Characteristics and regulation of 11β-hydroxysteroid dehydrogenase of proximal and distal nephron

et Biophysics Acta

ELSEVIER

Biochimica et Biophysics Acta 1243 (1995) 461-468

Characteristics and regulation of 11 P-hydroxysteroid proximal and distal nephron

dehydrogenase

of

Nadia Alfaidy a, Marcel Blot-Chabaud a, Daniel Robic b, Sabine Kenouch a, Richard Bourbouze b, Jean-Pierre Bonvalet a, Nicolette Farman a,* a INSERM (1246, 1JERX. Bichat, Institut F&d&&if de Recherches ‘cellules kpitheliales’, BP 416, 75870 Paris CMex 18, France b Fact&

de Pharmacie, lJniuersit& Paris V, 75006 Paris, France

Received 28 June 1994; accepted 19 October 1994

Abstract Enzymatic properties of the enzyme lip-hydroxysteroid dehydrogenase (ll-HSD), which confers mineralocorticoid selectivity, have been explored in the aidosterone-sensitive collecting duct (CCD) and the aldosterone-insensitive Pars Recta (PR) of the rat kidney. After incubation of freshly isolated tubular segments with [3H]corticosterone c3H-B) or [3H]dehydrocorticosterone c3H-A), the rate of transformation of 3H-B into 3H-A (dehydrogenase activity), or the reverse reaction (reductase activity) were measured by HPLC. V,,, for dehydrogenase activity was found to be 8- to lo-fold higher in CCD than PR. The enzyme functions over a very wide range (0.1-5000 nM) of corticosterone conclentration. In CCD, enzyme kinetics suggest either the presence of two ll-HSD forms, differing by their affinity for corticosterone, 01: complex kinetics. Addition of NAD or NADP to permeabilized tubules revealed that dehydrogenase activity is NAD-dependent in CCD and NADP-dependent in PR. Cofactor addition was ineffective in intact tubules. CCD exhibited an exclusive dehydrogenase activity, whereas in PR dehydrogenase and reductase activity were found. No regulation of dehydrogenase activity could be evidenced in adrenalectomized rats receiving or not aldosterone, corticosterone or dexamethasone, for 2 h, 3 days or 4 days. We conclude that ll-HSD in the CCD and PR differs by its V,,, and cofactor dependence. Corticosteroid hormones do not influence 1 l-HSD activity. Keywords:

Aldosterone; Glucocorticoids; Kidney; NAD; NADP

1. Introduction Mineralocorticoid where

aldosterone

selectivity can selectively

consists bind

in

a

to its own

situation miner-

(MR), despite much higher circulating levels of glucocorticoid hormones, which display the same affinity for MR than aldosterone [1,2]. Protection of MR against illicit occupancy by glucocorticoid hormones has been shown to depend on the presence of 11 P-hydroxysteroid dehydrogenase (ll-HSD), an enzyme which metabolizes natural glucocorticoids into 1 l-dehydro-derivatives which have a low affinity for MR [2]. In cells which possess MR the presence of ll-HSD is thus required to express aldosterone-specific regulated effects. It has been shown that ll-HSD catalytic activity along the nephron [3-61 was the highest in the distal parts of the alocorticoid

receptor

* Corresponding author. Fax: +33 1 42291644. 0304-4165/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved SSDI 0304.4165(94)00173-l

nephron, i.e. in the aldosterone-sensitive tubular segments, which express MR [7], a finding also reported by NarayFejes-Toth et al. [S], on immunoselected cells of rabbit cortical collecting tubule. However, 1 l-HSD catalytic activity is also present, although at lower levels, in the proximal tubule, which is aldosterone-insensitive [3-61. On the other hand, when using antibodies against the purified liver ll-HSD for immunolocalization within the kidney, several groups [1,9,10] showed immunolabelling in proximal, not distal tubule. Since these latter results were hardly compatible with the role of ll-HSD as protector of MR, several investigators suggested the possibility of several forms of ll-HSD, with different tissue-specific expression, cofactor dependence and physiological role [ 1 l141. Whereas the hepatic form of the enzyme has been characterized at the molecular level [15], the molecular characteristics of a putative enzymatic form responsible for MR protection is still unknown. The kinetics of ll-HSD

462

N. Alfaidy et al. /Biochimica

et Biophysics Acta 1243 (1995) 461-468

has been investigated in details in microsomial fractions from the rat liver but little is known on these parameters, concerning the rat collecting duct enzyme. In rabbit collecting duct cells, the enzyme has an apparent affinity for corticosterone of about 3-6 . 10-a M [8,16] and a V,,, of 17 fmol per min and per 1000 cells [8]. The aim of this study was to compare the kinetics of ll-HSD in proximal and distal nephron of rat kidney, which are unknown at the present time. Indeed, the enzyme is not expressed at the same levels along the nephron of the rabbit, as compared to the rat [6]. Whether it could correspond to different forms of ll-HSD is important to know in view of the search for the molecular species involved in MR protection. Another feature that appears to distinguish different forms of ll-HSD is their cofactor dependence. Recently, Walker et al. [17] reported that NAD as well as NADP increased the ll-HSD activity in rat kidney homogenates (to a similar extent). This NAD and NADP dependence was also found on tubular suspensions, however without difference between proximal and distal tubule preparations [17]. The cofactor dependence was further examined in rabbit collecting duct cells by the group of Naray-Fejes Toth [8,16,18,19]: they demonstrated clearly that the enzyme was NAD, not NADP dependent in collecting duct cells isolated from the rabbit kidney. In proximal tubule cell lysates, ll-HSD was more NADP than NAD-dependent [16]. Using an histochemical method, Mercer et al. [20] found an NAD-dependency of the rat distal nephron enzyme. However, the extend of this metabolic activity could not be quantified by this approach. One aim of our experiments was to measure the cofactor dependency of the rat ll-HSD in collecting duct and proximal tubule, in conditions of maximal activity of the enzyme (i.e. at saturing concentrations of substrate). Finally, little is known about possible regulation of this MR-related ll-HSD, despite its clear involvment in some forms of human hypertension, such as the syndrome of apparent mineralocorticoid excess, or liquorice intoxication [3,11,13,21]. The goal of this study was to examine, within the rat kidney, some of the enzymatic properties of ll-HSD in intact cells of the collecting duct and proximal tubule, and to know whether the enzyme could be influenced by corticosteroid treatments of the animals. Results show that the enzyme differs in the aldosterone-sensitive collecting duct and in the aldosterone-insensitive proximal tubule by its abundance (V,,,), much higher in the collecting duct, its cofactor dependence (on NAD in the collecting duct, on NADP in the proximal tubule) and the ability of the proximal enzyme to function either as an oxidase or a reductase, by contrast with the collecting tubule, which exhibits an exclusive oxidase activity. The wide range of activation of the enzyme by corticosterone, together with enzyme kinetics, suggests either two forms of ll-HSD in CCD, with different affinities, or complex kinetics. Finally

neither aldosterone nor glucocorticoids seem to be implicated in the regulation of the enzyme, in both proximal tubule and collecting duct.

2. Material and methods

2.1. Animal treatments Experiments were performed on adult female Wistar rats (150-200 g body wt.). Animals were fed a standard diet and had free access to tap water. Adrenalectomy (ADX) was performed after ether anaesthesia. ADX rats had access to 0.9% NaCl for drinking. Corticosteroid treatments were performed three days after ADX. Acute treatments consisted of a single i.p. injection of aldosterone or dexamethasone (100 pg/lOO g body wt.) 2 h before sacrifice. Other ADX rats had an i.p. implantation of an osmotic minipump (Alzet 2001) delivering 10 pg/lOO g body wt./day aldosterone or dexamethasone or corticosterone for 70 h (3 days treatment) or 96 h (4 days treatment). 2.2. Obtention

of tubular segments

Rats were anaesthetized with pentobarbital sodium (5 mg/lOO g body wt.) and perfused via the aorta with an ice-cold perfusion solution (containing, in millimolar concentration: NaCl, 137; KCl, 5; MgSO,, 0.8; Na,HPO,, 0.33; KH,PO,, 0.44; MgCl,, 1; CaCl,, 1; D-Glucose,5; Tris-HCl, 10; pH 7.4) followed by perfusion of collagenase solution (similar solution to which O,l% collagenase (Serva) 0.7 U/mg) was added. At the end of perfusion, kidneys were removed and thin pyramid pieces were incubated at 30°C for 1 h in the same collagenase solution. Microdissection was performed at 4°C in 4 ml of microdissection solution (similar to the perfusion solution except for the absence of collagenase and addition of 0.1% bovine serum albumin (BSA, Sigma Chemical Co., St. Louis, MO, USA)). Proximal tubules in their final part (Pars Recta -PR) and Cortical Collecting Ducts (CCD) were isolated under stereomicroscope as previously described [22]. Tubular length was measured using a millimeter scale placed under the microdissection dish. 2.3. II-HSD

assay

For ll-HSD assay, pools of 3 mm PR and CCD (2-3 tubular segments in each) were prepared and transferred in a minimal volume (1 1.~11into the 5 ~1 incubation solution tube. Incubation solution was identical to the microdissection solution, to which [ 3H]corticosterone (2.70 TBq/mmol, Amersham) or 3H-ll-dehydrocorticosterone (prepared in the laboratory as described below) was added. The cofactor dependence of ll-HSD was examined as

N. Alfaidy et al. / Biochimica et Biophysics Acta 1243 (1995) 461-468

follows: 1 mM NAD or NADP (for assay of dehydrogenase activity), or NADH or NADPH (for assay of reductase activity) was added to the incubation medium of either intact tubules or lrubules permeabilized by three successive freezing-thawing steps. Cofactors were from Boehringer (Mannheim, Germany). Incubation of tubules was performed at 37°C for 10 min, except in the experiments designed to examine the time-course of ll-HSD activity. The reaction was stopped by addition of 95 ~1 mobile phase of HPLC (methanol/H,O:l/l) containing 10e4 M unlabelled corticosterone and ll-dehydrocorticosterone (Sigma) as inlernal standards. Samples were stored at -20°C up to HPLC analysis. In order to examine tlhe concentration dependence of ll-HSD, 3H-corticosterone was used as substrate, at concentrations ranging from 0.1 to 5000 nM (with dilution of specific activity with unla’belled corticosterone for concentrations above 100 nM). In experiments designed to explore possible hormonal re:gulation of dehydrogenase activity, tubules were incubated with 1000 nM [3H]corticosterone. To study reductase activity, tubules were incubated with 1000 nM 3H-ll-dehydrNocorticosterone ([3~]-ii-~~~), repared as previously described [23]. In order to obtain P H-ll-DHC, the substrate of the reductase reaction, rat kidney slices were incubated with 3H-corticosterone for 2 h at 37°C. The incubation medium was then extracted twice with ethyl-acetate, dried and dissolved in methanol. Samples were injected into the HPLC system, and fractions corresponding to [3~]-ii-~~~ were collected and concentrated. The resultant fraction contained > 90% ‘Hll-DHC, about 5% [3H]corticosterone, and about 4% unknown metabolites. 2.4. HPLC analysis A CN reversed-phase column ( PBondapak CN, Waters Associates, Milford, MA., USA) with precolumn (C,,, 5 pm, SociCtC Frangaise de Chromatographie, Paris, France) was used. Samples were injected onto the column and eluted isocratically with the mobile phase (methanol/H,O:l/l), 1 ml/min, with the elution profile of unlabelled standards rnonitored by absorbency at 240 nm (Beckman Gold HPLC System, Beckman Instruments, Gagny, France). Fractions of 0.5 ml were collected every 30 seconds (Retriever III, ISCO, Lincoln, Nebraska, USA) in counting vials containing 3 ml scintillation fluid (Optiphase; Packard Instruments, Downes Grove, Illinois, USA) for 10 min. The radioactivity was counted in liquid scintillation counter (Rackbeta; LKB Instruments, Inc, Gaitherburg, MD, USA). Thus, for each sample, the profile of radioactivity was superimposed to the elution profile of unlabeled corticosterone and ll-DHC. Results are expressed as femtomoles of the metabolite produced per 3 mm and per 10 min (fmol/3 mm/l0 min). Data are given as mean f SEM. Variance analysis (ANOVA) was per-

formed to compared steroid status.

ll-HSD

463

activity following

changes in

3. Results In order to establish adequate conditions of measurement of 1 l-HSD catalytic activity on intact isolated tubules, the time-dependence of the reaction was examined on CCD, after incubation with a low concentration (0.6 nM) of the substrate [3H]corticosterone. Fig. 1 shows that the reaction increased linearly with incubation time, from 5 to 30 min. The assay condition chosen (3 mm tubule in 5 ~1 incubation solution) allows to be in conditions of initial velocity for at least 30 min. All further determinations of ll-HSD activity have been performed at 10 min incubation time. Since it has been shown [16,17,20] that the presence and nature of some cofactors is critical for the functioning of ll-HSD, we have evaluated the effect of NAD and NADP, when added to intact and permeabilized isolated tubules incubated with 1000 nM 3H-corticosterone. Fig. 2 illustrates the lack of effect of these cofactors on ll-HSD dehydrogenase activity in intact PR and CCD: no change in the enzyme activity was observed when NAD, NADP, or both were added or not to intact PR or CCD. Permeabilization of tubules did not modify ll-HSD activity in PR, in the absence of cofactors added to the medium, whereas this manoeuvre reduced ll-HSD activity in the CCD. Addition of NADP, not NAD, increased values by about 100% in PR, whereas the reverse cofactor specificity was observed in CCD, i.e. an effect of NAD, not NADP. Addition of both cofactors did not result in a further increase, as compared to the respective effects of NADP in PR and NAD in CCD. We have also searched for the presence of reductase activity in PR and CCD, and the effects of cofactor addition on this activity in intact or permeabilized tubules

10 20 INCUBATION

30

40

TIME

(min)

Fig. 1. Time course of dehydrogenase activity (conversion of [3H]corticosterone into [3H]-ll-dehydrocorticosterone-11DHC) in isolated cortical collecting tubules. Tubules (3 mm tubular length per assay) were incubated in 5 pl with [3H]corticosterone (0.6 nM) at 37”C, for various times (S-30 min). Dehydrogenase activity increased linearly with time.

464

N. Alfaidy et al./Biochimica

et Biophysics Acta 1243 (1995) 461-468

(Fig. 3). Indeed, it has been described that ll-HSD can function as dehydrogenase or as reductase, depending on the cell context [13]. Reductase activity of PR and CCD was explored using [ 3H]-1 1 -dehydrocorticosterone ( 10e6 M) as substrate. No reductase activity was found in intact or permeabilized CCD, even in the presence of NADH or NADPH in the incubation medium. In contrast, reductase activity was found to be present in the Pars Recta (Fig. 3): cofactor addition allowed to measure a clear reductase activity in intact PR (36 k 5 fmol/3 mm/l0 min, n = 6, in the presence of NADH, and 58 + 5, n = 9, in the presence of NADPH). When PR were permeabilized, similar results were obtained, i.e. a mixed cofactor specificity of reductase activity. Thus, it appears that the CCD enzyme functions exclusively as dehydrogenase and is NAD-dependent whereas the PR enzyme exhibits both dehydrogenase (with NADP-dependence) and reductase (depending on both NADH and NADPH) activity. We next examined the dose dependence of ll-HSD dehydrogenase activity in the mineralocorticoid-sensitive CCD, as compared to that in the mineralocorticoid-insensitive PR (Fig. 4). Plateau values of ll-HSD were strikingly

IPR

%

&

200

PERMEABILIZED Fig. 3. Cofactor dependence of reductase activity in PR. Catalytic activity was measured in intact and permeabilized PR (3 mm tubular length per assay), in absence (control) or in presence of 1 mM NADH or NADPH, and with 10e6 M [“HI-11-dehydrocorticosterone as substrate. Incubation was at 37°C for 10 min. In both intact and permeabilized PR, reductase activity was barely detectable in the absence of exogenous cofactors (control). Addition of cofactors resulted in a clear reductase activity (higher with NADPH than with NADH), at levels close to that of dehydrogenase activity (see Fig. 2). n = number of tubular samples. Data are mean + SE.

t

14

13

12

INTACT

7

10. 10

10

4

PERMEABILIZED

Fig. 2. Cofactor dependence of dehydrogenase activity in PR and CCD. Catalytic activity was measured in intact and permeabilized tubules (3 mm tubular length per assay), in absence (control) or in presence of 1 mM NAD, NADP or both, and with 10m6 M [3H]corticosterone as substrate. Incubation was at 37°C for 10 min. Cofactor addition did not affect enzyme activity of intact tubules. The enzyme appeared to be NADP-dependent in permeabilized PR, and NAD-dependent in permeabilized CCD. n = number of tubular samples. Data are mean f SE.

different in PR and CCD: maximal level of enzyme activity was 8-10 fold higher in CCD than in PR. Saturation of the enzyme was apparent only at corticosterone concentrations above 1000 nM. On the other hand, insets from Fig. 4 show that ll-HSD activity can be readily detected even at low corticosterone concentrations (0.1-50 nM). While no clear difference in the overall apparent affinity for corticosterone can be evidenced between ll-HSD of PR and CCD (with an apparent K,, of about 250 nM in each case), the very wide range of concentration dependence suggests that the enzyme activation by corticosterone could in fact correspond to complex kinetics, possibly reflecting the presence of more than one enzymatic form. To explore this hypothesis, analyses were performed separately for the lower concentration range (< 50 nM) and for data obtained at 200-1000 nM. Fig. 5 is Lineweaver-Burk plots of these two sets of data obtained in CCD. It appears that kinetic parameters differ at low and high corticosterone concentration (K,,: 65 nM, Vmax:39 fmol/3 mm/l0 min

N. Alfaidy et al. / Biochimica et Biophysics Acta 1243 (1995) 461-468

3

ADX

465

ALDO

DEX

I’\\.. ,‘,’ ,,,

II #‘ \\\,‘,’ I,,

/‘#‘#’

0

250

so0

750

looo

”

I’,‘,’

5ooo

/‘#‘#’ /‘,‘#’ ,‘#LN ,‘\\\ ,‘,’

CORTICOSTRRONE CONCENTRATION ( nM )

Fig. 4. Concentration dependence of ll-HSD dehydrogenase activity in PR and CCD. Intact tubules (3 mm tubular length per assay) were incubated at 37°C for 10 min in the presence of [3H]corticosterone (0.1 to 5000 nM). Insets illustrate the enzyme activity at low substrate concentration (0.1-50 nM), in PR and CCD. Saturation occured only at high corticosterone concentration (1000 nM). V,,, was S- to lo-times higher in CCD than in PR, without evidence for differences in overall apparent affinity. Data include 103 individual experimental points for CCD, and 94 for PR, dissected from 11 different rats.

- panel A - and K,,: 544 nM, Vmax:526 fmol/3 mm/l0 min - panel B-respectively). The NAD dependence of the CCD enzyme, which was observed at high corticosterone concentration (see above) was also evaluated in permeabilized CCD incubated with low (5 nM) corticosterone concentration: results show a clear exclusive NAD dependence (without cofactor: 5.11 k 0.24 fmol/3 mm/l0 min; in the presence of NAD: 12.78 k 0.65; in the presence of NADP: 4.25 + 0.17). Because ll-HSD plays a critical role in the protection of MR, the possibility that changes in plasma levels of aldosterone or glucocorticoids could intervene in modulating the level of enzyme activity was examined. As an attempt to mimic situations such as stress where endogenous corticosteroids vary acutely, the short-time effect on ll-HSD activity of an aNcute injection of a high dose of

‘\I\‘. I,,

,‘,‘#’ #L’ \\\ ,’ I,,

,‘,‘#’

,‘.\\ ,‘#’

41 ADX

22

30

ALDO

DEX

Fig. 6. Lack of effect of acute corticosteroid injections on ll-HSD dehydrogenase activity in PR and CCD. Adrenalectomized rats were injected with 100 pg/lOO g body wt. of aldosterone (ALDO) dexamethasone (DEX) or solvant (ADX) 2 h before sacrifice. Isolated intact tubules were incubated with 10m6 M [3H]corticosterone as substrate. ll-HSD activity did not differ between groups, in PR as well as in CCD. n = number of tubular samples. Data are mean f SE. Values do not differ significantly between treatments.

corticosteroids (aldosterone or dexamethasone) was determined in CCD and PR from ADX rats. As shown on Fig. 6, the V,,x of ll-HSD activity was not modified either by aldosterone or by dexamethasone, in both tubular epithelia. We have also examined the effects of more prolonged administration of corticosteroids. To this purpose, adrenalectomized rats were continuously exposed to physiological doses of aldosterone, of the synthetic glucocorticoid dexamethasone, or of corticosterone, the natural substrate of the enzyme in the rat, for 3 or 4 days. As shown in Fig. 7, it appears clearly that these modifications

Z

B

0.007

El

L

2 l/B(xlOg)

4

6

g

0.005

m ‘Z

0.003

$! 0.001 >’ 0 5,

0 ~

0 0

0.001

0.003

0.005

l/B(xlO’)

Fig. 5. Lineweaver-Burk plots #ofdehydrogenase ll-HSD activity in CCD. Data from Fig. 4 were analysed separately in the 0.1-50 nM range (panel A) and in the 200-1000 nM range (Ipanel B). Points are individual values from Fig. 4, except at 1000 nM where mean value of the 46 determinations is given. Solid lines are the linear regression lines: panel A: y = 1.669x + 0.026, r = 0.94; pane1 B: y = 1.034.x + 0.002, r = 0.91. Calculated V,,, and K,,, are given in the text.

466

N. Alfaidy et al./Biochimica

et Biophysics Acta 1243 (1995) 461-468

tors. To this purpose, dehydrogenase activity was measured in PR and CCD from rats receiving aldosterone for 4 days. Tubules were permeabilized and incubated with NADP (for PR) or NAD (for CCD) and lop6 M [3H]corticosterone. Table 1 shows the results of these experiments, as compared to those observed in intact tubules from the same animals. While the NAD or NADP dependence of the enzyme (in CCD and PR, respectively) was confirmed, no effect of animal treatment with aldosterone could be detected in permeabilized tubules, thus confirming the lack of effect of this treatment observed in intact tubules (Fig. 7).

ICCD

4. Discussion

C

ADX

ALDO 3d4d

DEX 3d4d

CORTICO 3d4d

Fig. 7. Lack of effect of continuous infusion of corticosteroids on ll-HSD dehydrogenase activity in PR and CCD. Osmotic minipumps delivering 10 pg/lOO g body wt./day of aldosterone (ALDO), dexamethasone (DEX) or corticosterone (CORTICO) were inserted into the peritoneal cavity of adrenalectomized rats (ADX) for 3 or 4 days. C: intact control rats. Isolated intact tubules were incubated with 10e6 M 3H-corticosterone as substrate. ll-HSD activity did not differ significantly between groups (Dunnett t test) for either PR or CCD. n = number of tubular samples. (ANOVA: PR: F = 12.168, p = 0.001, CCD: F = 1.983, p = 0.0625, without significance between groups. Dunnett t test for PR as well as for CCD).

of corticosteroid status did not influence the V,,, of ll-HSD, in CCD as well as in PR. In addition, we have questioned whether some regulation by corticosteroid hormones may be apparent only in permeabilized tubules incubated in the presence of cofac-

Table 1 Lack of effect of aldosterone lized tubules

treatment on ll-HSD

Permeabilized

Intact

Aldosterone Control rats

treated rats

activity in permeabi-

CCD + NAD

CCD

PR

476 + 23 n= 10 434*19 n = 14

50 + 4 754 + 72 n= 16 n=13 48+4 729+30 n=13 n=lO

PR + NADP 109+9 n=13 116+12 n = 10

Dehydrogenase activity was mesured in tubules from adrenalectomized rats treated for four days with 10 pg/lOOg bw/day of aldosterone. ll-HSD was determined in permeabilized tubules in the presence of 1 mM NAD (for CCD), or NADP (for PR), and 10m6 M 3H-corticosterone as substrate. Results are given as femtomoles/3 mm/l0 min of 3H-dehydrocorticosterone produced. For comparison purposes, data obtained in control rats are recalled (data from Fig. 2). Aldosterone treatment did not affect ll-HSD activity in intact as well as permeabilized tubules. n = number of tubular samples. Data are mean f SE.

The aim of these experiments was to examine whether the functional characteristics of the enzyme ll-HSD could differ in an aldosterone sensitive tubular epithelium, the collecting duct, and an aldosterone insensitive tubular epithelium, the proximal tubule of the rat kidney, and whether enzyme activity could be regulated by corticosteroid hormones. Several lines of evidences support the notion that at least two forms of the enzyme could be present in the kidney: one corresponding to the MR protector enzyme, predominant in the collecting tubule, and another one, putatively involved in some regulatory processes of glucocorticoid action, which could be located in the aldosterone-insensitive proximal tubule. Arguments in favour of this hypothesis are (1) the detection of several forms of the enzyme in the kidney, at the mRNA [24,25] or protein [12] level, (2) the contradiction between the proximal immunolocalisation of the enzyme (with an antibody directed against the hepatic enzyme) [1,9,10] and the large prevalence of ll-HSD catalytic activity in the collecting duct [4-6,201 and (3) the reported presence of a NAD dependent ll-HSD activity in the kidney [16,17,20,26], as well as in other organs such as placenta [27], contrasting with the NADP-dependence of the hepatic enzyme [28,29]. The present results clearly demonstrate that, within the rat kidney, ll-HSD in the proximal and the collecting tubule differs by its cofactor dependence: in permeabilized tubules, enzyme activity was found to be NAD-dependent in CCD, and NADP-dependent in PR. These findings are at variance with those of Walker et al. [17], who found a mixed cofactor dependence of ll-HSD in both proximal and distal tubular preparations. This mixed cofactor dependence may be attributed to reciprocal contaminations of the preparations, as pointed out by the authors. The NAD dependence of the rat CCD enzyme is in accordance with previous reports of the group of Naray-Fejes-Toth [16,19] on cortical collecting duct cell lysates of the rabbit kidney. These authors also described a low dehydrogenase activity in the proximal tubular cells, dependent on both NAD and

N. Alfaidy et al. /Biochimica

et Biophysics Acta 1243 (1995) 461-468

NADP, possibly due to a few contaminant distal cells, as noted by the authors. Thus, both species (rat and rabbit) exhibit a highly active ll-HSD in their collecting duct cells. However, they differ in the amount of enzyme expressed in the proximal tubule (higher catalytic activity in the rat than in the rabbit). Of interest, cofactors did not modulate dehydrogenase activity when added to intact tubules, as already noticed for NADP [4,29]. This is most likely due to sufficient intracellular stores of NAD and NADP in these intact cells. As a matter of fact, cell permeabilization appears to be a very useful tool to determine the cofactor specificity since it allows addition of one single cofactor on cell preparations in which a depletion in endogenous cofactors is likely to occur as a consequence of preparation procedures. This depletion is probably of variable extent, depending on experimental procedures: of small magnitude in our study on permeabilized fresh tubules and much more pronounced in the study of Rusvai and Naray-Fejes-Toth [16] on microsomal fractions of immunoselected cortical collecting cells. Since ll-HSD can function in a bidirectional way, we have tested the reductase activity in rat CCD and PR. In CCD, no reductase activity was detectable, in both intact tubules and tubules permeabilized in the presence of cofactors, a result in agreement with those of Rusvai and Naray-Fejes-Toth [16] obtained on rabbit tubular cells. In rat PR, reductase activity was undetectable in the absence of exogenous cofactors. However, addition of NADH or NADPH allowed to recover reductase activity, at a level equivalent to that of dehydrogenase activity. The need of cofactor addition to detect reductase activity could be due to the instability of NADH and NADPH, resulting in difficulties to recover reductase activity in vitro, as reported by Lakshmi and al. 1301. The dose-response study was performed in intact tubules. Another possibility would have consisted in the use of permeabilized tubules, to estimate kinetic parameters in the presence of saturating levels of cofactors. Indeed, in this condition, the apparent V,,, (as measured at lo-” M corticosterone) was higher than that observed in intact tubules, as shown on Fig. 2. The rationale for using intact tubules was to explore the enzyme in conditions of normal cell environment, leading to estimates of apparent Kd and V,,, in the specific tubular cell context. Whether permeabilization process and addition of exogenous cofactors are susceptible to modify the apparent K, remains to be determined. However, this appears unlikely, since the group of Naray-Fejes-Toth reported close K, values in intact CCD [8] and in CCD microsomial preparations, in the presence of NAD [16]. Results show that in both PR and CCD saturation of 111-HSD activity occurred for corticosterone concentration of about 1000 nM. V,,, were strikingly different in these two tubular types since it was 8-10 fold higher in CCD than in PR. The V,,, of the rat CCD (400-450 femtomoles per 10 min per 3 mm tubule,

467

i.e. per 1000-1500 cells) is a little higher than that reported for the rabbit CCD enzyme [8]: 170 femtomoles per 10 min per 1000 cells. Of particular interest is the very wide range of enzyme activity, from the nanomolar to the micromolar level. The association of a high V,,, and a large range of activity in CCD is in very good accordance with the proposed mineralocorticoid receptor protecting function of the enzyme at both low or very high glucocorticoid concentration, such as those which can occur during stress. However, one can wonder whether these kinetics correspond to one single enzyme or to the combination of several enzymatic forms. Indeed, the expression of several forms of ll-HSD has been reported in the kidney [12,24,25]. In addition, previous reports [26,31,32] have mentioned that unpurified microsomal ll-HSD showed curvilinear Eadie plots, whereas homogeneous enzyme gave rectilinear plots. This observation is compatible with either more than one enzyme, or complex kinetics. One of these form should display high affinity for glucocorticoid hormones, in order to exclude completely glucocorticoids from the mineralocorticoid receptor in the aldosterone-sensitive distal nephron. Thus we have focused our analysis on enzyme activity at low corticosterone concentration (0.1-50 nM) as compared to higher substrate concentration (200-1000 nM). Lineweaver-Burk plots of enzyme activity in the CCD support the possibility of coexistence of two forms of enzyme, differing in affinity for corticosterone by an order of magnitude. The higher affinity form (K,,, = 65 nM) of the rat CCD form is in agreement with that reported by the group of Naray-FejesToth et al. [8,16,18] in rabbit CCD immunoselected cells, and Brown et al. [27] in human placenta. The exclusive NAD dependence of these two putative CCD forms differentiates them from the NADP dependent enzyme of PR. Differences in cofactor dependence might reflect adequation to specific cell environment and metabolic pathways. Along this line, it is noticeable that a number of metabolic pathways (glycolysis, neoglucogenesis, fatty acid and ketone body metabolism) are NAD-dependent in CCD and PR [33]. In contrast, the NADP-dependent pentose phosphate pathway is active only in the proximal tubule [33]. Since ll-HSD intervenes in the expression of mineralocorticoid and glucocorticoid effects, it was of interest to examine whether corticosteroid hormones could regulate ll-HSD activity. Few studies addressed this question in epithelial cells [13,34-361. Gaeggeler et al. [13] found that ll-HSD is induced by corticosterone in a toad bladder cell line. Smith et al. [36] dit not detect any effect on ll-HSD as evaluated on whole kidney. However, this study could not eliminate the possibility that corticosteroid regulation could be restricted to a particular cell population. For instance, a regulation affecting only the cortical collecting tubule, which represents about 5% of the renal cell population, could be undetectable by experiments on whole kidney cortex. This is why we have performed a series of experiments on isolated PR or CCD. Neither adrenalec-

N. Alfaidy et al. /Biochimica

468

et Biophysics Acta 1243 (1995) 461-468

tomy nor administration of aldosterone, corticosterone or dexamethasone delivery to ADX rats for 2 h, 3 days or 4 days provoke significant changes of ll-HSD activity in intact PR or CCD. In addition, experiments on intact tubules or in permeabilized tubules from aldosteronetreated rats (4 days), incubated in the presence of NAD (CCD) or NADP (PR), yielded values of dehydrogenase activity very similar to those obtained in previous series on control rats. Altogether, these results do not support the occurrence of ll-HSD regulation by corticosteroid hormones administration to ADX rats, at least in an acute or subacute manner. In conclusion, our results clearly demonstrate that llHSD differs in the aldosterone insensitive proximal tubule and in the aldosterone sensitive collecting duct of the rat kidney, by the abundance of the enzyme, 8-10 fold higher in the collecting tubule than in the proximal tubule, and cofactor dependence (on NADP in the proximal tubule, and on NAD in the collecting duct). In addition, the CCD enzyme displays exclusive dehydrogenase activity, contrasting with both dehydrogenase and reductase activity in PR. Within CCD, kinetics analyses suggest the coexistence of two forms of ll-HSD, that differ in affinity for corticosterone by a factor of ten. ll-HSD catalytic activity is not modified by changes in corticosteroid status, in PR as well as in CCD.

Acknowledgements We wish to thank V. L&&que for typing the manuscript.

References [I] Edwards, C.R.W., Stewart, P.M., Burt, D., Bret, L., McIntyre, M.A., Sutanto, W.S., de Kloet, E.R. and Monder, C. (1988) Lancet 2, 986-989. [2] Funder, J.W., Pearce, P.T., Smith, R. and Smith, Al. (1988) Science 242, 583-585. [3] Bonvalet, J.P. (1991) News Physiol. Sci. 6, 201-205. [4] Bonvalet, J.P., Doignon, I., Blot-Chabaud, M., Pradelles, P. and Farman, N. (19901 J. Clin. Invest. 86, 832-837. [5] Kenouch, S., Alfaidy, N., Bonvalet, J.P. and Farman, N. (1994) Steroids 59, 100-104. [6] Kenouch, S., Coutry, N., Farman, N. and Bonvalet, J.P. (1992) Kidney Int. 42, 56-60. [7] Farman, N. (1992) Semin. Nephrol. 12, 12-17.

[81 Naray-Fejes-Toth,

A., Watlington, CO. and Fejes-Toth, G. (19911 Endocrinology 129, 17-21. [91 Castello, R., Schwarting, R., Muller, C. and Hierholzer, K. (1989) Renal Physiol. Biochem. 12, 320-327. [lOI Rundle, S.E., Funder, J.W., Lakshmi, V. and Monder, C. (1989) Endocrinology 125, 1700-1704. illI Bonvalet, J.P. (1993) Exp. Nephrol. I, 334-338. [I21 Monder, C. and Lakshmi, V. (1990) Endocrinology 126,2435-2443. [I31 Seckl, J.R. (1993) Eur. J. Clin. Invest. 23, 589-601. [141 Stewart, P.M., Whorwood, C.B., Barber, P., Gregory, C., Monder, C., Franklyn, J.A. and Sheppard, M.C. (19911 Endocrinology 128, 2129-2135. [I51 Agarwal, A.K., Monder, C., Eckstein, B. and White, P.C. (1989) J. BioI. Chem. 264, 18939-18943. [I61 Rusvai, E. and Naray-Fejes-Toth, A. (1993) J. Biol. Chem. 268, 10717-10720. [I71 Walker, B.R., Campbell, J.C., Williams, B.C. and Edwards, C.R.W. (1992) Endocrinology 131, 970-972. I181 Naray-Fejes-Toth, A., Rusvai, E., Denault, D.L., St Germain, D.L. and Fejes-Toth, G. (1993) Am. J. Physiol. 265, F896-F900. I191 Naray-Fejes-Toth, A., Rusvai, E. and Fejes-Toth, G. (1994) Am. J. Physiol. 266, F76-F80. 130, 1201Mercer, W.R. and Krozowski, Z.S. (1992) Endocrinology 540-543. ml Ulick, S., Levine, L.S., Gunczler, P., Zanconato, G., Ramirez, L.C., Rauh, W., Rosler, A., Bradlow, H.L. and New, Ml. (1979) J. Clin. Endocrinol. Metab. 49, 757-764. D21 Farman, N., Vandewalle, A. and Bonvalet, J.P. (1982) Am. J. Physiol. 242, F69-F77. 1231 Duperrex, H., Kenouch, S., Gaeggeler, H.P., Seckl, J.R., Edwards, C.R.W., Farman, N. and Rossier, B.C. (1993) Endocrinology 132, 612-619. [241 Krozowski, Z., Obeyesekere, V., Smith, R. and Mercer, W. (1992) J. Biol. Chem. 267, 2569-2574. 1251 Krozowski, Z., Stuchbery, S., White, P., Monder, C. and Funder, J. (1990) Endocrinology 127, 3009-3013. [261 Siebe, H., Baude, G., Lichtenstein, I., Wang, D., Buhler, H., Hoyer, G.A. and Hierholzer, K. (1993) Renal Physiol. Biochem. 16, 79-88. 1271Brown, R.W., Chapman, K.E., Edwards, C.R.W. and Seckl, J.R. (1993) Endocrinology 132, 2614-2621. D81 Monder, C., Lakshmi, V. and Miroff, Y. (1991) Biochim. Biophys. Acta 1115, 23-29. D91 Monder, C., Stewart, P.M., Lakshmi, V., Valentino, R., Burt, D. and Edwards, C.R.W. (1989) Endocrinology 125, 1046-1053. [301 Lakshmi, V. and Monder, C. (1985) Endocrinology 116, 552-560. [311 Lakshmi, V. and Monder, C. (1988) Endocrinology 123,2390-2398. [321 Monder, C. and Laksmi, V. (1989) J. Steroid Biochem. 32, 77-83. I331 Wirthensohn, G. and Guder, W.G. (1986) Physiol. Rev. 66,469-497. I341 Gaeggeler, H.P., Duperrex, H., Hautier, S. and Rossier, B.C. (1993) Am. J. Physiol. 264, C317-C322. [351 Low, S.C., Assad, S.N., Rajan, V., Chapman, K.E., Edwards, C.R.W. and Seckl, J.R. (1993) Endocrinology 139, 27-35. [361 Smith, R.E. and Funder, J.W. (1991) J. Steroid Biochem. Molec. Biol. 38, 265-267.