary amine. The nitrogen of the pyridine, in meta position to the anionic group, should correspond to the sulphonamide group offurosemide and related compounds (Schlatter et al. 1983). As evident from the doseresponse curves, obtained when torasemide was added either to the lumen or to the bath perfusate (fig. 10), this substance acts mainly from the luminal side by interfering with the Na+/2Cl-/K+ carrier. Half maximal inhibition of active NaCI reabsorption, measured as the equivalent short circuit current (lsd, was achieved at 3 X 10- 7 mol/L (Wittner et al. 1986). Figure II shows that this value is comparable to that found for powerful diuretics such as furosemide derivatives and bumetanide (Schlatter et al. 1983). The torasemide-mediated inhibition of the equivalent short circuit current, which is rapidly reversible, is due to a fall in the transepithelial lumen positive voltage and a moderate increase in the transepithelial resistance. These effects and the hyperpolarisation of the basolateral membrane potential in the presence of luminal torasemide (Wittner et al. 1986) are qualitatively and Ouabain Furosemide
o - 20
;;-
- 40
S - 60 21
o
0..
- 80
I
o
I
0.5
I
3.0
Time (min)
Fog. 9. Inhibition of the (Na+jK+) -ATPase in the presence and absence of furosemide (5 X 10- 5 moljL, lumen) ina cortical
thick ascending limb cell. Furosemide leads to a hyperpolarisation of the potential difference across the basolateral membrane (PDbl) during which phloretin (5 X 10- 6 moljL, bath) has no effect on PDbl. As soon as furosemide is removed, phloretin depolarises PDbl. Upon removal of phloretin, the potential difference of the cell recovers, and furosemide again leads to a small hyperpolarisation. Ouabain, like phloretin, is then without effect on PDbl. Only after removal of furosemide can ouabain lead to its typical depolarisation (data from Greger & Schlatter 1983).
How Do Loop Diuretics
9
Act~
300 (10)
crE
200
"
100
.J!'
o
\ (6\
~~
(11)
u
(4)
\,---1 '\~3) I",' "
(5)
(5)
'I__ -- __(2) ,
~.~I----,,------,,------,,----~,
10- 7
10- 6
10-5
10-4
Torasemide (mol' L- ')
Fig. 10. Dose response curves for torasemide added to the luminal (X---X) and to the peritubular perfusates (0-0 ). The equivalent short circuit current (I se) is plotted versus the torasemide concentration. A half maximal effect of torasemide is achieved at 3 x 10- 7 moljL in the luminal perfusate and at 3 x lQ-5 mol/L in the peritubular perfusate. Numbers in brackets refer to the number of tubules tested. Mean values ± SEM (data from Wittner et al. 1986).
even quantitatively similar to those obtained for other diuretics of the furosemide type (Schlatter et al. 1983). Like these loop diuretics, torasemide may interfere by means of its negatively charged sulphonyl urea moiety with one of the Cl- binding sites of the N a +/2CI-/K + transport protein. The blocked uptakes of Na+, K+ and Cl- lead to a fall in cytosolic Cl- activity. This leads to a hyperpolarisation of the cell towards the K+ equilibrium potential (Schlatter et al. 1983) and to a collapse of the transcellular current flow. Torasemide, unlike furosemide, acts also from the peri tubular side (fig. 10), probably by interfering with the Cl- channels of the basolateral membrane. When compared with the effects exerted from the luminal side, far higher doses are required to inhibit the equivalent short circuit current from the peritubular side. A concentration of 3 X 10- 5 mol/L is needed to inhibit the equivalent short circuit current by 50% (Wittner et al. 1986). The discovery that the effects of luminal and peritubular torasemide are qualitatively similar and that torasemide, unlike furosemide, is far more lipophilic (Wittner et al. 1986), suggests that the peri tubular effects of torasemide are caused by a diffusion
across the cell and an interference with the luminal Na+ /2CI -/K+ cotransporter. Nevertheless, the following findings seem to disprove this hypothesis: torasemide acts equally rapidly from both sides of the cell, whereas furosemide and related compounds at very high concentrations (see below) act quickly from the luminal side, but only slowly from the peri tubular side; with furosemide, peritubular effects are only seen if the peritubular concentration exceeds 10- 3 mol/L. These effects are only slowly reversible; as shown by the patch-clamp recording given in figure 12, torasemide, unlike furosemide but like other Cl- channel blockers (Oi Stefano et al. 1985; Wangemann et al. 1986), inhibits Cl- channels in excised membrane patches of the luminal membrane of the shark rectal gland (Greger & Gbgelein 1985). These Cl- channels are similar if not identical to those present in the basolateral cell membrane of the thick ascending limb (Oi Stefano et al. 1985; Greger & Gbgelein 1985). The interference of torasemide with CI- channels leads to an inhibition of the conductive exit of this ion from the cell and thus reduces the transcellular current and the rate of Cl- reabsorption.
5. Structure-Activity Relationship Figure II shows several well-known loop diuretics and their respective IC50 values. These values correspond to the luminal concentration needed to inhibit 50% of NaCI reabsorption. In 2 previous studies (Schlatter et al. 1983; Wittner et al. 1987) the structure-activity relationships of analogues of furosemide and torasemide have been examined. The following are requirements for reversible interaction with the Na+/2Cl-/K+ carrier: an anionic group such as a sulphonate, carboxylate, tetrazolate or sulphonyl urea is mandatory; a secondary (e.g. furosemide, torasemide) or a tertiary (e.g. piretanide) amine is needed in ortho position or meta position to the anionic group; this amino group must link the anionic moiety to an apolar residue; in meta position to the amino group a sulphonamide group (e.g. furosemide, piretanide) or a pyridino ring (e.g. torasemide) is required. As was previously shown (Greger, in press; Gre-
Drugs 41 (Suppl. 3) 1991
10
BTS 39542 Xi pam ide
Torasemide
L-Ozolinone
Indacrinone CH3 0
CI
@C@r o :
:
CI
:
O-CH2-COO: Muzolimine .
Furosemide
I 10- 8
,, , I 10- 7
t
t
I 10-- 6
,
C'if~-\. CI J.8.I N..( . . NH2
, I 10- 5
CH3 0
I 10-4
,
:.... ..... .. :.......~ »
10-4
I
10- 3
IC50 (mmol· L- 1) Fig. 11. Loop diuretics and their· respective 1C50 values. The IC50 value is the luminal concentration at which the substance
inhibits NaCI reabsorption half maximally. Data were obtained on isolated perfused cortical thick ascending limbs of the rabbit nephron (data from Schlatter et al. 1983; Wittner et al. 1987).
ger et al. 1987a,b), loop diuretics have structural similarities to Cl- channel blockers and probably also to blockers of the Cl- /HC03- exchanger of the red blood cell membrane (Cousin & Motais 1982). Figure 13 shows some Cl- channel blockers and the concentrations needed to inhibit NaCI reabsorption half maximally. One of the most effective Cl- channel blockers is 5-nitro-2-(3-phenylpropylamino)-benzoate, which inhibits the equivalent short circuit current half maximally at 8 x 10- 8 mol/L (Wangemann et al. 1986). Torasemide, with its chemical structure placing it between loop diuretics and Cl- channel blockers, has an affinity for Cl-channels similar to that of diphenylamine2-carboxylate, the first compound shown to specifically inhibit Cl- channels in the thick ascending limb (Di Stefano et al. 1985). Triflocine and niflu-
mate, which are inefficient on the Cl- channel and on the Na+/2CI-/K+ carrier, are known to inhibit the band 3 protein in red blood cells (Cousin & Motais 1982). 5.1 The Common Features of, and the Differences between, Loop Diuretics and Cl- Channel Blockers All these compounds, most of which are carboxylates, are weak acids with pK-values ,., 4 to 6. All compounds carry an amino nitrogen at a certain distance from the carboxylate group. The ortho-positioning and meta-positioning between the 2 groups leads to active compounds, whereas greater distances lead to inactive ones. Chloride channel blockers mayor may not have a substituent on the
II
How Do Loop Diuretics Act?
right hand side ring (figs 11, 13); a chloride substitution in para position or a nitro substitution in meta position to the carboxylate group. This moiety carries a partial negative charge. The blockers of the Na+/2Cl-/K+ carrier possess an electronegative sulphonamide group in the same position. Finally, both classes of substances possess an apolar moiety (phenyl moiety, furfuryl moiety, etc.). For blockers of the Na+/2Cl-/K+ carrier, the restrictions on this moiety are not very marked, for example, in the case of furosemide, a Cl-substitution can replace the apolar moiety. For chloride channel blockers, the apolar residue is subject to very narrow spatial constraints; it must be a ring structure, and this ring must be separated by a well-defined distance from the amino group. In general, the blockers of chloride channels are more hydrophobic than the blockers of the N a +/2Cl-/K + carrier. Finally, loop diuretics can be 'converted' into Cl- channel blockers and vice versa by small modifications of the molecule (Greger et al. 1987b). For
example, removal of chloride in the furosemide molecule reduces the inhibitory effect on the Na+/ 2Cl-/K+ carrier. Further replacement of the sulphonamide group by an amino group abolishes any inhibitory activity. Replacement now of the amino group by a nitro group leads to an inhibitor of Clchannels, with no inhibitory effect on the Na+ /2Cl-/ K+ carrier. Comparable observations have been made in an extensive study of torasemide analogues; replacement of the pyridino nitrogen by a N02-Substituted carbon atom abolished the ability to block the Na+/2Cl-/K+ carrier, but did not affect the inhibitory potency for Cl- channels (Wittner et al. 1987). The structural similarities between loop diuretics and Cl- channel blockers and the fact that minor structural modifications convert a loop diuretic into a Cl- channel blocker and vice versa suggest that the different membrane transport proteins share in common some portion of the molecule.
NaCI/NaCI, -24mV Torasemide 1 mmol· L-l, 0-8 sec
9-16 sec
17-24 sec
Control. 0-8 sec
Control, 9-16 sec
2pA
L 800 msec
Fig. 12. Patch-clamp recording of an excised luminal chloride channel of the rectal gland of the shark. Torasemide
(10- 4 mol/L) reversibly blocks chloride channel activity (data from Greger & G6gelein 1985).
Drugs 41 (Suppl. 3) 1991
12
Flufenamate
•
:
H(D)
Torasemide
~Nr@
N02~COO-
CH3)j-j(
~-.lQ\
.~c:.!l- ~ : s coo133B
DPC
Triflocin
oj~-@ @COO-
:
d ~\Q;
F3
N~COO-
coo-
OO© ,
10- 8
• I
10-7
, I
10- 6
I
10- 5
••
L....• » I
10-4
10-4
I
10-3
1C50 (mmol' L-1) Fig. 13. Chloride channel blockers and their respective 1C50 values. i.e. the basolateral concentrations needed for half-maximal
inhibition of active NaCI reabsorption in the thick ascending limb of the loop of Henle. Data were obtained on isolated perfused cortical thick ascending limbs of the rabbit nephron. Torasemide is also a CI- channel blocker with a similar inhibitory potency to diphenylamine-2-carboxylate (OPC) [data from Oi Stefano et al. 1985; Wangemann et al. 1986; Wittner el al. 1986].
6. Conclusions The pharmacodynamics of loop diuretics are very similar. Loop diuretics inhibit the Na+/2CI-/ K+ carrier. Some are also Cl- channel blockers, but their affinity is rather poor. For torasemide, the affinity for the Na+/2Cl-/K+ carrier is IOO-fold higher than that for the Cl- channel. The inhibition of the Na+/2CI-/K+ carrier will therefore be the dominant effect of this substance if given in vivo. The pharmacokinetics of loop diuretics are variable, as some are cleared by metabolism, whereas others are metabolised to a lesser extent.
Some have a high lipid solubility with consequently slow and prolonged action, and others are predominantly water soluble. Some of the possible pharmacological applications of these substances other than for their diuretic effect are prevention of glial cell swelling, prevention of diarrhoea, relaxation of smooth muscle, etc. The possible use of loop diuretics for the treatment of brain oedema, however, would only be feasible with drugs of sufficient lipid solubility. To produce any of these additional effects, it would be necessary to concentrate the drug within the organ in question.
How Do Loop Diuretics Act?
References Brezis M, Rosen S, Silva P, Epstein FH. Renal ischemia: a new perspective. Kidney International 26: 375-383, 1984 Burg MB, Green N. Effect of mersalyl on the thick ascending limb of Henle's loop. Kidney International 4: 245-251, 1973 Burg MB, Stoner L, Cardinal J, Green N. Furosemide effect on isolated perfused tubules. American Journal of Physiology 225 (I): 119-124, 1973 Cousin JL, Motais R. Inhibition of anion transport in the red blood cell by anionic amphiphilic compounds. II. Chemical properties of the flufenamate-binding site on the band 3 protein. Biochimica et Biophysica Acta 687: 156-164, 1982 Oi Stefano A, Wittner M, Schlatter E, Lang HJ, Englert H, et al. Oiphenylamine-2-carboxylate, a blocker of the Cl- conductive pathway in Cl- transporting epithelia. Pfliigers Archiv 405 (Suppl. I): S95-S 100, 1985 Friimter E, Rumrich G, Ullrich KJ. Phenomenologic description of Na+, Cl- and HC03 absorption from proximal tubules of the rat kidney. Pfliigers Archiv 343: 189-220, 1973 Garg LC, Knepper MA, Burg MB, et al. Mineralocorticoid effects on Na-K-A TPase in individual nephron segments. American Journal of Physiology 240: F536-F544, 1981 Greger R. Coupled transport of Na+ and Cl- in the thick ascending limb of Henle's loop of rabbit nephron. Scandinavian Audiology 14 (Suppl.): 1-15, 1981a Greger R. Chloride reabsorption in the rabbit cortical thick ascending limb of the loop of Henle. A sodium dependent process. Pfliigers Archiv 390: 38-43, 1981 b Greger R. Ion transport mechanisms in thick ascending limb of Henle's loop of mammalian nephron. Physiological Reviews 65: 760-797, 1985 Greger R. Pathophysiologie der renalen Ischamie. Zeitschrift fUr Kardiologie 76 (Suppl. 4): 81-86,1987 Greger R. Chloride channel blockers. In Fleischer B, Fleischer S (Eds) Methods in enzymology, Vol. 5, Academy Press, Orlando, in press Greger R, Giigelein H. Patch clamp analysis of the ionic con. ductances in the rectal gland of squalus acanthias. Bulletin of the Mont Oesert Island Biological Lab 25: 116-119, 1985 Greger R, Lang HJ, Englert HC, Wangemann P. Blockers of the Na+j2Cl-jK+ carrier and of chloride channels in the thick ascending limb of the loop of Henle. In Puschett (Ed.) Oiuretics II. Chemistry, pharmacology and clinical applications, pp. 131137, Elsevier, Amsterdam, 1987a Greger R, Oberleithner H, Schlatter E, Cassola AC, Weidtke C. Chloride activity in cells of isolated perfused cortical thick ascending limbs of rabbit kidney. Pfliigers Archiv 399: 29-34, 1983 Greger R, Schlatter E. Cellular mechanisms of the action of loop diuretics on the thick ascending limb of Henle's loop. Klinische Wochenschrift 61: 1019-1027, 1983 Greger R, Schlatter E, Wittner M. Cellular mechanisms of action of furosemide-like diuretics in the thick ascending limb of the loop of Henle. In Puschett J (Ed.) Oiuretics I. Chemistry, pharmacology and clinical applications, pp. 215-221, Elsevier, Amsterdam, 1984c Greger R, Velasquez H. The cortical thick ascending limb and early distal convoluted tubule in the urinary concentrating mechanism. Kidney International 31: 590-596, 1987 Greger R, Wangemann P. Wittner M, Oi Stefano A, Lang HJ, et al. Blockers of active transport in the thick ascending limb of
13
the loop of Henle. In Adreucci, Oal Canton (Eds) Oiuretics: basic, pharmacological and clinical aspects, pp. 33-38, 1987b Greger R, Weidtke C, Schlatter E, Wittner M, Gebler B. Potassium activity in cells of isolated perfused cortical thick ascending limbs of rabbit kidney. Pfliigers Archiv 401: 52-57, 1984a Greger R, Wittner M, Schlatter E, Oi Stefano A. Na+j2Cl-jK+ co-transport in the thick ascending limb of Henle's loop and mechanism of action ofloop diuretics. In Hoshi T (Ed.) Coupled transport in nephron. Miura Foundation, Tokyo, 1984b Imai M. Effect of bumetanide and furosemide on the thick ascending limb of Henle's loop of rabbits and rats perfused in vitro. European Journal of Pharmacology 41: 409-416, 1977 Lesne M, Clerckx-Braun F, Cuvelier R, Van Ypersele de Strihou CH. Comparison between diuretic effects and pharmacokinetic parameters of torasemide and furosemide in human volunteers. Naunyn-Schmiedeberg's Archives of Pharmacology 330: 125, 1985 Oberleithner H, Lang F, Greger R, Wang W, Giebisch G. Effect of luminal potassium on cellular sodium activity in the early distal tubule of Amphiuma kidney. Pfliigers Archiv 396: 3440, 1983 Palmer LG, Frindt G. Amiloride-sensitive Na channels from the apical membrane of the rat cortical collecting tubule. Proceedings of the National Academy of Science USA 83: 2767, 1986 Sands JM, Knepper MA, Spring KR. Na-K-Cl cotransport in apical membrane of rabbit renal papillary surface epithelium. American Journal of Physiology 251: F475-F484, 1986 Schlatter E, Greger R, Weidtke C. Effect of 'high ceiling' diuretics on active salt transport in the cortical thick ascending limb of Henle's loop of rabbit kidney: correlations of chemical structure and inhibitory potency. Pfliigers Archiv 396: 210-217, 1983 Schlatter E, Schaefer JA. Electrophysiological studies in principal cells of rat cortical collecting tubules. Pfliigers Archiv 409: 8192, 1987 Sonnenberg H. Effects of furosemide, acetazolamide and mannitol on medullary collecting duct function in the rat kidney. Pfliigers Archiv 373: 113-123, 1978 Wangemann P, Wittner M, Oi Stefano A, Englert He, Lang HJ, et al. Cl- channel blockers in the thick ascending limb of the loop of Henle. Structure-activity relationship. Pfliigers Archiv 407 (Suppl. 2): 128-141, 1986 Wittner M, Oi Stefano A, de Rouffignac C, Roinel N. Basal Ca++ and Mg++ reabsorption in the mouse cTAL is driven by the transepithelial potential difference. 31 sl International Congress of Physiological Sciences, Finland, 1989 Wittner M, Oi Stefano A, Schlatter E, Oelarge J, Greger R. Torasemide inhibits NaCI reabsorption in the thick ascending limb of the loop of Henle. Pfliigers Archiv 407: 611-614,1986 Wittner M, Oi Stefano A, Wangemann P, Oelarge J, Greger R. Analogues of torasemide-structure-function relationships - experiments in the thick ascending limbs of the loop of Henle of rabbit nephron. Pfliigers Archiv 408: 54-62, 1987 Wittner M, Weidtke C, Schlatter E, Oi Stefano A, Greger R. Substract utilization in the isolated perfused cortical thick ascending limb of rabbit nephron. Pfliigers Archiv 402, 52-62, 1984
Author's address: Or M. Wittner, Service de Biologie Cellulajre, Oepartement de Biologie CEA-CEN Saclay, 91191 Gif-sur-Yvette, Cedex, France.
