Na+-dependent biotin transport into brush-border membrane vesicles from rat kidney BARBARA BAUR, HUGO WICK, Metabolic Unit, University Children’s
AND E. REGULA BAUMGARTNER Hospital, CH-4058 Base& Switzerland
BAUR,BARBARA,HUGOWICK,AND E. REGULABAUMGARTNER. Na’-dependent biotin transport into brush-border membrane vesicles from rat kidney. Am. J. Physiol. 258 (Renal Fluid
Electrolyte Physiol. 27): F840-F847, 1990.-The mechanisms of biotin reabsorption in rat kidney cortex were investigated using isolated brush-border membrane vesicles. An inwardly directed Na+ gradient specifically stimulated a transient biotin overshoot. Biotin transport was not affected by a valinomycininduced K+-diffusion potential, and biotin--Na+ stoichiometry was found to be 1: 1. As a function of concentration, the uptake showed saturation in the presence of a Na+ gradient with an apparent Michaelis constant (I&) of 55 PM and Vmax of 217 pm01 . mg protein-‘. 25 s-l. Desthiobiotin, 250 PM, norbiotin, bisnorbiotin, thioctic acid, Valerie acid, probenecid, and nonanoic acid inhibited the transport of 30 PM biotin, whereas other biotin derivatives, as well as biocytin and organic acids found in the urine of biotinidase-deficient patients, did not. Preloading of the vesicles with biotin, desthiobiotin, norbiotin, and thioctic acid in the presence of Na+ increased initial uptake of biotin from the incubation medium (trans-stimulation). Our results indicate that biotin absorption in rat kidney fulfills the criteria for a specific carrier-mediated and electroneutral Na+biotin- cotransport in a 1: 1 ratio. The results are discussed in context with congenital biotinidase deficiency in humans. biotinidase deficiency; electroneutrality; trans-stimulation
renal absorption;
THE PHYSIOLOGICAL IMPORTANCE of biotin as a covalently bound prosthetic group for carboxylation reactions is well established. In mammals all four biotin-dependent carboxylases (pyruvate carboxylase, acetyl-CoA carboxylase, propionyl-CoA carboxylase, and 3-methylcrotonoyl-CoA carboxylase) play crucial roles in intermediary metabolism. Nutritional biotin deficiency and the recently discovered genetic disorders of biotin metabolism result in impaired carboxylase activities (2l), i.e., multiple carboxylase deficiency, leading to the accumulation of a typical pattern of abnormal organic acids in body fluids and to a serious metabolic derangement with possible lethal outcome (12). One cause of multiple carboxylase deficiency is biotinidase deficiency (23). The enzyme biotinidase (EC 3.5.1.12) cleaves biotin from biocytin, the product of metabolic degradation of biotin-dependent carboxylases. Patients suffering from biotinidase deficiency are not able to recycle endogenous biotin or to process proteinbound biotin, the main form of this vitamin in the diet. Biotin is lost in the urine in the form of biocytin, which F840
results in progressive biotin depletion in early infancy or childhood. Previous investigations in our laboratory have shown that the high requirement for free biotin in biotinidasedeficient patients is not only the result of defective recycling but also of increased renal excretion of biotin itself (2, 20). Clearance studies revealed that, at normal biotin plasma concentration, renal biotin clearance in patients was equal to or slightly higher than the creatinine clearance. In healthy children and adults biotin clearance was 40% of the creatinine clearance, approximating creatinine clearance with increasing plasma biotin concentrations up to 50 times normal (1). These results suggest a saturation phenomenon either at the site of glomerular filtration attributable to a biotinbinding protein in plasma and/or of tubular reabsorption of biotin. So far little is known about the renal handling of biotin or how biotin is transported in the blood. In a recent report biotinidase itself was identified as a biotintransporti .ng protein in pla .sma (7). We focused our interest on the mechanisms of renal biotin reabsorption and, particularly, whether it can be influenced by biocytin and possibly also by other abnormal metabolites. For our investigations we used brush-border membrane vesicles isolated from rat kidney cortex. To our knowledge there are only two publications concerning biotin reabsorption in kidney. Podevin et al. (13) described an active Na+-dependent electrogenic biotin transport into brush-border membrane vesicles from rabbit kidney, and recently Spencer and Roth (19) reported an energy-dependent and saturable biotin transport in brush-border membrane vesicles from rat kidney. More data are available on intestinal absorption of biotin observed by use of several in vivo and in vitro techniques (3, 6, 8, 15-17).
In the present study we confirm the existence of a carrier-m .ediated, Na+-dependent biotin transport system in brush-border membrane vesicles isolated from rat kidney cortex and characterize some of its properties in more detail. MATERIALS
AND METHODS
Materials. [8,9-3H] biotin (30-50 Ci/mmol) was purchased from Du Pont de Nemours (Zurich, Switzerland). Unlabeled biotin, desthiobiotin, diaminobiotin, biotin methyl ester, and biocytin were from Sigma Chemical (Poole, Dorset, UK). Biotin-D-sulfoxide, biotin+sulfoxide, and biotin sulfone were gifts from Dr. Bowers-
0363-6127/90 $1.50 Copyright 0 1990 the American Physiological
Society
Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 15, 2019.
BIOTIN
TRANSPORT
IN
RENAL
BRUSH-BORDER
Komro (Woodruff Medical Center, Atlanta, GA). Norbiotin was a gift from Dr. J. Bausch (HoffmannLaRoche, Basel, Switzerland), and bisnorbiotin was synthesized in 95% purity by Dr. W. Holick (HoffmannLaRoche). Thioctic acid was obtained from Merck (Zurich, Switzerland). Probenecid, nonanoic acid, and N-benzoyl-&aminovaleric acid were from Sigma. 3Methylcrotonic acid was obtained from Fluka (Buchs, Switzerland). 3-OH-isovalerate (ammonium salt) was a gift from Dr. C. Bachmann (Inselspital, Bern, Switzerland). Rialuma scintillation cocktail was purchased from Lumac (Landgraaf, The Netherlands). All other chemicals were obtained from commercial sources and were of analytical grade. Methods. Brush-border membrane vesicles from rat kidney cortex slices (Wistar rats) were prepared by an ethylene glycol-bis(P-aminoethyl ether)-N,N,N’,N’-tetraacetic acid (EGTA)-magnesium precipitation method as described by Biber et al. (4) with slight modifications. The rats were killed by an overdose of chloroform. The abdomen was opened, and the kidneys were removed. After removal of the capsule, the cortex slices were dissected with a razor blade. The slices were homogenized with a polytron (PT 10-35, Kinematica, Kriens, Switzerland) on setting 7 for 2 min. The homogenate was subjected to magnesium precipitation and differential centrifugation. The high-speed centrifugation steps were performed at 18,000 revolutions/min (rpm) and 16,000 rpm for 30 min in a Sorvall RC-5C centrifuge with a SS34 rotor. After the purification the vesicles were loaded with buffers as indicated in the legends to Figs. l-8 and Table 1. The final pellet was suspended in the desired buffer and rehomogenized with a glass-Teflon potter. Then the membranes were collected by centrifugation (20,000 rpm) and resuspended at a concentration of 510 mg/ml in an appropriate volume of the final buffer by repeated suction through a fine needle (gauge 0.5 mm). After equilibration of the vesicles for 90 min at room temperature, transport measurements were performed as follows. The uptake of [3H]biotin was started by adding an aliquot (lo-20 ~1) of the vesicle suspension to the incubation medium (100-200 ~1) containing tritiated biotin. It was terminated by the addition of 4 ml icecold stop solution containing 90 mM mannitol, 100 mM NaCl, 10 mM N-2-hydroxyethyl piperazine-N’-2-ethanesulfonic acid and tris( hydroxymethyl)aminomethane (HEPES-Tris), pH 7.4, to the reaction mixture and subsequent rapid filtration through nitrocellulose filters (Millipore, HAWP 02412, pore size 0.45 pm) kept under suction. The filters were washed with another 4 ml of chilled stop solution and the retained radioactivity was counted in 10 ml Rialuma by use of a Tri-Carb 4530 scintillation counter (Packard Instrument, Downers Grove, IL). Values for the nonspecific binding of C3H]biotin (0.02% of the total radioactivity) on the filters were determined by filtering a reaction mixture without vesicles; those values were then subtracted from the incubated samples. Nonspecific retention remained unchanged in the presence of 0.1 mM biotin or biotin-D/Lsulfoxide in the stop solution. All incubations were carried out at 25°C and at least
MEMBRANE
VESICLES
FE341
in triplicate with freshly prepared membrane vesicles. Each experiment was repeated at least three times. Although the quantitative uptake of [3H]biotin varied from one preparation to another, the qualitative results were in good agreement among the different experiments. Therefore the data of individual experiments are presented in most instances. The results are generally expressed as femtomoles or picomoles per milligram protein t SD or as percentage t SE of the mean. Unless indicated, SD or SE values do not exceed the size of the symbols in the figures. The purity of the isolated brush-border membrane vesicles was routinely checked by measuring the enrichment of the brush-border marker enzyme, alkaline phosphatase. The enrichment was lo- to &fold with respect to the starting homogenate. Alkaline phosphatase activity was assayed as described by Bowers and McComb (5) using an automated analysis kit (Boehringer Mannheim, Mannheim, FRG). Biotinidase activity was measured according to Wastell et al. (22). Activity was found to be -40-100 pmol . min-l . mg protein-l in the homogenate and below detection limits (-0.1 nmol/min) in the brush-border membrane fraction. Protein was determined using the Bio-Rad reagent with bovine plasma albumin as a standard (Bio-Rad, Richmond, CA). To confirm the identity of the transported biotin, vesicles were incubated for 90 min at 25°C in a medium containing 0.5 PM [3H]biotin, 100 mM mannitol, 100 mM NaCl, and 50 mM HEPES-Tris (pH 7.4). The membranes were collected by filtration and the radioactivity was extracted as described by Podevin et al. (13). Fifty microliters of the extracts were spotted on thinlayer chromatography aluminum sheets with si.lica ge 160 (Merck). The following solvent systems were used: butanol-acetic acid-water (4:1:5; upper phase) and benzene-methanol-acetone-acetic acid (14:4:1: 1).With both systems >93% of the radioactivity comigrated with the biotin standard (Rf = 0.71 and 0.51,respectively, where Rf is the migration relative to the solvent front). RESULTS
Time course and Na+ specificity. Biotin uptake as a function of time is shown in Fig. 1. In four experiments essentially the same results have been obtained. During the first 25-40 s biotin transport was apparently linear (Fig. 1, inset). Extrapolation of the initial linear transport rate to zero time corresponds nearly to zero uptake, indicating that there is no substantial binding to the membranes. An overshoot phenomenon was observed in the presence of an extravesicular-to-intravesicular N .a+ gradient. Maximal accumulation was reached within 46 min of incubation and was about twice the equilibrium value. Thereafter the uptake of biotin slowly diminished to equilibrium at 90 min of incubation, indicating efflux from the vesicles. None of the other monovalent cations tested (Li’, K+, choline+) was able to produce an overshoot effect when imposed as inwardly directed gradients with Cl- as the accompanying anion. The uptake in the presence of LiCl, KCl, and choline Cl was not signifi-
Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 15, 2019.
F842
BIOTIN
TRANSPORT
IN RENAL
30
BRUSH-BORDER
MEMBRANE
VESICLES
60
lime
(set)
I
,,,,,~,,~~~, 2
4
6
8
(Osmolar
10
12
it y)”
FIG. 2. Medium osmolarity. Renal brush-border membrane vesicles were loaded with 100 mM sucrose, 50 mM HEPES-Tris, pH 7.4. Uptake was started by adding 20 ~1 of vesicle suspension to 100 ~1 of incubation medium containing 100 mM NaCl, 20 nM [3H]biotin, 50 mM HEPESTris (pH 7.4), and increasing concentrations of sucrose (loo-1,000 mM). Incubation was performed for 90 min. Values are means +: SD of 1 experiment performed in triplicate. Relationship was linear according to Y = rnx + b, where n is slope, b is y-intercept, and r is correlation coefficient.
Time
(mid
1. Cation requirements. Renal brush-border membrane vesicles were resuspended in 300 mM mannitol, 50 mM HEPES-Tris, pH 7.4. At start of incubation, 20 ~1 of vesicle suspension were added to 100 ~1 of incubation medium containing the following: 300 mM mannitol, 50 mM HEPES-Tris (pH 7.4), and 20 nM [3H]biotin (A), or 100 mM mannitol, 50 mM HEPES-Tris (pH 7.4), 20 nM [3H]biotin, and 100 mM NaCl (0); 100 mM LiCl (0); 100 mM KC1 (w); or 100 mM choline Cl (cl). Values are means t SD for triplicate assays of 1 experiment. Inset: uptake of 20 nM [3H]biotin during first 60 s of incubation under NaCl conditions: values are means & SD of 1 experiment performed in quadruplicate. FIG.
cantly different from the biotin uptake in the absence of any salt gradient. After 90 min of incubation, the same equilibrium value was reached under all conditions. Effect of medium osmolarity. To ensure that the biotin accumulation represents transport into the vesicles’ inner space and not simple binding to the membranes, the effects of increasing medium osmolarity on biotin uptake were examined. Figure 2 shows that the equilibrium value of biotin uptake decreased with increasing sucrose concentration in the incubation medium. Extrapolating to zero, i.e., infinite osmolarity, reveals a binding component of 13.1 fmol/mg protein, which accounts for 20% of the equilibrium value under isotonic conditions. Membrane potential dependence. The influence of the electrical membrane potential was tested by eliciting a strong transient intravesicular potential. To this aim the vesicles were preloaded with 100 mM KC1 and valinomycin and subsequent transport measurements were performed in K+-free medium. As illustrated in Fig. 3, the inside negative membrane potential failed to stimulate both initial and maximal biotin uptake in the presence of an inwardly directed Na+ gradient (curue A vs. C). Curve D (control) shows the biotin uptake as a function of time when a Na+ gradient (out > in) was imposed in
A
-5 100
I
b k
3
P
.
A 0 Na’ K’ I3 0 Na’ K’ C Wla+ K' 0 c] Na’
out s in in > out, d/al out = in in + out, l Val outdn in D out, +EtOH out, in, l EtOH
T
4) < 50 4
.-E H m
Time
(mid
3. Membrane potential dependence. Vesicles were loaded with 300 mM mannitol, 50 mM HEPES-Tris, pH 7.4 (o), or with 100 mM mannitol, 100 mM KCl, 50 mM HEPES-Tris (pH 7.4) with (0) and without valinomycin (m), and with an additional 100 mM NaCl and valinomycin (0). Valinomycin, 1 ~1 [4 mg/ml ethanol (EtOH)], was added per 100 ~1 of vesicle suspension. To controls without valinomycin, ethanol was given in same amount. Incubation was started by adding 20 ~1 of vesicles to 100 ~1 of medium of the following composition: 100 mM NaCl, 20 nM [3H]biotin, 50 mM HEPES-Tris (pH 7.4), and 100 mM mannitol (0, n , IX); or 100 mM NaCl, 20 nM [3H]biotin, 50 mM HEPES-Tris (pH 7.4), and 300 mM mannitol(0). Values are means * SD of 1 experiment performed in triplicate. FIG.
Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 15, 2019.
BIOTIN
TRANSPORT
IN RENAL
BRUSH-BORDER
MEMBRANE
F843
VESICLES
the absence of K+. The presence of KC1 at the trans side of the vesicles had no significant influence (curve C vs. D) nor did valinomycin itself alter significantly the timedependent biotin uptake (data not shown). With Na+ present at equilibrium on both sides of the vesicle membrane, a K+ diffusion potential alone did not result in biotin accumulation above equilibrium as illustrated in curve B.
In a control experiment (data not shown) a valinomycin-induced K+ diffusion potential enhanced the accumulation of 0.1 mM [3H]glucose. Under Na+ gradient conditions the overshoot was strongly increased (5-6 times the equilibrium value) compared with a control of KCl-preloaded vesicles without valinomycin (2-3 times the equilibrium value). Under Na+-equilibrated conditions, an inside negative potential also produced a distinct overshoot (2 times the equilibrium value). Na+ actiuation. In Fig. 4 the hyperbolic dependence of biotin transport on the Na+ concentration is shown. The membrane potential was clamped by equimolar concentrations of KC1 on both sides of the vesicles’ membranes and valinomycin. The Hill transformation of the data revealed a stoichiometric factor of n = 0.89 t 0.07, indicating that roughly one sodium ion is transported per biotin molecule across the membrane. Biotin concentration dependence. When biotin uptake was measured as a function of increasing biotin concentration, transport was found to be saturable (Fig. 5). The contribution of simple diffusion measured in the presence of a choline Cl gradient was minimal within 25 s. With increasing biotin concentration, the specific activity of
0. .+---* */- --*-“-i = 0.993 e-l
/-
,,-(I’-
I
I
1 logha+l
I
I
I
100
Na+
concentration
m = -1.51 b = 0.89
2
I
200
()rM 1
FIG. 4. Na+ activation. Vesicles were preloaded with 500 mM mannitol, 100 mM KCl, 50 mM HEPES-Tris (pH 7.4), and 1~1 valinomycin (4 mg/ml ethanol) per 100 ~1 of vesicle suspension. Incubation was performed for 25 s by adding 20 ~1 of vesicle suspension to 100 ~1 of incubation medium containing 100 mM mannitol, 100 mM KCl, 50 mM HEPES-Tris (pH 7.4), 20 nM [3H]biotin, and increasing NaCl concentrations (0, 10, 20, 50, 100, and 200 mM). At concentrations ~200 mM NaCl was replaced isosmotically by choline Cl. Na’-independent biotin uptake in presence of 200 mM choline Cl was subtracted. Values are means t SE of 3 experiments performed in quadruplicate. Inset: Hill-type plot of data. Relationship was linear according to y = mx + b, where 172is slope, b is y-intercept, and r is correlation coefficient.
I
50
Biotin
I
100
concentration
I
150
I
200
(pM)
FIG. 5. Concentration dependence. Vesicles were preloaded with 300 mM mannitol, 50 mM HEPES-Tris, pH 7.4. Incubation was performed for 25 s by adding 20 ~1 of vesicle suspension to 100 ~1 of incubation medium containing 100 mM mannitol, 100 mM NaCl, 1 PM [3H]biotin, 50 mM HEPES-Tris (pH 7.4), and increasing amounts of cold biotin. Final concentrations ranged from 1 to 200 PM biotin. Values are means & SE of 3 experiments performed in quadruplicate. Inset: Eadie-Hofstee plot of data. Relationship was linear according to y = mx + b, where lt2 is slope, b is y-intercept, and r is correlation coefficient.
[3H]biotin was lowered by dilution with cold biotin. The uptake resulting from diffusion was so small that it became statistically indistinguishable from the values for nonspecific retention on the filters. In Fig. 5, therefore, only the background values were subtracted. The outer concentrations ‘ranged from 1 to 200 PM biotin. The uptake at smaller concentrations (20 nM-1 PM) appeared to be linear (data not shown). The time-dependent uptake of 200 PM biotin was also linear up to 30-40 s (data not shown), but, when extrapolated to zero time, showed as lmall membra ne binding compon .ent that was not visible at lower biotin concentrations. This was taken into account when evaluating the kinetic parameters. The apparent kinetic parameters calculated from the Eadie-Hofstee plot (Fig. 5, inset) were a Michaelis constant (&J of 55 t 4 PM and Vmax = 217 t 10 pmolmg protein-‘. 25 s-l. Inhibition studies. Several biotin derivatives, structural analogues, and other related monocarboxylates were tested at concentrations of 250 PM for their ability to inhibit the uptake of 30 ,uM biotin (Table 1). Of special interest were also the organic acids accumulating in the urine of biotinidase-deficient patients. Among the biotin derivatives tested desthiobiotin, norbiotin, and biotin itself inhibited the apparent transfer of 30 PM [3H]biotin into the vesicles. The uptake relative to the control was
Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 15, 2019.
F844
BIOTIN
TRANSPORT
IN
RENAL
BRUSH-BORDER
MEMBRANE
VESICLES
1. Effect of biotin derivatives and related monocarboxylic acids on biotin uptake
TABLE
Inhibitor
(250 pM)
% of Control
Biotin derivatives None (control) Biotin Biocytin Biotin methylester Diaminobiotin Desthiobiotin Biotin sulfone Biotin-D-sulfoxide Biotin-L-sulfoxide Norbiotin Bisnorbiotin Other monocarboxylic acids None (control) Thioctic acid N-benzoyl-&amino-valeric Probenecid Nonanoic acid Valerie acid Propionic acid Lactic acid 3-OH-isovalerate 3-Methylcrotonic acid
P Values
lOO.Ot2.7
38.2t2.4 96.2k3.8 98.3t3.9 94.8t3.5
42.6t3.8 102.8k2.9 103.723.6 98.6k3.9 42.8t2.1 90.2t2.3
AND E. REGULA BAUMGARTNER Hospital, CH-4058 Base& Switzerland
BAUR,BARBARA,HUGOWICK,AND E. REGULABAUMGARTNER. Na’-dependent biotin transport into brush-border membrane vesicles from rat kidney. Am. J. Physiol. 258 (Renal Fluid
Electrolyte Physiol. 27): F840-F847, 1990.-The mechanisms of biotin reabsorption in rat kidney cortex were investigated using isolated brush-border membrane vesicles. An inwardly directed Na+ gradient specifically stimulated a transient biotin overshoot. Biotin transport was not affected by a valinomycininduced K+-diffusion potential, and biotin--Na+ stoichiometry was found to be 1: 1. As a function of concentration, the uptake showed saturation in the presence of a Na+ gradient with an apparent Michaelis constant (I&) of 55 PM and Vmax of 217 pm01 . mg protein-‘. 25 s-l. Desthiobiotin, 250 PM, norbiotin, bisnorbiotin, thioctic acid, Valerie acid, probenecid, and nonanoic acid inhibited the transport of 30 PM biotin, whereas other biotin derivatives, as well as biocytin and organic acids found in the urine of biotinidase-deficient patients, did not. Preloading of the vesicles with biotin, desthiobiotin, norbiotin, and thioctic acid in the presence of Na+ increased initial uptake of biotin from the incubation medium (trans-stimulation). Our results indicate that biotin absorption in rat kidney fulfills the criteria for a specific carrier-mediated and electroneutral Na+biotin- cotransport in a 1: 1 ratio. The results are discussed in context with congenital biotinidase deficiency in humans. biotinidase deficiency; electroneutrality; trans-stimulation
renal absorption;
THE PHYSIOLOGICAL IMPORTANCE of biotin as a covalently bound prosthetic group for carboxylation reactions is well established. In mammals all four biotin-dependent carboxylases (pyruvate carboxylase, acetyl-CoA carboxylase, propionyl-CoA carboxylase, and 3-methylcrotonoyl-CoA carboxylase) play crucial roles in intermediary metabolism. Nutritional biotin deficiency and the recently discovered genetic disorders of biotin metabolism result in impaired carboxylase activities (2l), i.e., multiple carboxylase deficiency, leading to the accumulation of a typical pattern of abnormal organic acids in body fluids and to a serious metabolic derangement with possible lethal outcome (12). One cause of multiple carboxylase deficiency is biotinidase deficiency (23). The enzyme biotinidase (EC 3.5.1.12) cleaves biotin from biocytin, the product of metabolic degradation of biotin-dependent carboxylases. Patients suffering from biotinidase deficiency are not able to recycle endogenous biotin or to process proteinbound biotin, the main form of this vitamin in the diet. Biotin is lost in the urine in the form of biocytin, which F840
results in progressive biotin depletion in early infancy or childhood. Previous investigations in our laboratory have shown that the high requirement for free biotin in biotinidasedeficient patients is not only the result of defective recycling but also of increased renal excretion of biotin itself (2, 20). Clearance studies revealed that, at normal biotin plasma concentration, renal biotin clearance in patients was equal to or slightly higher than the creatinine clearance. In healthy children and adults biotin clearance was 40% of the creatinine clearance, approximating creatinine clearance with increasing plasma biotin concentrations up to 50 times normal (1). These results suggest a saturation phenomenon either at the site of glomerular filtration attributable to a biotinbinding protein in plasma and/or of tubular reabsorption of biotin. So far little is known about the renal handling of biotin or how biotin is transported in the blood. In a recent report biotinidase itself was identified as a biotintransporti .ng protein in pla .sma (7). We focused our interest on the mechanisms of renal biotin reabsorption and, particularly, whether it can be influenced by biocytin and possibly also by other abnormal metabolites. For our investigations we used brush-border membrane vesicles isolated from rat kidney cortex. To our knowledge there are only two publications concerning biotin reabsorption in kidney. Podevin et al. (13) described an active Na+-dependent electrogenic biotin transport into brush-border membrane vesicles from rabbit kidney, and recently Spencer and Roth (19) reported an energy-dependent and saturable biotin transport in brush-border membrane vesicles from rat kidney. More data are available on intestinal absorption of biotin observed by use of several in vivo and in vitro techniques (3, 6, 8, 15-17).
In the present study we confirm the existence of a carrier-m .ediated, Na+-dependent biotin transport system in brush-border membrane vesicles isolated from rat kidney cortex and characterize some of its properties in more detail. MATERIALS
AND METHODS
Materials. [8,9-3H] biotin (30-50 Ci/mmol) was purchased from Du Pont de Nemours (Zurich, Switzerland). Unlabeled biotin, desthiobiotin, diaminobiotin, biotin methyl ester, and biocytin were from Sigma Chemical (Poole, Dorset, UK). Biotin-D-sulfoxide, biotin+sulfoxide, and biotin sulfone were gifts from Dr. Bowers-
0363-6127/90 $1.50 Copyright 0 1990 the American Physiological
Society
Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 15, 2019.
BIOTIN
TRANSPORT
IN
RENAL
BRUSH-BORDER
Komro (Woodruff Medical Center, Atlanta, GA). Norbiotin was a gift from Dr. J. Bausch (HoffmannLaRoche, Basel, Switzerland), and bisnorbiotin was synthesized in 95% purity by Dr. W. Holick (HoffmannLaRoche). Thioctic acid was obtained from Merck (Zurich, Switzerland). Probenecid, nonanoic acid, and N-benzoyl-&aminovaleric acid were from Sigma. 3Methylcrotonic acid was obtained from Fluka (Buchs, Switzerland). 3-OH-isovalerate (ammonium salt) was a gift from Dr. C. Bachmann (Inselspital, Bern, Switzerland). Rialuma scintillation cocktail was purchased from Lumac (Landgraaf, The Netherlands). All other chemicals were obtained from commercial sources and were of analytical grade. Methods. Brush-border membrane vesicles from rat kidney cortex slices (Wistar rats) were prepared by an ethylene glycol-bis(P-aminoethyl ether)-N,N,N’,N’-tetraacetic acid (EGTA)-magnesium precipitation method as described by Biber et al. (4) with slight modifications. The rats were killed by an overdose of chloroform. The abdomen was opened, and the kidneys were removed. After removal of the capsule, the cortex slices were dissected with a razor blade. The slices were homogenized with a polytron (PT 10-35, Kinematica, Kriens, Switzerland) on setting 7 for 2 min. The homogenate was subjected to magnesium precipitation and differential centrifugation. The high-speed centrifugation steps were performed at 18,000 revolutions/min (rpm) and 16,000 rpm for 30 min in a Sorvall RC-5C centrifuge with a SS34 rotor. After the purification the vesicles were loaded with buffers as indicated in the legends to Figs. l-8 and Table 1. The final pellet was suspended in the desired buffer and rehomogenized with a glass-Teflon potter. Then the membranes were collected by centrifugation (20,000 rpm) and resuspended at a concentration of 510 mg/ml in an appropriate volume of the final buffer by repeated suction through a fine needle (gauge 0.5 mm). After equilibration of the vesicles for 90 min at room temperature, transport measurements were performed as follows. The uptake of [3H]biotin was started by adding an aliquot (lo-20 ~1) of the vesicle suspension to the incubation medium (100-200 ~1) containing tritiated biotin. It was terminated by the addition of 4 ml icecold stop solution containing 90 mM mannitol, 100 mM NaCl, 10 mM N-2-hydroxyethyl piperazine-N’-2-ethanesulfonic acid and tris( hydroxymethyl)aminomethane (HEPES-Tris), pH 7.4, to the reaction mixture and subsequent rapid filtration through nitrocellulose filters (Millipore, HAWP 02412, pore size 0.45 pm) kept under suction. The filters were washed with another 4 ml of chilled stop solution and the retained radioactivity was counted in 10 ml Rialuma by use of a Tri-Carb 4530 scintillation counter (Packard Instrument, Downers Grove, IL). Values for the nonspecific binding of C3H]biotin (0.02% of the total radioactivity) on the filters were determined by filtering a reaction mixture without vesicles; those values were then subtracted from the incubated samples. Nonspecific retention remained unchanged in the presence of 0.1 mM biotin or biotin-D/Lsulfoxide in the stop solution. All incubations were carried out at 25°C and at least
MEMBRANE
VESICLES
FE341
in triplicate with freshly prepared membrane vesicles. Each experiment was repeated at least three times. Although the quantitative uptake of [3H]biotin varied from one preparation to another, the qualitative results were in good agreement among the different experiments. Therefore the data of individual experiments are presented in most instances. The results are generally expressed as femtomoles or picomoles per milligram protein t SD or as percentage t SE of the mean. Unless indicated, SD or SE values do not exceed the size of the symbols in the figures. The purity of the isolated brush-border membrane vesicles was routinely checked by measuring the enrichment of the brush-border marker enzyme, alkaline phosphatase. The enrichment was lo- to &fold with respect to the starting homogenate. Alkaline phosphatase activity was assayed as described by Bowers and McComb (5) using an automated analysis kit (Boehringer Mannheim, Mannheim, FRG). Biotinidase activity was measured according to Wastell et al. (22). Activity was found to be -40-100 pmol . min-l . mg protein-l in the homogenate and below detection limits (-0.1 nmol/min) in the brush-border membrane fraction. Protein was determined using the Bio-Rad reagent with bovine plasma albumin as a standard (Bio-Rad, Richmond, CA). To confirm the identity of the transported biotin, vesicles were incubated for 90 min at 25°C in a medium containing 0.5 PM [3H]biotin, 100 mM mannitol, 100 mM NaCl, and 50 mM HEPES-Tris (pH 7.4). The membranes were collected by filtration and the radioactivity was extracted as described by Podevin et al. (13). Fifty microliters of the extracts were spotted on thinlayer chromatography aluminum sheets with si.lica ge 160 (Merck). The following solvent systems were used: butanol-acetic acid-water (4:1:5; upper phase) and benzene-methanol-acetone-acetic acid (14:4:1: 1).With both systems >93% of the radioactivity comigrated with the biotin standard (Rf = 0.71 and 0.51,respectively, where Rf is the migration relative to the solvent front). RESULTS
Time course and Na+ specificity. Biotin uptake as a function of time is shown in Fig. 1. In four experiments essentially the same results have been obtained. During the first 25-40 s biotin transport was apparently linear (Fig. 1, inset). Extrapolation of the initial linear transport rate to zero time corresponds nearly to zero uptake, indicating that there is no substantial binding to the membranes. An overshoot phenomenon was observed in the presence of an extravesicular-to-intravesicular N .a+ gradient. Maximal accumulation was reached within 46 min of incubation and was about twice the equilibrium value. Thereafter the uptake of biotin slowly diminished to equilibrium at 90 min of incubation, indicating efflux from the vesicles. None of the other monovalent cations tested (Li’, K+, choline+) was able to produce an overshoot effect when imposed as inwardly directed gradients with Cl- as the accompanying anion. The uptake in the presence of LiCl, KCl, and choline Cl was not signifi-
Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 15, 2019.
F842
BIOTIN
TRANSPORT
IN RENAL
30
BRUSH-BORDER
MEMBRANE
VESICLES
60
lime
(set)
I
,,,,,~,,~~~, 2
4
6
8
(Osmolar
10
12
it y)”
FIG. 2. Medium osmolarity. Renal brush-border membrane vesicles were loaded with 100 mM sucrose, 50 mM HEPES-Tris, pH 7.4. Uptake was started by adding 20 ~1 of vesicle suspension to 100 ~1 of incubation medium containing 100 mM NaCl, 20 nM [3H]biotin, 50 mM HEPESTris (pH 7.4), and increasing concentrations of sucrose (loo-1,000 mM). Incubation was performed for 90 min. Values are means +: SD of 1 experiment performed in triplicate. Relationship was linear according to Y = rnx + b, where n is slope, b is y-intercept, and r is correlation coefficient.
Time
(mid
1. Cation requirements. Renal brush-border membrane vesicles were resuspended in 300 mM mannitol, 50 mM HEPES-Tris, pH 7.4. At start of incubation, 20 ~1 of vesicle suspension were added to 100 ~1 of incubation medium containing the following: 300 mM mannitol, 50 mM HEPES-Tris (pH 7.4), and 20 nM [3H]biotin (A), or 100 mM mannitol, 50 mM HEPES-Tris (pH 7.4), 20 nM [3H]biotin, and 100 mM NaCl (0); 100 mM LiCl (0); 100 mM KC1 (w); or 100 mM choline Cl (cl). Values are means t SD for triplicate assays of 1 experiment. Inset: uptake of 20 nM [3H]biotin during first 60 s of incubation under NaCl conditions: values are means & SD of 1 experiment performed in quadruplicate. FIG.
cantly different from the biotin uptake in the absence of any salt gradient. After 90 min of incubation, the same equilibrium value was reached under all conditions. Effect of medium osmolarity. To ensure that the biotin accumulation represents transport into the vesicles’ inner space and not simple binding to the membranes, the effects of increasing medium osmolarity on biotin uptake were examined. Figure 2 shows that the equilibrium value of biotin uptake decreased with increasing sucrose concentration in the incubation medium. Extrapolating to zero, i.e., infinite osmolarity, reveals a binding component of 13.1 fmol/mg protein, which accounts for 20% of the equilibrium value under isotonic conditions. Membrane potential dependence. The influence of the electrical membrane potential was tested by eliciting a strong transient intravesicular potential. To this aim the vesicles were preloaded with 100 mM KC1 and valinomycin and subsequent transport measurements were performed in K+-free medium. As illustrated in Fig. 3, the inside negative membrane potential failed to stimulate both initial and maximal biotin uptake in the presence of an inwardly directed Na+ gradient (curue A vs. C). Curve D (control) shows the biotin uptake as a function of time when a Na+ gradient (out > in) was imposed in
A
-5 100
I
b k
3
P
.
A 0 Na’ K’ I3 0 Na’ K’ C Wla+ K' 0 c] Na’
out s in in > out, d/al out = in in + out, l Val outdn in D out, +EtOH out, in, l EtOH
T
4) < 50 4
.-E H m
Time
(mid
3. Membrane potential dependence. Vesicles were loaded with 300 mM mannitol, 50 mM HEPES-Tris, pH 7.4 (o), or with 100 mM mannitol, 100 mM KCl, 50 mM HEPES-Tris (pH 7.4) with (0) and without valinomycin (m), and with an additional 100 mM NaCl and valinomycin (0). Valinomycin, 1 ~1 [4 mg/ml ethanol (EtOH)], was added per 100 ~1 of vesicle suspension. To controls without valinomycin, ethanol was given in same amount. Incubation was started by adding 20 ~1 of vesicles to 100 ~1 of medium of the following composition: 100 mM NaCl, 20 nM [3H]biotin, 50 mM HEPES-Tris (pH 7.4), and 100 mM mannitol (0, n , IX); or 100 mM NaCl, 20 nM [3H]biotin, 50 mM HEPES-Tris (pH 7.4), and 300 mM mannitol(0). Values are means * SD of 1 experiment performed in triplicate. FIG.
Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 15, 2019.
BIOTIN
TRANSPORT
IN RENAL
BRUSH-BORDER
MEMBRANE
F843
VESICLES
the absence of K+. The presence of KC1 at the trans side of the vesicles had no significant influence (curve C vs. D) nor did valinomycin itself alter significantly the timedependent biotin uptake (data not shown). With Na+ present at equilibrium on both sides of the vesicle membrane, a K+ diffusion potential alone did not result in biotin accumulation above equilibrium as illustrated in curve B.
In a control experiment (data not shown) a valinomycin-induced K+ diffusion potential enhanced the accumulation of 0.1 mM [3H]glucose. Under Na+ gradient conditions the overshoot was strongly increased (5-6 times the equilibrium value) compared with a control of KCl-preloaded vesicles without valinomycin (2-3 times the equilibrium value). Under Na+-equilibrated conditions, an inside negative potential also produced a distinct overshoot (2 times the equilibrium value). Na+ actiuation. In Fig. 4 the hyperbolic dependence of biotin transport on the Na+ concentration is shown. The membrane potential was clamped by equimolar concentrations of KC1 on both sides of the vesicles’ membranes and valinomycin. The Hill transformation of the data revealed a stoichiometric factor of n = 0.89 t 0.07, indicating that roughly one sodium ion is transported per biotin molecule across the membrane. Biotin concentration dependence. When biotin uptake was measured as a function of increasing biotin concentration, transport was found to be saturable (Fig. 5). The contribution of simple diffusion measured in the presence of a choline Cl gradient was minimal within 25 s. With increasing biotin concentration, the specific activity of
0. .+---* */- --*-“-i = 0.993 e-l
/-
,,-(I’-
I
I
1 logha+l
I
I
I
100
Na+
concentration
m = -1.51 b = 0.89
2
I
200
()rM 1
FIG. 4. Na+ activation. Vesicles were preloaded with 500 mM mannitol, 100 mM KCl, 50 mM HEPES-Tris (pH 7.4), and 1~1 valinomycin (4 mg/ml ethanol) per 100 ~1 of vesicle suspension. Incubation was performed for 25 s by adding 20 ~1 of vesicle suspension to 100 ~1 of incubation medium containing 100 mM mannitol, 100 mM KCl, 50 mM HEPES-Tris (pH 7.4), 20 nM [3H]biotin, and increasing NaCl concentrations (0, 10, 20, 50, 100, and 200 mM). At concentrations ~200 mM NaCl was replaced isosmotically by choline Cl. Na’-independent biotin uptake in presence of 200 mM choline Cl was subtracted. Values are means t SE of 3 experiments performed in quadruplicate. Inset: Hill-type plot of data. Relationship was linear according to y = mx + b, where 172is slope, b is y-intercept, and r is correlation coefficient.
I
50
Biotin
I
100
concentration
I
150
I
200
(pM)
FIG. 5. Concentration dependence. Vesicles were preloaded with 300 mM mannitol, 50 mM HEPES-Tris, pH 7.4. Incubation was performed for 25 s by adding 20 ~1 of vesicle suspension to 100 ~1 of incubation medium containing 100 mM mannitol, 100 mM NaCl, 1 PM [3H]biotin, 50 mM HEPES-Tris (pH 7.4), and increasing amounts of cold biotin. Final concentrations ranged from 1 to 200 PM biotin. Values are means & SE of 3 experiments performed in quadruplicate. Inset: Eadie-Hofstee plot of data. Relationship was linear according to y = mx + b, where lt2 is slope, b is y-intercept, and r is correlation coefficient.
[3H]biotin was lowered by dilution with cold biotin. The uptake resulting from diffusion was so small that it became statistically indistinguishable from the values for nonspecific retention on the filters. In Fig. 5, therefore, only the background values were subtracted. The outer concentrations ‘ranged from 1 to 200 PM biotin. The uptake at smaller concentrations (20 nM-1 PM) appeared to be linear (data not shown). The time-dependent uptake of 200 PM biotin was also linear up to 30-40 s (data not shown), but, when extrapolated to zero time, showed as lmall membra ne binding compon .ent that was not visible at lower biotin concentrations. This was taken into account when evaluating the kinetic parameters. The apparent kinetic parameters calculated from the Eadie-Hofstee plot (Fig. 5, inset) were a Michaelis constant (&J of 55 t 4 PM and Vmax = 217 t 10 pmolmg protein-‘. 25 s-l. Inhibition studies. Several biotin derivatives, structural analogues, and other related monocarboxylates were tested at concentrations of 250 PM for their ability to inhibit the uptake of 30 ,uM biotin (Table 1). Of special interest were also the organic acids accumulating in the urine of biotinidase-deficient patients. Among the biotin derivatives tested desthiobiotin, norbiotin, and biotin itself inhibited the apparent transfer of 30 PM [3H]biotin into the vesicles. The uptake relative to the control was
Downloaded from www.physiology.org/journal/ajprenal by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 15, 2019.
F844
BIOTIN
TRANSPORT
IN
RENAL
BRUSH-BORDER
MEMBRANE
VESICLES
1. Effect of biotin derivatives and related monocarboxylic acids on biotin uptake
TABLE
Inhibitor
(250 pM)
% of Control
Biotin derivatives None (control) Biotin Biocytin Biotin methylester Diaminobiotin Desthiobiotin Biotin sulfone Biotin-D-sulfoxide Biotin-L-sulfoxide Norbiotin Bisnorbiotin Other monocarboxylic acids None (control) Thioctic acid N-benzoyl-&amino-valeric Probenecid Nonanoic acid Valerie acid Propionic acid Lactic acid 3-OH-isovalerate 3-Methylcrotonic acid
P Values
lOO.Ot2.7
38.2t2.4 96.2k3.8 98.3t3.9 94.8t3.5
42.6t3.8 102.8k2.9 103.723.6 98.6k3.9 42.8t2.1 90.2t2.3
