THE MECHANISMS·OF FLUORIDE UPTAKE BY THE BRUSH BORDER MEMBRANE VESICLES (BBMV) OF RABBIT SMALL INTESTINE AND KIDNEY
by
HanHe
Submitted to the Faculty of the School of Graduate Studies of the Medical College of Georgia in partial fulfillment of the Requirements of the Degree of Doctor of Philosophy
June 1997 THE MECHANISMS OF FLUORIDE UPTAKE BY THE BRUSH-BORDER
MEMBRANE VESICLES (BBMV) OF ~BIT SMALL INTESTINE AND KIDNEY
This dissertation is submitted b.Y HAN HE and has been examined and approved by an appointed committee of the faculty of the School .of Graduate Studies of the Medical
College of Georgia.
The signatures which appear below verify the fact that all required changes have been incorporated and that the dissertatioii has received fmal approval with reference to content, form and accuracy of presentation.
This dissertation is therefore in partial fulfillment of the requirements for the degree of Doctor of Philosophy.
Date
Department Chairperson
Dean, School of Graduate studies iii
DEDICATION
This work is dedicated to my wife, Rong Zou, my daughter Cassie and my parents,
Kechun He, Fushou Xiao. iv
ACKNOWLEDGMENTS
I would like to express my sincere gratitude to the following people for without
their help and support, I would have been unable to successfully complete this project.
First and foremost, I wish to thank my major advisor, Dr. Gary M. Whitford, for
0 0 0 • L '•, his guidance and support and wisdom throughcJµtJ!le course of this· work.
I would like to thank Dr. Vadivel Ganapathy for his invaluable advice and for
teaching me .the methods required for the preparation of brush border membrane vesicles. I
also wish to e~press my appreciation to the rest of my advisory committee, Dr~.· James L.
Borke, David H. Pashley, and Frederick H. Leibach, for their critical suggestions and
guidance.
I wish to thank Ors. Carlos Isales, Lailin Min, and Kehong Ding for helping me measure intravesicular pH and valuable discussion.
I also want to thank Mary Maness for her technical assistance and to Michelle
Burnside and Shirley Jonston for their secretarial support. My special thanks to Ms.
Chenjin Zhou and Puttur D. Prasad for their supporting. V
TABLE OF CONTENTS
DEDICATION ...... iii ACKNOWLEDGMENTS...... iv TABLE OF CONTENTS...... V LIST OF FIGURES...... Vil LIST OF TABLES...... viii LIST OF ABBREVIATIONS...... ix INTRODUCTION ...... ·...... 1 Statement of Problem and Specific Aims.:...... 1 Review of Literature...... 4 B. l. Overview...... -4 B.2. Fluoride Absorption...... 5 MATERIALS· AND METHODS~ ... ~ .. ~ ...... · ... :...... 9 A. Agents Used ...... -...... ·...... ~ ...... ~ ...... ~...... 9 B. Experimental Animals...... 9 C. Experimental Procedure ...... •...... _...... 9 C .1. Preparation of Everted Sacs of .Rat Intestine...... 9 · · - C.1.a. Effect of pH on Migration of Fluoride Using Everted Intesiinal Sacs ...... ·~.- ...... 10 -c.2., Brush Border Membrane Vesicle (BBMV) Studies ...... ~...... 13 C.2.a. Preparation of BBMV~.~ ...... · ...... 13 C.2.b. · . Fluoride Uptalce Studies: Ge.neral Method ...... 14 C.2.c .. Preparation· of. Solutions ...... - ...... : ...... 16 C.2.d. Control Studies ...... - ...... 17 C.2.e. Time Course of Fluoride Uptake ...... :...... 17 C.2.f. Influence of pH on Fluoride Uptake ...... 17 C.2.g. Influence of Medium Osmolarity on Fluoride Uptake ...... 18 C.2.h. Effects of DIDS and DEP on Fluoride Uptake...... 19 C.2.i. Effect of diBAC on Fluoride Uptake ...... 20 C.2.j. Effect of PCMBS on Fluoride Uptake...... 20 C.2.k. Effect of DIDS, DEP, PCMBS, and diBAC on Transmembrane pH Gradients ...... , ...... 20 C.2.1. F--c1- and/or F--Hco3 - Exchange Study ...... 21 C.2.m. Fluoride Conductance Pathway Study...... 22 C.2.n. Effect of Sodium Gradient on Fluoride Uptake ...... 23 D. Data Analysis ...... 24 RESULTS ...... 25 1. Control Studies ...... ,...... 25 2. Time Course of Fluoride Uptake ...... ~ ...... 28 3. Influence of pH and Transmembrane pH Gradients on Fluoride Uptake...... 31 4. Effect of Medium Osmolarity on fluoride Uptake ...... 34 5. Effects of DIDS and DEP on Fluoride Uptake ...... 37 6. Effect of PCMBS on Fluoride Uptake ...... 40 7. Effect of diBAC on Fluoride Uptake ...... ~ ...... 42 8. Effect of DIDS, DEP, PCMBS, and diBAC on Transmembrane pH Gradients ...... ~ ...... 45 vi
9. F--c1- and/or F--Hco3- Exchange Study ...... - ...... 51_ 10. Fluoride Conductance Pathway Study...... 54 11. Effect of Sodium or Potassium Gradient on Fluoride Uptake ...... L ...... 58 12. EffectofNOJ- and c1- on Fluoride Uptake ...... 61 13. Effect of pH on Migration of Fluoride Using Everted Intestinal Sacs ...... 64 DISCUSSION ....· ...... -...... 66 SUMMARY ...... 76 REFERENCES ...... 77 vii
LIST OF FIGURES
Figure Page
1. Apparatus used for studies with everted intestinal sacs...... 12 2. Schematic showing the filtration apparatus...... 15 3. Effectiveness of number of washings on removal of fluoride from the filters ...... 26 4. Time course of fluoride uptake by small intestinal BBMV ...... 29 5. Time course of fluoride uptake by renal BBMV...... 30 6. Effect of medium osmolarity on F uptake by small intestinal BBMV ...... 35 7. Effect of medium osmolarity on F __uptake by renal BBMV ...... 36 8. Effect of PCMBS on fluoride· uptake ...... 41 , 9. Effect of diBAC on fluoride uptake by small intestinal BBMV...... 43 10. Effect of diBAC on fluoride uptake by renal BBMV ...... 44 11. Effect of DIDS, DEP, PCMBS and diBAC on transmembrane pH gradient ...... 48 12. Effect of DIDS, DEP, PCMBS and diBAC on BCECF fluorescence signal ...... 49 13. Effect of DIDS, DEP, PCMBS and diBAC on transmembrane pH gradient ...... ·...... 50 14. Cl-/F- and HCO3-/F- exchange study using small intestinal BBMV ...... 52 15. c1-1F- and HCO3-/F- exchange study using renal BBMV ...... 53 16. Effect of an inwardly-redirected K+ diffusion potential on fluoride uptake by small intestinal BBMV ...... : ...... 55 17. Effect of an inwardly-redirected K+ diffusion potential on fluoride uptake by renal BBMV ...... 56 18. Effect of reversal ofK+ diffusion potential on fluoride uptake ...... 57 19. Effect of Na+ and K+ gradient on F uptake by small intestinal BBMV ...... 59 20. Effect of Na+ and K+ gradient on F uptake renal BBMV ...... 60 21. Effect of c1- and N~- on F uptake by small intestinal BBMV ...... 62 22. Effect of c1- and NO3 - on F uptake renal BBMV ...... 63 Vlll
LIST OF TABLES Table Page
1. The Content of Magnesium in Different Membrane Preparations ...... 27 2. Influence of pH per se and a Transmembrane pH Gradient on Fluoride Uptake in Small Intestinal BBMV ...... 32 3. Influence of pH per se and a Transmembrane pH Gradient on- Fluoride Uptake in Renal BBMV ...... 33 4. The effect of DIDS and DEP on Fluoride Uptake in Small Intestinal BBMV ...... 38 5. The effect of DIDS, DEP and PCMBS on Fluoride Uptake in Renal BBMV ...... 39 6. Effect of pH on Migration of Fluoride Using Everted Intestinal Sacs ...... 65 ix
LIST OF- ABBREVIATIONS
BBMV, brush border membrane vesicles;
BCECF acid, 2',7'-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein;
DEP, diethylpyrocarbonate; diBAC, bis-(1,3-dibutylbarbituric acid)pentamethine ~xonol;
DIDS, 4,4'-diisothiocyanostilbene-2,2'-disulfonate;
DMSO, dimethylsulfoxide;
EGTA, ethylene glycol bis (beta-aminoethyl-ether), N, N, N', N'-tetraacetic acid;
HEPES, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid;.
HMDS, hexamethyldisiloxane;
IC50, half-maximal inhibitory concentration;
MES, 2-(N-morpholino)ethanesulfonic acid;
PCMBS, p-chloromercuribenzene sulfonate; pHi, intravesicular pH; ·
pH0 , extravesicular pH; pHm, mucosa! pH;
pH8, serosal pH;
Tris, tris(hydroxymethyl)aminomethane;
TMA-OH, tetramethylammonium; INTRODUCTION
A. Statement of the Problem
Ingested fluoride is absorbed rapidly from the gastrointestinal tract. The literature indicates the following about the gastrointestinal absorption of fluoride:
1. the rate and extent of absorption are reduced by certain divalent and trivalent
cations;
2. the rate of absorption from the intact stomach is dependent on the magnitude of
the transinucosal pH gradient;
,·. 3. the rate of absorption froni the intact small intestine is not affected significantly
by the pH of the bulk solution;
4. there is evidence that some unidentified, energy-requiring mechanism(s) may be
involved in the intestinal absorption of fluoride;
5. although DIDS (4,4'-diisothiocyanostilbene-2,2'-disulfonate) has little or no
effect on fluoride movement across the human red blood cell membrane or
the cortical collecting duct of the rabbit, it partially inhibits uptake by rat
submandibular salivary cells.
This study was carried out to test the hypothesis that the intestinal and renal absorption of fluoride occurs by more than one mechanism. That is, some fluoride is I absorbed in a pH-dependent manner by HF diffusion through the apical membrane of
intestinal cells, possibly via aquaporins. The bulk of fluoride, however, is absorbed as
ionic fluoride via other routes poss~bly including passage through intercellular gap junctions, aquaporins, F--c1- and/or. F--Hc~- exchanger(s), conductive pathway and
specific anion-transporting protein carriers.
The specific aims of this study were to determine the effects of pH, pH gradients,
1 2 sodium, potassium, chloride and bicarbonate gradi~nts, DIDS ( a specific inhibitor of anion transport), bis-(1,3-dibutylbarbituric acid) pentamethine oxonol (diBAC, another specific inhibitor of anion transport ), diethylpyrocarbonate (DEP, a histidine-specific reactive reagent), parachloromercuri-benzene sulfonate (PCMBS, a sulfhydryl-modifying reagent which is an inhibitor of aquaporins), and valinomycin on the movement of fluoride across brush border membrane vesicles (BBMV) prepared from the proximal intestinal mucosa and renal cortex of the rabbit. The effect of pH gradients on fluoride transfer in everted intestinal sacs from the rat was also studied.
To accomplish the goals stated, _two models were used: ( 1) the "everted sac" technique using tissue from rats; and (2) BBMV prepared from intestinal mucosa! scrapings and renal cortex of rabbits. Both methods allow control of the compositions of the solutions on both surfaces. The everted sac technique involves the full thickness of the intestinal wall whereas BBMV involve only cell membranes from the luminal aspect of the mucosa! cells (mainly enterocyte and some Goblet cells, enteroendocrine cells, Paneth cells, etc.). A,comparison of the findings using these two approaches may show that the variables to be tested have different effects depending on the model and thus reveal the existence of different absorptive mechanisms or pathways.
The variables to be tested included: (1) pH; (2) pH gradients; (3) specific inhibitors of (a) anion transport (DIDS and diBAC, which combine with the anion transporter to form a stable derivative to inhibit anion transport) and (b) proton transport (DEP, a histidine specific reactive reagent which has been used to 9haracterize proton and proton-coupled transport); (4) PCMBS, a sulfhydryl-modifying reagent which is a non-specific inhibitor of several transport systems including aquaporins; (5) valinomycin; and (6) gradients of sodium and potassium.
The mechanism(s) involved in the absorption of fluoride from the intestinal tract remain poorly defined. The literature indicates that the rate of fluoride transport or migration across several types of epithelia and ·cell membranes is directly proportional to the 3 magnitude of the pH gradient. This dependency, however, appears not to apply to the intestinal epithelium. The present studies were designed to clarify this matter and thus contribute to an improved understanding of the basic mechanisms and pathways involved in the transport of fluoride in biological systems. The findings may have implications for the transport of other anions as well. 4
B. REVIEW OF LITERATURE
B. 1. Overview
Fluoride is the ionic form of fluorine, the most electronegative of all elements. It has chemical and physiological properties of importance to human health and well-being. It is a natural component of the biosphere and is the thirteenth most abundant element in the earth's crust. As such, it is not surprising that it has been found in a wide range of concentrations in virtually all inanimate and living things.
Research on the biology of fluorides has been very extensive during this century, especially after it was demonstrated in the 1930s that fluoride in very small doses has remarkable influences on the dental system: on the one hand, a strong inhibition of dental caries; on the other, and with higher doses, a disturbance of enamel formation known as dental fluorosis. Fluoride was first added to public drinking water in 1945 to prevent tooth decay (USPHS, 1991). Today, various fluoride compounds are added to dentifrices, mouth rinses, topical gels, lacquers and other products. The remarkable decline in dental caries that is now occurring throughout much of the world has been largely attributed to the use of the ingested and topical forms of fluoride. Indeed, fluoride is now widely regarded as the cornerstone of modern preventive dentistry. Fluoride is also an important agent in the treatment of osteoporosis, a metabolic bone disease characterized by a loss of normal bone mineral mass that principally afflicts postmenopausal women (Harrison et al., 1981;
Riggs et al., 1982).
At the same time, fluoride is a hazardous substance when ingested acutely in large amounts or when ingested in lower amounts over a period of time. I ts dose-dependent chronic effects include dental fluorosis and skeletal fluorosis (Singh and Jolly, 1970;
Bhussry, 1970). Its acute toxic effects range from transient gastric distress and reduction in renal concentrating ability (Mazze et al., 1971; Whitford and Taves, 1973) to severe hyperkalemia, hypocalcemia, cardiovascular collapse and death (Hodge and Smith, 1965). 5
Most commonly, fluoride is absorbed and enters the _body fluids by way of the lungs or gastrointestinal tract. It may. also be released within the body following the
biotransformation of fluorine-containing organic molecules such as the volatile anesthetics like methoxyflurane, halothane or enflurane, among others. Plasma is regarded as the central compartment because it is this fluid from which and into which the ion must pass for its subsequent distribution and elimination. The quantitatively important fates of absorbed fluoride are uptake by calcified tissues and excretion in the urine. Approximately
99% of fluoride in the body is associated with calcified tissues. The fluoride in calcified tissues is not irreversibly bound. It may exchange isoi,onically with fluoride ions or heteroionically with other ions available from the extracellular fluids. Within the various soft tissues, fluoride rapidly establishes a steady-state distribution between the e~tracellular . and intracellular fluids. The major route for the elimination of fluoride is excretion in the
_urine. The bulk of fecal fluoride represents ingested fluoride that was not absorbed from the gastrointestinal tract (Whitford, 1996).
B.2. Fluoride absorption
When taken by mouth, the absorption of fluoride begins in the oral cavity. Gabler
(1968) and Patten et al. (1978) placed fluoride solution into the mouths of esophageal ligated rats. Based on analyses of blood, urine, soft tissues and bone, it was concluded that absorption occurred at a measurable but substantially lower rate than that which occurs after swallowing. Whitford et al. (1982) investigated the influence of solution pH and fluoride concentration on the absorption of fluoride, 3H20 ,12s1 and 14C-DMO (5,5- - dimethyl-2,4 oxazolidinedione) from the hamster cheek pouch in situ. Based on the fact that hydrog~n fluoride is a weak acid _(HF, pKa = 3.4) and that, in generat the permeabilities of cell membranes and epithelia are greater for the undissociated molecules than for the anions, they hypothesized that the rate of absorption would be inversely proportional_·to solution pH. The first order kinetics and strong pH..:dependence of the 6
absorptive process was consistent with the hypothesis that fluoride was absorbed by
diffusion in the form of the undissociated.acid, HF.
Several reviews of the gastrointestinal absorption of fluoride (Smith, 1966; Cremer
and Buttner, 1970; Whitford, 1996) indicate that, in the absencec of relatively high levels of
di .. or trivalent cations such as calcium, magnesium an~ aluminll:m that can combine with
fluoride to form insoluble compounds, the rate of absorption is unusually rapid and that the
degree of absorption is nearly complete. Peak plasma fluoride levels are usually achieved
within 20 to 40 minutes after 1ngestiqn and the extent of absorption ranges from 80 to
almost 100 percent of the administered dose, depending on the form and method of
administration.
Most published- reports suggest oi- indicate that fluoride is absorbed by rapid .
passive diffusion along the entire gastrointestinal tract. Whitford and Pashley (1984)
studied the influence of gastric acidity on the absorption of intragastrically administered
fluoride in rats. Their findings were consistent with the hypothesis that fluoride is
absorbed from the gastric lumen principally as the undissociated molecule, HF, which also appears to be the principal permeating moiety ina variety of cells and epithelia including the oral mucosa, renal tubules, urinary bladder, red blood cells and isolated hepatocytes
(Whitford, 1996). Ionic fluoride (F-) appears to be virtually incapable of such permeation in these systems. Using cholesterol .. lecithin lipid bilayer membranes, Gutknecht and
Walter found that the permeability coefficient of HF was 106 times greater than that of ionic
fluoride (Gutknecht and Walter, 1981), presumably because of the ion's charge and large
hydrated radius (Whitford arid Williams, 1986). This finding was consistent with the
results of in vivo animal studies showing that fluoride absorption is increased when the pH
of the gastric contents is in the acidic range (Whitford and Pashley, 1984).
Nopakun etal. (1989, 1990) and Messer and Ophaug (1993) reported that, in intact
rats without any measures taken to influence gastric acidity, the rate of increase in plasma
fluoride concentrations was dependent on the rate' of gastric emptying into the small 7 intestine. This suggested that the rate of absorption from the upper intestine was greater than from the stomach. They further investigated gastric fluoride absorption in relation to gastric acidity and, in agreement with earlier findings, concluded that the rate and extent of absorption from the stomach was inversely related to pH. Absorption from the small intestine, however, compensated for the low gastric absorption at high pH so that the extent of absorption after an hour or more was ind_ependent of the pH of the gastric· contents.
More recently they studied fluoride transport in vitro by int~stinal and renal epithelial monolayers (Messer and Birner, 1995). They concluded that renal fluoride transport occurred largely as undissociated HF while intestinal fluoride transport occurred ~s ionic fluoride.
The results of some studies, however, have suggested that fluoride absorption may involve mechanisms other than diffusion. Stookey et al. (1963, 1964a, 1964b) reported that there was an inverse relationship. between: the rate of fluoride absorption and plasma fluoride levels in rats. They suggested that there might be a negative feedback system operating at the level of the gut to homeostatically control plasma fluoride concentrations.
Parkins et al. (1966, 1971) obtained evidence from in vitro studies that the intestinal handling of fluoride in several species of rodent~-was an active process in the secretory direction, being capable of 'uphill' transport and subject to inhibition by metabolic poisons and low temperature. These findings suggested that diffusion, whether as HF or ionic fluoride, was not the only mechanism for the absorption of fluoride. However, the authors did not consider the possibility of alkalinization of the mu~osal solution subsequent to the secretion of bicarbonate. To the extent that this occurred, it would have established a diffusion gradient fqr HF subject to inhibition by a variety of means.
The literature contains conflicting reports concerning the effects of the disulfonic stilbene, DIDS, on the transmembrane or transepithelial transport of fluoride. _DIDS is a
potent inhibitor of the cell membrane band-3 protein anion exchanger. Whitford et al.
(1985) reported that DIDS almost completely inhibited the efflux of chloride from red blood 8 cells but had little or no effect on the efflux of fluoride. Similarly, Rouch et al. (1992) found no effect of DIDS on the transport of fluoride from the isolated, perfused cortical collecting duct of the rabbit. Melvin et al. ( 1990), however, used rat submandibular gland cell aggregates and found that fluoride uptake was partially inhibited by both bumetanide,
an inhibitor of the Na/K/Cl cotransporter, and DIDS. They concluded that fluoride movement in salivary epithelia occurred primarily via chloride pathways .and that passive fluoride fluxes were of minor importance. Along with the lack of a pH gradient effect on.
the intestinal absorption of fluoride, these ,divergent results stemming from DIDS studies with red blood cells, renal tubules and submandibular cells suggest the possibility that . fluoride transport may share a common pathway with chloride that is peculiar to certain
cells, particularly glandular cells. MATERIALS AND METHODS
A. Agents used
DIDS {4,4'-diisothiocyanostilbene-2,2'-disulfonate), HEPES (N-2-
hydroxyethyl piperazine-N'-2-ethanesulfonic acid), MES (2-(N-morpholino)ethanesulfonic
acid), PCMBS (p-chloromercuribenzene sulfonate), Tris
(tris(hydroxymethyl)aminomethane), TMA-OH (tetramethylanimonium), Sigma 104®
phosphatase substrate and valinomycin were purchased from Sigma (St. Louis, MO). DEP
(diethylpyrocarbonate) and p-nitrophenol were kindly provided by Dr. Vadivel Ganapathy.
BCECF acid (2',7'-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein) and diBAC (bis
(1,3-dibutylbarbituric acid)pentamethine oxonol) were obtained from Molecular Probes
(Eugene, OR). The membrane filters were obtained from Millipore corporation (Bedford,
MA). All other chemicals used were of analytical grade.
B. Experimental animals
Both male and female Sprague-Dawley rats weighing 200-250g were purchased from Harlan Sprague-Dawley, Madison, Wisconsin. White New Zealand rabbits were
obtained from Shelton Bunny Farm, Waverly, Georgia. All animals were housed in an
environmentally controlled room on a 12h/12h day-night cycle and tap water was given ad
libitum.
C. Experimental procedures
C. l. Preparation of everted sacs of rat intestine
Male and female rats weighing 200-250g were used. Rats were deprived of food
.9. 10 for 24 hours before the experiment. Following ether anesthesia, the abdomen was opened by a midline incision. The intestine was removed from each animal by carefully stripping off the mesentery at the line of its attachment to the intestine. The intestine was immersed in oxygenated saline at room temperature. _Sections of .intestine were removed from the proximal intestine (Wiseman, 1961).
A modification of the method of Crane· artd Wilson ( 1958) was used in this study.
The apparatus (Figure 1) consisted of a plastic'test tube, C, and a one-hole rubber stopper,
B, through which was inserted a glass cannula, E. The needles, A and D, ·were forced through the stopper at slight angles and a length of polyethylene tubing was attached to needle D. One end of a 4-cm segment of ·everted small intestine wa~ closed with a ligature while the opposite end was tied around the lower end of the glass cannula The serosal test solution (2.5 mL) was introduced into the upper portion of the cannula and the sac was filled until the glass cannula contained a column of fluid 1-2 ·cm in height above the level in the outer tube. Hydrostatic pressures greater than about 4 cm of water were avoided. The gas mixture (95% 0 2 and 5% CO2) was continuously passed into the solution, entering at needle D and leaving by way of needle A. The ra~e of gas flow was between 10-to-30 mL/min. The test tube was 'placed in a water bath held at the desired temperature (37°C).
To obtain samples of the serosal fluid at different time points, 0.1 mL of the serosal solution was withdrawn from the middle of the sac using a syringe and polyethylene tubing inserted through the glass cannula.
C.1.a. Effect of pH on migration· of fluoride using everted intestinal ' / sacs
The experim~nt was carried out using proximal everted sacs. Media samples were collected at 5, 10, 15, 30, 60, 90 and 120 minutes. Migration of fluoride was measured in the presence of an inward-directed H+ gradient (serosal pH8 = 7.5, mucosal pHm = 5.5), no H+ gradient (pH8= pHm =7.5) and an outward-directed H+ gradient (pH8= 6.5, pHm = 11
8.0). The everted sacs were immersed in the uptake medium (mucosa! solution) containing
Ringer's solution, 10 mM glucose, 1 mM NaF, 20 mM HEPES/Tris, pH 7.5 or 20 mM
MES/Tris, pH 5.5 or 20 mM Tris/HEPES, pH 8.0. The serosal solution contained
Ringer's solution, 10 mM glucose, 20 mM HEPES/Tris, pH 7.5. 12
Figure 1. Apparatus used/or studies with everted intestinal sacs. A, number 18-gauge needle, blunted; B, size number 3 rubber stopper; C, plastic centrifuge tube (30 mL); D, number 18 needle with polyethylene tubing; E, Pyrex tube. -E
c- 13
C.2. Brush Border Membrane Vesicle (BBMV) .Studies
C.2.a. Preparation of _BBMV
The brush-border membrane vesicles were prepared from rabbit small intestine and kidney cortex by Mg+ 2 precipitation in the presence of EGTA (Miyamoto et al., 1985;
Takuwa et al., 1985; Tiruppathi et al., 1988). After sacrifice of the animal with carbon dioxide, the first half of the small intestine was removed and rinsed with ice-cold 0.9 % saline. The intestine was opened longitudinally. The mucosa was gently scraped off and homogenized in 20 volumes (v/w) of ice-cold 75 mM mannitol, 5 mM EGT A, 20 mM
Tris/NaOH buffer, pH 7.5, for 90 seconds using a INST Blender. One mole/L MgCh was added to the homogen~te to adjust the MgC}i concentration to 10 mM. The homogenate was stirred for 1 minute and allowed to stand for 30 minutes. It then was centrifuged at
3,000g for 15 minutes. The supernatant fraction was then centrifuged at 60,000g for 30 minutes. The pelleted material representing BBMV was suspended in preloading buffer, consisting of320 mM mannitol, 20 mM HEPES/Tris pH 7.5. This suspension was again centrifuged at 60,000g for 30 minutes and the pellet was resuspended in the sa~e buffer using a syringe with a 25-gauge needle. The protein concentration of the final brush border membrane preparation was determined (Lowry et al., 1951) and adjusted to 10 mg/mL with preloading buffer. The brush-border membrane vesicles were now ready for transport studies.
For the preparation of renal BBMV, the cortices from fresh rabbit kidney were dissected from the medulla and subjected to the same procedure described above for intestine except that the homogenization was done for 4 minutes using a INST Blender. The purity of brush-border membrane prepared from small intestine and kidney was assessed by assaying the apical membrane marker enzyme, al~aline phosphatase (Murer et al., 1976; Knickelbein et at, 1983). The alkaline phosphatase activity was enriched 13- fold in the renal BBMV and 11-fold in the small intestinal BBMV compared to the homogenat~. 14
C.2.b. Fluoride uptake studies: general method
The fluoride uptake measurements were carried out using a rapid filtration technique
(Tiruppathi etal., 1988). The filtration system is shown in figure 2. In general, uptake was initiated by addition of 40 µL of the membrane suspension containing 0.4 mg of membrane protein to 160 µL of uptake buffer containing a known concentration of NaF.
The mixture was shaken gently at room temperature in a gyratory shaker. At the end of the incubation period, the uptake process was stopped by adding 5 mL of ice-cold "stop" buffer and then it was filtered immediately through a Millipore filter (p~re size, 0.65 µm).
The filter was washed four times with 5 mL of stop buffer to remove the last traces of uptake buffer while retaining the vesicles. Washing four times was sufficient to remove extravesicular fluoride (see results). Finally, the filter was transferred to a petri dish for fluoride analysis using the ion-specific electrode after diffusion using the hexamethyldisiloxane (HMDS)-facilitated method. Appropriate stop buffers were used based on the different studies. The uptake rate for each time point or group was· calculated from the analysis of BBMV on six separate filters. BBMV were normally suspended in preloading buffer: 20 mM Hepes/Tris, 320 mM mannitol, pH 7.5. The uptake buffer usually.consisted of: 10 mM NaF (8 mM NaF after mixing with the vesicle suspension), 20 mM MES/Tris, 150 mM NaCl, pH 5.0; 20 mM HEPES/Tris, 150 mM NaCl, pH 7.5; or
20 mM Tris/HEPES, 150 mM NaCl, pH 9.0. Mannitol, KCl or other salts replaced NaCl in some experiments as required. Stop buffer was usually the same as uptake buffer except that there was no fluoride and sodium chloride was replaced by potassium chloride.
The HMDS-facilitated diffusion method for fluoride analysis was adopted in this study. The HMDS diffusion technique was developed by Taves (1968) and modified by
Whitford and Reynolds (1979). In the presence of a strong acid, the HMDS molecule 15
Figure 2. Schematic showing the filtration apparatus. ..◄ t---- F i I t e r
Vacuum pump --◄------16
dissociates. A fluoride ion from the sample or standard bonds with the silane radical to
form the volatile trimethylfluorosilane. When this molecule diffuses into the alkaline trap
CH:3 CI-13 CI-13 CI-13 + I I I I CH3-Si-O-Si-CH3 ----►- CI-13-Si-OH + -Si-CI-13 I I I I CI-13 CI-13- CH3 CI-13
(0.05N NaOH), the fluoride exchanges with an hydroxyl ion thus releasing the fluoride
ion with the formation of trimethylsilanol. This occurs by mass action because the
affinities of hydroxyl and fluoride _ions for the silane radical are similar. The major value of
this preparatory method is the fact that the fluoride of the original samples is quantitatively
transferred to a smaller volume and, thus, the analyzed so,lution has a fluoride concentration
well above the electrode's limit of sensitivity.
CI-13 CH3 I I CH3-Si-F + OH ------t•~ CI-13-Si-OH + p I l - CI-13 CH3
C.2.c. Preparation of solutions
DIDS was dissolved in the preloading b1:1ffer and adjusted to 2 or 10 mM in desired
volume.
Fresh DEP stock solution (60 mM) was prepared just prior to use by mixing 0.1
mLof commercially available DEP solution (6.91 M) with 0.1 mL of ethanol and diluting
1 the mixture with 11.3 mL of preloading buffer.
PCMBS solutions were made by dissolving the inhibitor in preloading buffer.
DiBAC was dissolved in dimethylsulfoxide (DMSO) to 9.2 mM as stock solution.
BCECF acid stock solution was prepared by dissolving 1 mg BCECF in 2 mL 17
anhydrous DMSO. This stock solution was divided into small aliqu~ts (200 µL/tube) and
stored at -20°C.
Valinomyciri was dissolved in ethanol. .
C.2.d. Co.ntrol studies
1). To find the number of washings adequate for our experiments, the
effectiveness of washing number on removal of fluoride from the filters was determined in
the presence of an inward-directed transmembrane proton gradient (pHi = 7.5, pH0 = 5.0).
BBMV were prepared from small intestine. Uptake buffer was prepared that contained 10
mM fluoride. The period of incubation was 15 seconds.
2). Magnesium can form an insoluble salt with fluoride. The total magnesium
concentrations in the intestinal BBMV preparations were determined by atomic absorption
· spectrometry (Varian, Model Spectra A-20) using an acetylene flame for magnesium.
C.2.e. Time course of fluoride uptake .
Experiments were done to determine the small intestinal and renal BBMV uptake of
fluoride after incubation periods ranging from 5 seconds to one hour. Uptake was
measured in the presence of an inward-directed transmembrane proton gradient (pHi = 7.5,
pH0 = 5.0), no transmembrane proton gradient (pHi = pH0 = 7.5) and an outward-directed .
transmembrane proton gradient (pHi =7.5, pH0 = 9.0).
C.2.f. Influence of pH on fluoride uptake
These experiments were conducted to determine fluoride uptake by the small
intestinal and renal BBMV in the presence and absence of a transmembrane proton
gradient. To_ determine the effect of H+ per se on fluoride uptake, the uptake rates were
measured at fixed pH values such that intravesicular pH was equal to extravesicular pH. 18
The pH values were 9.0, 7.5 and 5.5. The incubation time was 15 seconds in each study.
The uptake by a fourth group was determined in the presence of a transmembrane pH gradient (pHi = 7.5, pH0 =5.5). A comparison of uptake values when the intravesicular pH was 7.5 in the presence or absence of a transmembrane pH. gradient would show whether the acidic pH had an influence on fluoride uptake or not. The intravesicular pH was altered by preloading the membrane vesicles with buffers of desired pH. The preloading procedure involved incubation of the membrane vesicles with desired buffer at room temperature for 1 hour, followed by centrifugation and suspension of the resulting membrane pellet in a small volume of the ~ame buffer. It was assumed that the pH of the intravesicular medium became equal to the pH of the incubation buffer by this procedure
(Ramamoorthy et al., 1993).
C.2.g. Influence of medium osmolarity on fluoride uptake
In order to evaluate the possibility of fluoride binding to the small _intestinal and renal brush border membrane vesicles (as opposed to transport or migration into the intravesicular space), equilibrium uptakes (one hour) were studied in the presence of a transmembrane pH gradient (pHi =7.5, pH0 =5.5) and a range of extravesicular buffers in which the mannitol concentration was varied to give a wide range of osmolarities (100-900 mosmol/L). As the medium osmolarity increases, the intravesicular space or volume shrinks. The extrapolation of the curve relating 1/osmolarity to fluoride uptake would
• ' • 0 reveal no fluoride uptake provided that it does not bind to the membranes since there would be no intravesicular volume at infinitely high osmolarity. The possibility of binding was determin~d by plotting fluoride uptake at equilibrium against 1/osmolarity.
C.2.h. Effects of· DIDS and DEP on fluoride ·uptake
The effects of treatment of the small intestinal and renal BBMV with DIDS and DEP on fluoride uptake in the presence of a transmembrane pH gradient (pHi = 7.5, pH0 = 5.5; 19
15 seconds incubation) were examined. Immediately after addition of DIDS, the vesicle suspension was mixed and incubated at room temperature for 60 minutes.
BBMV were treated with DEP in different concentrations in total volume 5 mL.
Aliquots of the DEP stock solution were added to the membrane suspension to give final
,concentrations of 1 and 5 mM. Immediately after the addition of DEP, the suspension was
~gitated to ensure complete mixing and incubated at room temperature for 10 minutes
(Miyamoto et al., 1986). During DEP treatment, the ethanol concentration in membrane suspension did not exceed 0.5%. An equal concentration of ethanol was added to the control BBMV suspension which was taken through various steps similar to the DEP treated membrane vesicles.
The reaction with DEP was terminated by addition of 25 mL of ice-cold preloading buffer. The excess unreacted reagent was removed by centrifugation at 60,000g for 30 minutes. The resulting membrane pellets were washed twice with and finally suspended in preloading buffer whose composition was varied depending upon the individual experiment. Throughout these experiments, the protein concentrations in these suspensions were adjusted to 10 mg/mL.
C.2.i. Effect. of diBAC on fluoride uptake
Knauf-et al. (1995) studied the effect of diBAC (50 to 400 nM) on anion exchange using human red blood cells and concluded that diBAC is the most potent known inhibitor of band 3-mediated anion exc~ange. To test wheth~r the diBAC ~nhibited fluoride transport
I . . . . . or not, 1, 10, and 50 µM diBAC were used to determine the effect.of diBAC on fluoride transport. The membrane vesicles ·from the small intestine and kidney were treated with diBAC on the ice for 10 minutes and washed free of the reagent after incubation. The uptake rate of fluoride was measured in the presence of a pH gradient (pHi = 7.5,
pH0 =5.5). The incubation period was 15 seconds. 20
C.2.j. Effect of PCMBS on fluoride uptake
PCMBS, a specific sulfhydryl-modifying reagent, is often used to inhibit water movement through aquaporins_and to characterize the p-aminohippurate transporter (~okol etal., 1986) and tetraethylammonium transporter (Hori et al., 1987; Katsura et al., 1992).
In this experiment, PCMBS was used to study whether ~CMBS had an effect on fluoride transport or not. PCMBS was dissolved in the preloading buffer (320 mM mannitol, 20 mM HEPES/Tris, pH 7.5) and then diluted into membr~ne suspensions to final concentrations of 0.1, 0.25, 0._5, 1.0, 5.0 and 10.0 mM. · Membrane vesicles were preincubated with PCMBS for 30 minutes. Uptake buffers contained 20 mM MES/Tris,
150 mM NaCl, 10 mM NaF, pH 5.0. The incubation period was 15 seconds and then the uptake of fluoride was stopped as described above.
C.2.k. Effect of DIDS, DEP, PCMBS, and diBAC on transmembrane pH gradients
Preliminary results indicated that fluoride uptake by BBMV is enhanced when pHi
was greater than pH0 and that uptake was negligible in the absence of a pH gradient. Because the rate of fluoride uptake in the experiments with DIDS, DEP, PCMBS, and diBAC was determined in the presence of a pH gradient, it was _considered important to determine the effects of pretreatment with DIDS, DEP, PCMBS, and diBAC on the magnitude of the initial pH gradient. If pretreatment of the BBMV with these compounds reduced the gradient and the effect was not detected, the reduction of uptake could be mistakenly attributed to an effect on membrane carriers. Therefore, the dissipation rate of the pH gradient across the brush-borde:r membrane, if any, was estimated using a recording spectrofluorimeter and a pH-sensitive fluorescent dye BCECF acid (2',7'-bis-(2- carboxyethyl)-5-(and-6)-carboxyfluorescein), a compound commonly used to estimate transmembrane pH gradients. The membrane vesicles were loaded with BCECF stock· solution: the membrane 21 vesicles were added to the tube containing BCECF acid; final concentration of BCECF acid was 96 µM. Since BCECF acid could not cross the biological membrane, the following procedure was used to load BCECF acid into the vesicles: the vesicles mixed with BCECF acid were frozen in liquid nitrogen for 20 minutes, then thawed with warm water. The
. . . . freeze-thaw procedure was repeated two more times. · It was assumed that during the freeze-thaw process, the vesicles would be broken and resealed so that BCECF acid would
be loaded into the vesicles·. We obtained good BCECF signals using this method.
Extravesicular BCECF was removed by centrifugation at 60,000g for 30 minutes. The
. . . pellets were resuspended in preloading buffer and centrifuged again. The resulting membrane pellets were washed twic_e with preloading buffer and finally suspended in
preloading buffer. As usual, the protein concentrations in these suspensions were adjusted to 10 mg/mL. The membrane vesicles were placed in a cuvette to obtain a baseline
fluorescence signal. DIDS, DEP, PCMBS and diBAC were then added to the cuvette at final concentrations of 5 mM, 1 mM, 1 mM and 50 µM, respectively. Fluorescence signals· were measured for up to 6 hours with a dual-wavelength spectrofluorimeter (Photo
Technologies International, South Brunswick, NJ) and using excitation wavelengths of 440 and 490 nm and an emission wavelength of 530 nm (Gasalla-Herraiz etal., 1995).
C.2.1. F·-c1- and/or F·-HC03 • exchange study The purpose of these experiments was to determine whether there is F--c1- and/or
p--HC~-exchange in the renal and small intestinal brush-border membrane vesicles. The
experiments were carried out in the absence of a transmembrane pH gradient (pHi = pH0 = 7.5) and with 15-second incubation periods. Uptake buffers for three groups were the
same: 20 mM HEPES/Tris, 10 mM NaF, 300 mM mannitol, pH 7.5. Preloading buffers
were different for the different groups. The Group A buffer contained 50 mM potassium
gluconate (control), 20 mM HEPES/Tris, 220 mM mannitol. The buffer for Group B
contained 50 mM KCI, 20 mM HEPES/Tris, 220 mM mannitol to test for F--CI- exchange. 22 ,
The Group C buffer contained 50 mM KHC~, 20 mM HEPES/T~s, 220 mM mannitol to test F--HC~- exchange.
C.2.m. Fluoride conductance pathway study
To test for the presence of a fluoride ion conductance pathway in BBMV, fluoride uptake was measured in response to a K+ diffusion potential induced with valinomycin.
This ionophore increases the K + conductance of biological membranes. When there is a inward-directed K+ gradient, the intravesicular potential becomes more positive in the presence of valinomycin. The inside positive potential may attract the negatively charged fluoride ion into the vesicles through the conductive pathway.
The experiments with small intestinal and renal BBMV were carried out in the
absence of a transmembrane pH gradient (pHi = pH0 = 7.5). Incubation times were 15 seconds. Prepared valinomycin solution was added to vesicle suspensions. Final concentrations in the incubation mixture of valinomycin and ethanol were 10 µg/mg protein and 0.1 %, respectively. The composition of the uptake buffers for all groups were the same: 20 mM HEPES/TMA-OH (tetramethylammonium hydroxide), 10 mM NaF, 100 mM potassium gluconate, 100 mM mannitol. Preloading buffers for the five groups were as follows: group A, 20 mM HEPES/TMA-OH, 100 mM potassium gluconate, 120 mM mannitol and valinomycin (no K+ gradient); group A', 20 mM HEPES/TMA-OH, 320 mM mannitol and valinomycin (K+ gradient); group B, 20 mM HEPES/TMA-OH, 100 mM potassium gluconate, 120 mM mannitol, 2 mM DIDS and valinomycin (no K+ gradient plus DIDS); group B', 20 mM HEPES/TMA-OH, 320 mM mannitol, 2 mM DIDS and valinomycin (K+ gradient plus DIDS); group C, 29 mM HEPES/TMA-OH, 100 mM potassium gluconate, 120 mM mannitol (no K+ gradient or valinomycin); group C', 20 mM HEPES/TMA-OH and 320 mM inannitol {K+ gradient and no valinomycin) . . · : . The following table shows a simplified scheme for the composition of the preloading buffers: 23
K gradient Valinomycin DIDS Grou:12 A + A' + + B + + B' + + + C C' +
When there is an outward-directed K + gradient, the intravesicular potential becomes more negative in the presence of valinomycin. The inside negative potential may repel the negatively charged fluoride ion into the vesicles through the conductive pathway. The experiment was done in the absence of a pH gradient (pHi = pH0 = 7 .5). Membrane vesicles were prepared from small intestine. Preloading buffers for the three groups were as follows: control group and inside-negative group, 20 mM HEPES/TMA-OH, 100 mM potassium gluconate, 120 mM mannitol and valinomycin (100 mM intravesicular K+); in,side-positive group, 20 mM HEPES/TMA-OH, 320 mM mannitol and valinomycin (no intravesicular K + ). Uptake buffers for the three groups were as follows: control group and inside-positive group, 20 mM HEPES/TMA-OH, 10 mM NaF, 100 mM potassium gluconate and 100 mM mannitol (100 mM extravesicular K+); inside-negative group, 20 mM HEPES/TMA-OH, 10 mM NaF, 300 mM mannitol (no extravesicular K +).
The following table shows a simplified scheme for the composition of the uptake buffers:
·. K gradient Valinomycin Grou:12 Control + Inside+ + (inward) + Inside - + (outward) +
C.2.n. Effect of sodium ·or potassium gradient on fluoride uptake
The transport systems for many amino acids and sugars are sodium dependent. 24
These studies were done to determine the effects of a sodium or potassium gradient on small intestinal and renal BBMV uptake of fluoride. Fluoride uptake was measured with
and without a sodium or potassium gradient in the presence (pHi = 7.5, pH0 = 5.5) and in
the absence (pHi =pH0 = 7.5) of an inward-directed H+ gradient. The membrane vesicles were preloaded with 320 mM mannitol, 20 mM HEPES/Tris, pH 7.5. Uptake buffer A
(H+ gradient) contained 20 mM MES/Tris, 10 mM NaF, 150 mM NaCl (with sodium),
150 mM KCl (with potassium) or 300 mM mannitol (without sodium or potassium), pH
5.5. Uptake buffer B (no H+ gradient) contained 20 mM MES/Tris, 10 mM NaF, 150 mM
NaCl (with sodium), 150 mM KCl (with potassium) or 300 mM mannitol (without sodium or potassium), pH 7.5. The incubation period wa~ 15 seconds.
D. Data analysis
Uptake measurements under each condition were usually done by analyzing the fluoride on each of six different filters. The results are expressed as the mean ± SE. Statisti~lly significant differences were determined with the unpaired Student's t test, analysis of variance and Fisher's PLSD post hoc test. RESULTS
1. Control studies
la. Figure 3 shows. the effect_ of the n~mber of washings on the removal of fluoride from the filters. The study was done in the presence of an inward-directed transmembrane proton gradient (pHi = 7.5, pH0 = 5.0). The recovery of fluoride was reduced to a constant amount after four washings. This indicated that washing four times with 5 mL of "stop" buffer was adequate to remove virtually all extravesicular fluoride from the filters.
1b. Table I shows the concentration of magnesium and in the MgF2 product in two different membrane preparations. The magnesium concentration was determined using atomic absorption spectrometry. With two different intestinal BBMV preparations, the magnesium concentrations were 0.136 µmoles and 0.105 µmoles/mg of protein. The fluoride concentration was 8 mmoles/L in both cases. The ion products of MgF2 were 1.09 x 10-8 and 8.4 x 10-9 in the vesicle suspension ( 10 mg protein/mL) before adding to the uptake buffer. The ion products of MgF2 shown in the table I were the values after the vesicles were mixed with uptake buffers. These products of MgF2 were slightly less than the solubility product constant (7.1 x 10-9).
25 26
Figure 3. The effectiveness of number of washings on removal offluoride from the filters
in the presence of an inward-directed transmembrane proton gradient (pHi = 7.5~ pH0 =
5.0). Membrane vesicles were preparaedfrom small intestine and preloaded with 320 mM mannitol, 20 mM HEPES/Tris (pH 7.5). Uptake offluoride (10 mM in uptake medium) was measured after a 15-s incubation. 40
..-. tn
Lt) ,- 30 ...... (1) ■ = ~ca ..,(1) +■I 0 C. a.. ::, C. u. 20 C, ...... E tn (1) 0 E rn .._.C
0 -t-...-..--...-.,...... ,,...... ,----,.--.----.---.---.------.-...-...... -...---.,...... ,- - 1 0 2 3 4 5 6 7 8 9 10
Number of washings 27
Table I. The concentration ofmagnesium and the MgF2 ion product in different membrane preparations Preparation [Mg2+] [Mg2+] [FJ2
(moles/L) Ion product
1 1.36 x 10-3 2.18 X 10-9
2 1.05 X 10-3 1.68 X IQ-9 28'
2. Time course of fluoride uptake
The time course of fluoride uptake by rabbit small intestinal and renal brush-border membrane vesicles was studied under three different experimental conditions: (1) in the
presence of an inward-directed H+ gradient (pHi = 7.5, pH0 = 5.0); (2) in the absence of
H+ gradient (pHi = pH0 = 7.5); and (3) in the presence of an outward-directed H+ gradient
(pHi = 7.5, pH0 = 9.0). The uptake rates of fluoride by small intestinal and renal BBMV were rapid in the presen~e of an inward-directed H+ gradient and they exhibited the overshoot phenomena, indicating transient concentrative accumulation of fluoride inside the vesicles (Figure 4, 5). In the both cases, the initial uptake of fluoride was very fast. The uptake peaked at 15 seconds and then decreased toward an apparent equilibrium or steady state uptake at 60 minutes. The intravesicular fluoride concentrations at the overshoot were
2 times the equilibrium in renal BBMV and 2.3 times equilibrium in the small intestinal
BBMV. 29
Figure 4. Time course of fluoride uptake by small intestinal BBMV. Brush-border membrane vesicles were preloaded with 320 mMmannitol, 20 mM HEPES/Tris and pH
7.5. Uptake offluoride ( 10 mM) was measured from an uptake medium containing 150 mM NaCl, 10 mM NaF; 20 mM MES/Tris ( pH 5.0 ), or 20 mM HEPES/Tris ( pH 7.5) or 20 mM Tris/HEPES ( pH 9.0 ). Final pHs of uptake medium after mixing were 5.5 for pH 5.0 of original uptake medium and 8.5 for pH 9.0 of original uptake medium. 12
-■ -pH 5.0 -•-pH7.5 10 -•-pH 9.0
C ·a3 0.a. 8 C) E VJ -(1) 0 6 E .s (1) ~ ro 4 ~ 0.. ::, LL
2
0
1--.....--.----.--r--.....---r-~--,-----r----.--~~--r-~-....-""T"'""~--,---,l I I I I 0 1 2 3 4 5 6 7 8 55 60 65 Time (min) 30
Figure 5. Time course of fluoride uptake by renal BBMV. Brush-border membrane vesicles were preloaded with 320 mM mannitol, 20 mM HEPES/Tris and pH 7.5.. Uptake offluoride ( 10 mM) was measured from an uptake medium containing 150 mM NaCl, 10 mM NaF, 20 mM MES/Tris ( pH 5.0 ), or 20 mM HEPES/Tris ( pH 7.5 ) or 20 mM
Tris/HEP ES ( pH 9.0 ). Final pHs of uptake medium after mixing were 5.5 for pH 5.0 of original uptake medium and 8.5 for pH 9.0 oforiginal uptake medium. 12 -•-pH 5.0 T -•-pH 7.5 10 • -A-pH 9.0 -·cuC ' ..... 0 8 ~a. C) E (/) T -Q) 6 ~ 0 E C ~-I -Q) ~ 4 ...... ro a. :J u. T 2
0 ~ _l_ {. ,i l I I I 0 1 2 3 4 5 6 7 8 55 60 65 Time (min) 31
3 .. Influence of pH and transmembrane pH gradients on fluoride uptake
In order to ob~in specific information on the role of H+ concentrations per se, fluoride uptake was compared in the presence and absence· of a transmembrane proton gradient. The fluoride uptake rates with the small intestinal and renal BBMV were determined after adjusting the pH of both the intravesicular and extravesicular solutions to the same values (pH= 9.0, 7.5 or 5.5). As shown in Table 2, the uptake rates under these conditions by the small intestinal BBMV were close to zero. The study with the renal
BBMV produced similar results (table 3 ). The differences in fluoride uptake rate among different intravesicular pH groups in the absence of pH gradient had no statistical significance (p>0.05). The uptakes in both cases, however, were stimulated many-fold in
the presence of an inward-directed H+ gradient (pHi = 7,?, pH0 = 5.5). The uptakes in the presence of an inward-directed H+ gradient were higher than those in the absence of an
inward-directed H+ gradient (p<0.01). The results indicated that pHi or pH0 had no effect on fluoride uptake when there was no pH gradient. '32
Table II. Influence ofpH per se and a transniembrane pH gradient on fluoride uptake in small intestinal BBMV. Values are means ± SE. Membrane vesicles were preloaded with
320 mMmannitol, 20mM Hepes/Tris (pH 7.5), 20 mM Tris/HEPES (pH 9.0) or 20 mM
MES/Tris (pH 5.5). Uptake offluoride ( 10 mM in uptake medium) was measured after a
15-s incubation. Extravesicular pH values are the final pH values of the uptake media after mixing the membrane vesicles with the uptake buffer. *p<0.01 compared to other three groups. H Fuptake
Intravesicular pH Extravesicular pH Gradient (nmoles/mg of protein/15 s)
9.0 9.0 0 -1.36±0.25
7.5 7.5 0 -0.09±0.62
5.5 5.5 0 0.26±0.47
7.5 5.5 2.0 4.59±0.73 * 33
Table Ill. Influence ofpH per se and a transmembrane pH gradient on fluoride uptake in renal BBMV. Values are means ± SE. Membrane vesicles were preloaded with 320 mM mannitol, 20 mM HEPES/Tris (pH 7.5), 20 mM Tris/Hepes (pH 9.0) or 20 mM MES/Tris
(pH 5.5 ). Uptake of fluoride (10 mM in uptake medium) was measured after a 15-s incubation. Extravesicular pH values are the final pH values of the uptake media after mixing the membrane vesicles with the uptake buffer. *p<0.01 compared to other three groups. H Fuptake
Intravesicular pH Extra.vesicular pH Gradient !nmoles/m~ of Erotein/ 15 sl 9.0 9.0 0 0.40±0.24
7.5 7.5 0 1.45±0.48
5.5 5.5 0 1.49±0.38
7.5 5.5 2.0 5.67±0.50 * 34
4. Effect of medium osmolarity on fluoride uptake
The effect of medium osmolarity on fluoride uptake by intestinal and renal BBMV were studied by measuring the uptake of fluoride after 60-minute incubation. The extravesicular medium osmolarity was varied by changing mannitol concentrations. A 60- minute incubation period was chosen because the uptake of fluoride into the vesicles reaches an apparent equilibrium or steady state after that amount of time (Figure 4, 5).
Increasing concentrations of mannitol, to which the vesicles are impermeant, were correlated with decreasing uptakes of fluoride (Figure 6, 7). Extrapolation to infinite medium osmolarity (zero intravesicular space) in the case of intestinal BBMV resulted in no uptake. This indic~ted. that fluoride uptake was completely due to transport into intravesicular fluid space rather than binding to the membrane. However, in the case of renal BBMV (Figure 7), there appeared to be some residual uptake of fluoride even at infinite mediurp osmolarity suggesting that a portion of the uptake may have been due to binding to the membranes. This residual ·uptake was about 10% of the total uptake under isosmolar conditions (300 mosm/Kg).- 35
Figure 6. The effect ofmedium osmolarity on F uptake. Brush-border membrane vesicles prepared from small intestine were preloaded with 320 mM mannitol, 20 mM HEPES/Tris and pH 7.5. Membrane vesicles were incubated for 60 minutes at room temperature in uptake media containing 10 mM NaF, 20 mM MES/Tris (pH 5.5) and increasing mannitol concentrations from 100 mM to 800 mM. 6
,._ -::J 5 0 .c
T ...... 4
~ C ctS ! .... 0 C. ,._ 3 ::J C. l1. C') ...... E u, 2 I Cl) I 0 I E I .._..C 1 I . II I I 0 ...:-'------0 2 4 6 8 1 0 12 1 /Osmolarity 36
Figure 7. The effect ofmedium osmolarity on F uptake. Brush-border membrane vesicles prepared from kidney were preloaded with 320 mM mannitol, 20 mM HEPES/Trif and pH
7.5. Membrane vesicles were incubated for 60 minutes at room temperature in uptake media containing 10 mM NaF, 20 mM MES/Tris (pH 5.5) and increasing mannitol concentrations from 100 mM to 800 mM. 6
..-.. I., :::J 5 0 .c
~ ...... 4 C: Cl) Cl) ~ca ...... 0 a. I., 3 :::J a. LL C, ...... E u, 2 Cl) 0 E ._.C: 1
0 0 2 4 6 8 1 0 12
1 / O_ s m o_ I a r i t y 37
5. Effects of DIDS and DEP on fluoride uptake
_The effects of DIDS and DEP on fluoride uptake by the small intestinal BBMV were studied (table 4). In these experiments, there was an inward-directed H+ gradient
(pHi = 7.5, pH0 = 5.5). The vesicles were pretreated with DIDS and DEP for 60 and 10 minutes, respectively. The resµlts showed that uptake of fluoride into vesicles treated with
2 mM or 10 mM DIDS decreased 18 and 42% compared with control group. Fluoride uptake into vesicles treated with 1 mM or 5 mM DEP decreased 39 and 42% compared with control group. The fluoride uptake was decreased by 50% when the vesicles were treated with a combination of 2 mM DIDS and 1 mM DEP.
Table 5 shows the effects of DIDS, DEP and PCMBS on fluoride uptake by the
renal BBMV in the presence of a transmembrane proton gradient (pHi = 7.5, pH0 = 5.5). Uptake by vesicles pretreated with. 2 mM DIDS or 1 mM DEP was decreased by 56% compared with the control group. 38
Table IV. The effect of DIDS and DEP on fluoride uptake in small intestinal BBMV.
Values are means± SE. Membrane vesicles were preloaded with 320 mM mannitol, 20 mM HEPES/Tris (pH 7.5) and treated with DIDS for 60 minutes or DEP Jot 10 minutes.
Uptake offluoride was measured after 15-s. The uptake buffer contained 20 mM MES/Tris
(pH 5.0), 150 mM NaCl and 10 mM NaF. ,.)
Group Fuptake % Reduction
(nm.oles/ mg of protein /15-s)
Ethanol control 6.42±0.47 0
DIDS(lOmM) 3.82±0.30 40
DIDS (2mM) 5.27±0.30 18
DEP(5mM) 3.89±0.55 39
DEP(lmM) 3.74±0.08 42
DIDS (2mM)+DEP(l mM) 3.22±0.29 50 39
Table .V. The_ effect of DIDS, DEP mid PCMBS on fluoride uptake_ in renal BBMV.
Values are means± SE. Membrane vesicles were preloaded with..320 mM mannitol, 20 mM HEPES/Tris (pH 7.5) and treat¢d with DIDS for 60 minutes, DEP for 10 minutes and
PCMBS for 30 minutes. Uptqke offluoride was measured after 15-s. The uptake buffer contained 20 mM MES/Tris (pH 5.0), 150 mM NaCl and 10 mM NaF. Group Fuptak:e % Reduction
(nm.oles/ mg of protein /15-s)
Ethanol control 7.70±0.31 0
DIDS(2mM) 338±0.43 56
DEP(l mM) 3.36±0.29 56
PCivIBS ( 1 mM) 1.67±0.48 78 40
6. Effect of PCMBS on fluoride uptake
PCMBS, a specific sulfhydryl-modifying reagent, is often used to inhibit water movement through aquaporins and to ~haracterize p-aminohippurate transporter (Sokol et al., 1986) and tetraethylammonium transporter (Hori et al., 1987; Katsura et al., 1992). In· this experiment, PCMBS was used to study whether PCMBS had an effect on fluoride transport or not. Small intestinal BBMV were treated with different concentrations of
PCMBS for 30 minutes and then washed free of the reagent.: Fluoride uptake was measured as described above. H+ gradient-dependent uptake was inhibited by PCMBS in a dose-dependent manner (figure 8). The rate of fluoride uptake was inhibited by 34% when the PCMBS concentration was 0.1 mM and by 81 % when the PCMBS concentration was
5.0 mM. The inhibition of uptake when the PCMBS concentration was 10.0 mM was not significantly different from that at 5.0 mM.
The uptake rate of fluoride by renal vesicles treated with 1 mM PCMBS was decreased by78% (table V). 41
Figure 8. The effect of PCMBS on fluoride uptake in the presence of pH gradient - (pHi=7.5, pH0 =5.5). Membrane vesicles prepared from small intestine were treated with
PCMBS for 30 minutes at room temperature before adding the uptake buffer. 10 -"' ll) T"" ""'C: 8 Cl) 0 -11.. C.
0) 6 E ""'u, Cl) 0 E 4 -C: Cl) ~ ca -C. 2 ::I
LL
0 0 0.1 0.25 0.5 1.0 5.0 10.0
PCMBS,mmoles/L 42
7. Effect of diBAC on fluoride uptake
The small intestinal and renal BBMV were treated on the ice with 1, 10, and 50 µM diBAC for 10 minutes. The excess unreacted reagent was removed by centrifugation at
60,000g for 30 minutes. Measurements of fluoride uptake were carried out in the presence of an inward-directed H+ gradient (pHi =7.5, pHo = 5.5). Figures 9 and 10 show that diBAC had no effect on F uptake in either system. 43
Figure 9. The lack of an effect of diBAC on fluoride uptake in small intestinal BBMV in the
presence ofpH gradient (pHi=7.5, pH0 =5.5). The vesicles were treated with diBAC for 10 minutes on the ice. rn -(/J ,....LO ..... C: 8 ..•'?J? Q) '1.... :ff;' 0 -I.. C. 6 :1 C) .....E ~~l (/J ...-:/J.:: Q) --::-~ 0 E 4 -C: ~ Q) ~ m 2 C. -::::s
LL 0 C 1µM 10µM SOµM 44
Figure 10. The lack of an effect of diBAC on fluoride uptake in renal BBMV in the
presence ofpH gradient (pHF7.5, pH0 =5.5). The vesicles were treated with diBAC for 10 minutes on the ice. 8
-en LC) 6 ,- ...... ·-C: (1) ! ~ 0 ca ii,,, 'E.. C. 4 ::J u. C) E ...... en (1) 0 E 2 .._..C
0-+----~- C 1µM 45
8. Effect of DIDS, DEP, PCMBS and diBAC on transmembrane pH gradient
All previous findings indicated that fluoride uptake by BBMV occurred only when
pHi was greater than pH0 and that uptake was negligible in the absence of a pH gradient.
Because the rates of fluoride uptake in the experiments with DIDS, DEP, PCMBS, and diBAC were determined in the presence of a pH gradient, we considered the possibility that the apparent inhibitions of fluoride uptake by DIDS, DEP and PCMBS might be artifactual and secondary to an increased H+ permeability with an enhanced rate of dissipation of the imposed H+ gradient. Such an effect could indirectly reduce the rate of fluoride uptake, which is markedly sensitive to the transmembrane H+ gradient. Therefore, it was important to determine the effects of pretreatment with DIDS, DEP, PCMBS, and diBAC on the magnitude of the initial pH gradient. If pretreatment of the BBMV with these compounds reduced the pH gradient and the effect was not taken into account, the reduction of uptake could be mistakenly attributed to membrane carriers.
The dissipation rate of the pH gradient across brush-border membrane, if any, was estimated using a recording spectrofluorimeter and BCECF acid, a pH-sensitive fluorescent dye commonly used to estimate transmembrane pH gradients. The membrane vesicles were pretreated with the inhibitors and then loaded with BCECF. The results showed that intravesicular pHs with all four tested ~eagents in all treated groups were lower than that of the pH 5.0 control in both the small intestine and kidney BBMV (figure 11).
Theoretically,_ it is impossible that intravesicular pHs with DIDS, DEP, PCMBS and diBA C treated groups could be lower than that of pH 5. 0 contml. In order to clarify this, we tested the possjbility thatDIDS, DEP, PCMBS and diBAC may directly react with
BCECF to decrease the signals. To examine this possibility, we measured effects of
DIDS, DEP, PCMBS and diBAC on BCECF signal without BBMV in pH 7.5 preloading buffer. Figur~ 12 shows the effects of DIDS, DEP, PCMBS and diBAC on BCECF signal. Each panel consists of three tracks. The first track represents 0-500 second 46 periods. The second track represents the same sample after three hours. The third track represents the same sample after six hours. In panel A (coritrol), addition of 20 µLand
100 µL of solvent (ethanol) in track 1 and 2 decreased BCECF signal due to the dilution.
The signal was still stable after 6 hours (track 3)., In panel B, DEP quenched BCECF signal immediately after addition of 20 µL of 1 mM DEP in track 1. The signals of three hours and six hours were the same. This suggested that the DEP quenching of the
BCECF signal reached a maximum at 3 hours. Actually, the signal was reduced to close to zero in 30 minutes (data not shown). In panel C, 100 µL preloading buffer and 100 µL
PCMBS were added. The signal was decreased in 3 and 6 hours (track 2 and 3). In panel
D, Dl~S quenched the signal immediately, since the decrease of signal at adding 100 µL
DIDS was more than._that_ at adding 100 µL pre_loading buffer. The-result of diBAC was similar to the DIDS group (panel E). The quenching potency of the reagents was in this order: DEP>PCMBS>DIDS>diBAC.
T~ese results suggested that the ~nhibit~rs.. may have quenched BCECF signal even in the abs~nce of vesicles. This may have_ been *e reason that the fluores_cent signals of
BBMV pretreated with the inhibitors were lower than control ~ignals (figure ll). To test this possibility, the following experiment was done to determine the effect of DIDS, DEP,
PCMBS and diBAC on the transmembrane pH gradient. The membrane vesicles were loaded with BCECF and placed in a cuvette to get a baseline signal. DI:OS, DEP, PCMBS and diBAC were then added to the cuvette to achieve final concentrations of 5 mM, 1 mM,
1 mM and 50 µM, respectively. The results are shown in figure 13.
Panel A shows standard curve. BBMV were added to buffers at pH 7.5 (track 1), 6.0 (track 2) or 5.0 (track 3). Panel B shows that fluoride in the uptake buffer.had no effect on the imposed pH gradient compared to nonfluoride uptake buffers._ Panel C and D tested the effects of inhibitory agents on transmembrane pH gradient. The signal of the diBAC group was slightly reduced when diBAC was added to the cuvette (panel C, track
1) and continued declining for 2 hours (panel D, track 5). The signals of DIDS, DEP and 47
· PCMBS groups were unchanged or only slightly decreased. The magnitude of fluorescence change induced by adding pH 5.0 uptake buffer in 2 hours was similar in control and DIDS-, DEP-, PCMBS- and diBAC-treated membrane vesicles (panel D).
These results indicated that diBAC slightly reduced intravesicular pH prior to adding pH
5.0 uptake buffer and that DIDS, DEP and PCMBS_ had little or no effect on intravesicular pH prior to adding pH 5.0 uptake buffer, and did not facilitate dissipation rate of imposed pH gradient. 48
Figure 11. Effect of DIDS, DEP, PCMBS and diBAC on transmembrane pH gradient.
Membrane vesicles prepared from intestine (panel A) and kidney (panel B) were treated with DIDS, DEP, PCMBS and diBAC in pre loading buffer (pH 7.5 ). Then the membrane vesicles treated with the reagents were loaded with the pH-sensitive dye BCECF (96 µM).
The fluorescence was measured at 440/530 nm (excitation/emission). The membrane vesicles were incubated in pre loading. buffer containing 10 µM nigericin and then exposed to pH 5.0 uptake buffer ( 10 mM NaF, 20 mM HEPES/Tris, 300 mM NaCl ). In panel A, track 1 is control; track 2 represents the vesicles treated with diBAC; track 3 represents the vesicles treated with DEP; track 4 represents the vesicles treated with PCMBS; track 5 represents the vesicles treated with DIDS. In panel B, track 1 is control; track 2 represents the vesicles treated with diBAC; track 3 represents the vesicles treated with PCMBS; track
4 represents the vesicles treated with DIDS; track 5 represents the vesicles treated with
DEP. A ,. 200 I I xl0:3
Ill □ 2 J ••.,._:,•,•J. •. , ....1 •-•••··/·,.•','('• •""•\,,•\r:,,ll•,,,•~--.,•·•.•-.•.,.,• ..•.,.'-,"•,' ► .''l'1lf"•,•'i,.··,,•..,\.•.,.•1n••,,,.•._ •• ,\. •-'"• ',1, ,,,-,,,.,_...... ,,,•.,·• 4 •,•,•• ...... , ....., ..••,\~r··· '•\•'•'•'··'.--..,•..,,,.... ,•,,...-·,/'•.·,,J .,,.,.. _\'4_.•·-.r•••l-•,.,r,•1._r·•'••" -'-·.•t·,,i. ..,,, ►'•"\•• 1 ...... -.,1.,,1-\J-• ·-••'· \ ... , 5
a-
-'.j(J -----'------"•L-·------'------L-----=--__;-...i'L----__l {) 1()0 2()(1 ]00 Ti me (seconds)
B 15fl
,, J r·',',"r/\1.• •,• ... _.',• -·-· 'r ...... ,,-1~•-/•,1.•·.•1··•,v-,1, .•,.~ .... -.. ,f/1\-..•'·,·,•·:'···,•t'' ',,. ..- •.,.\v,_ ...... ,.,, ''if·•-•\"•:
llllJ i ! 2
-SU ______....__--,-.____-----1. __ ~__,---1- __ 0 100 200 JOO Ti1ne (secnnds) 49
Figure 12. Effect of DIDS, DEP,, PCMBS and diBAC on BCECF fluorescence signal.
The experiment was done in pH 7.5 preloading buffer only (no vesicles). 25 µL BCECF
(96 µM) was added to 2 mL preloading buffer in the cuvette and followed by the addition ofthe different reagents, ethanol and pre loading buffer. A B
20 ul ethanol xlO.; X 10.; 20 ul DEP 100 ul ethanol I 3 i ' -----~1' 1 __ ,., ... 2 2
0 0
-I -l.__,___ ~-~1---'---..__--1,1-J...--_,______J ,0 250 3.00 3.04 6.00 6.04 0 250 3.00 3.04 6.00 6.04
Time (second) Time (hour) Time (second) Time (hour) C D 4 4 100 ul preloading buffer xlO,; 100 ul preloading buffer 3 ! ,, ·100 ul P018S 3 * .,,,.--· 100 ul 0 IDS 77¥ '-,, ¥'
2 2
Ji-
0 0
-I -1.__.______,,f-'----'---lr_..______, 0 250 3.00 3.04 6.00 6.04 0 250 3.00 3.04 6.00 6.04
Time (second) Time (hour) Time (second) Time {hour)
E
100 ul preloading buffer
3 ~ / 100 ul d iBAC -- \_· ·-- __ .,, .... 2
0
-t.___.___ _.__ ___,~~------~,1-i------' 0 250 3.00 3.04 6.00 6.04
Time {second) Time {hour) 50
Figure 13. Effect of DIDS, DEP, PCMBS and diBAC on the transmembrane pH gradient.
Membrane vesicles were prepared from small intestine and preloaded with 320 mM mannitol, 20 mM HEPES/Tris and pH 7.5. The vesicles were then loaded with the pH sensitive dye BCECF (96 µM) and placed in a cuvette. The membrane vesicles were then exposed to pH_ 5.0 uptake buffer (JO mM NaF, ~? mM MES!Tri_s, 300 mM NaCl ). Fluorescence signals were measured with a dual-wave-:length, spectrofluorimeter using excitation wavelengths of440 and 490 nm and an emission waveleng~h of 530 nm. Panel
A shows the standarq Cl!,rve. Tracks 1, 2 and 3 represent the pH 7.5, 6.0 and 5.0 . '. standard, respectively. · lri panel B, -the drop· offluorescence signal ofirack 1 is the dilution effect. The track 2, 3 and 4 shows the effect offluoride and nonfluoride uptake buffers on the imposed pH gradient. Panel C is the effect of tested agents on transme'l!lbrane pH gradient. DIDS, DEP, PCMBS and diBAC were· added to the cuvette .at arrow. Two hours later (panel D ), the vesicles were_ expo~ed to pH 5.() uptake buffer.· A .B I.U 1.0 xlO,; xlO,;
0.8 '-"'·,\~.;Ir•' ·\11,>,' 0.8 •••.,'\ ►'1f\'r.,"' .,r'',b ~3'1:5f' U.6 · 0.6 I I
U,4 , 1,--~------~"--•-•·-•w I -...... _.: ::--~~..,...... ~'"=""~~~~...... 2. J md 4 0.2 0.2'
0 0
-0. 2 '--'---'---L---'---'---_.__.....______.'-----'----'--' -0 .2 .__ _.___...,__ _ __,__ _ __.__ _ _._ _ _._ _ _._ _ __._ _ _._ 0 100 200 300 400 0 100 200 300 400 T iine (seconds) Time (seconds)
C D
I 1.(1
xlll,; .,.,1-.... , O.IJ '·--:;_:,_ ...,. ..., .... , ..
0.(,
IJ.,1-
0.2 0.2
11 I)
-0. 2L._ _ __.__ _ _L_ __.__ ___:L____ _,____-L._ __._ __'----J -D .ZL._.-1.--~--1.Loo__ ..__-:-20-!-o:------'----:-3c:1-::iu:---____. 0 2.00· 2.03 2.06 2.09 2.12 Time (seconds) Time (hour) 51
9. F·-c1- and F·-HC03 • exchange s~udies To determine whether an anion exchange pathway for fluoride is present in the rabbit renal and small intestinal brush-border membrane vesicles, two anions (CI- and
HC03-) that might exchange with fluoride were tested in the absence of pH gradient (pHi = pH0 = 7.5). As shown in Figure 14 and 15~ an outward c1- or HC03- gradient (50 mM) did not accelerate fluoride influx (p > 0.05), indicating that there was no_F--c1- or F-
HC03- exchange in the small intestinal or renal brush-border membrane vesicles. An outward-directed gluconate gradient was used as a control because gluconate does not cross the membrane. 52
Figure 14. The Cz-JF- and HCO3-JF- exchange study. Membrane vesicles prepared from small intestine were preloaded with 20 mM HEPES/Tris (pH 7.5), 220 mM mannitol and 50 mM potassium gluconate or potassium chloride or potassium bicarbonate. The preloading buffer containing KHCO3 was gassed with 5% CO2-95% 02 during the experiment. Uptake medium contained 10 mM NaF, 300 mM mannitol, 20 mM
HEPES/Tris and pH 7.5. Incubation time was 15 seconds at room temperature. en 2.5 LO ,- -s:::: C1) 0 -1- C.
C') 1.5 E :::: 0 E -s:::: C1) ~ 0.5 ca C. -:J
LL
-0.5 Gluconate Cl- HC03- 53
Figure 15. The CI-IF- and HCO3-/F- exchange study. Membrane vesicles prepared from kidney were preloaded with 20 mM HEP ES/Tris (pH 7.5 ), 220 mM mannitol and 50 mM potassium gluconate or. potassium chloride or potassium bicarbonate. The preloading
buffer containing KHCO3 was gassed with 5% CO2-95% 02 during the experiment.
Uptake medium contained 10 mM NaF, 300 mM mannitol, 20 mM HEPES/Tris and pH
7.5. Incubation time was 15 seconds at room temperature. 3.0 -u, Lt) .....,- 2.5 C: G) -...0 C. 2.0
C, .....E U) G) 1.5 0 E C: 1.0
-G) ~ca -C. 0.5 :::s Li.
0.0 Gluconate Cl - Hcor 54
10. Fluoride conductance pathway study
To test for the presence of a membrane potential-sensitive transport of fluoride ion
in the renal and small intestinal BBMV, fluoride uptake was measured in response to a K +
diffusion potential induced by valinomycin. As an ionophore, valinomycin increases K+
conductance of biological membranes. When there is an inward.:.directed K + gradient,
. valinomycin induces an inside-positive K+-diffusion potential. An insi~e positive potential
could attract fluoride into the vesicles through a conductive pathway. The experiments were
carried out in the abs(}nce of a transmembrane pH gradient (pHi = pH0 = 7.5).
As shown in figure 16, fluoride uptake by small intestinal BBMV in the presence of
an inward potassium gradient (group A') was· higher than that in the absence of a
potassium gradient (group A, p<0.05). Since the uptake of group B' was higher than that
of group B (p<0.05), the increased uptake (group B') due to an inward potassium gradient
was not inhibited by 5 mM DIDS. Comparing group C to A, we found that valinomycin
itself had no effect on fluoride uptake (p>0.05). Comparing group C' to C, we found that
potassium gradient itself did not increase fluoride uptake (p>0.05). In the case of kidney
BBMV (figure 17), the fluoride uptake values in various groups were nearly identical.
That is, none of the variables altered fluoride uptake.
As shown in figure 18, fluoride uptake by the inside-negative potential group was
less than that of control, although the difference was not statistically significant. Auoride . . uptake by the inside-positive was significantly higher (p<0.0S)than that of the control or
inside-negative group. 55
Figure 16. The effect of an inwardly-redirected K+,diffusion potential on fluoride uptake.
Membrane vesicles prepared from small intestine were pre incubated as follows:
A. 20 mM HEPESITMA-OH, 100 mM potassium gluconate, 120 mM mannitol
and valinomycin ( no K+ gradient);
A'. 20 mM HEPESITMA-OH, 320 mM mannitol and valinomycin (K+ gradient);
B. 20 mM HEPESITMA-OH, 100 mM potassium gluconate, 120 mM mannitol, 2
mM DIDS and valinomycin (DIDS without K+ gradient);
B'. 20 mM HEPESITMA-OH,·320 mM mannitol, 2 mM DIDS a11:d valinomycin
(DIDS with K+ gradient);
C. 20 mM HEPESITMA-OH, 100 mM potassium gluconate, 120 mM mannitol
(no K+ gradient or valinomycin);
C'. 20 mM HEPESITMA-DH and 320 mM mannitol (K+ gradient without
valinomycin):
Valinomycin was added to the vesicle suspension (10 µglmg protein). The composition of the uptake buffers for all groups was the same: 20 mM HEPESITMA-OH
(tetramethylammonium hydroxide), 10 mM NaF, 100 mM potassium gluconate, 100 mM mannitol. There was no pH grq,dient in any of these experiments. 3.0 -en Ln ...... ,- Cl) C: ~ ca Cl) C. 0 2.0 -I.. -::::s C. Cl) "C 'i: C) 0 E ::::s ...... en u: Cl) 0 1.0 E -C:
A A 1 B c· 56
Figure 17. The effect of an inwardly-redirected K+ diffusion potential on fluoride uptake.
Membrane vesicles prepared from small kidney were pre incubated as follows:
A. 20 mM HEPESITMA-OH, JOO mM potassium gluconate, 120 mM mannitol
and valinomycin ( no K +gradient);
A'. 20 mM HEPESITMA-OH, 320 mM mannitol and valinomycin (K+ gradient);
B. 20mMHEPESITMA-OH, l00mMpotassiumgluconate, 120mMmannitol, 2
mM DIDS and valinomycin (DIDS without K+ gradient);
B'. 20 mM-HEPESITMA-OH, 320 mM mannitol, 2 mM DIDS and valinomycin
(DIDS with K+ gradient);
C. 20 mM HEPESITMA-OH, 100 mM potassium gluconate, 120 mM mannitol
( no K+ gradient or valinomycin );
C'. 20 mM HEPESITMA-OH and 320 mM mannitol (K+ gradient without
valinomycin);
Valinomycin was added to the vesicle suspension (10 µglmg protein). The composition of the uptake buffers for all groups was the same: 20 mM HEPESITMA-OH
(tetramethylammonium hydroxide), 10 mM NaF, 100 mM potassium gluconate, 100 mM mannitol. There was no pH gradient in any of these experiments. 3
-tn Lt) .....,- C a., 0 -I- a. 2
C, E .....u, a., 0 E -C a., ~ ca a. -::::,
LL
0 A A' B e· C c· 57
Figure 18. Effect of reversal of K + diffusion potential on fluoride uptake in the absence of a pH gradient. Membrane vesicles prepared from small intestine were preincubated with 120 mM mannitol, 20 mM HEPESITMA-OH (pH 7.5), and either 100 mM potassium gluconate for the control and inside-negative group or 200 mM mannitol for the inside positive group. Valinomycin was added to all groups. The· uptake medium ~onsisted of 20 mM HEPESITMA-OH (pH 7.5), 10 mM NaF, 100 mM potassium gluconate and 100 mM mannitol for the control and the inside-positive group and 200 mM mannitol for the in$ide negative group. -en Lt) ,- ....._ 2 C: Q) ~ Q) ca ...... 0 C. ,._ :s C. LL C) E :::: 0 E ._..C:
C Inside - Inside + 58
11. Effect of a sodium or potassium gradient on fluoride uptake
The effects of a Na+ or K + gradient on F uptake by the small intestinal and renal
BBMV were examined in the presence and absence of a pH gradient (Figures 19, 20).
Under pH gradient conditions (panel B, pHi = 7.5, .PHo = 5.5), uptake by the Na+, K+ or mannitol groups was higher than that of the corresponding groups in the absence of a pH gradient ( p<0.01 ). Fluoride uptake in BBMV in the Na+ and K+ groups were not different with statistical significance ( p>0.05 ), but fluoride uptakes in the mannitol groups of both_ intestine and kidney BBMVs were higher than that of 'Na+ and K+ groups
(p<0.01). This may have been due to chloride in the Na+ and K+ groups competing with fluoride for the same transport system. In the absence of a pH gradient .(panel A,
pHi=pH0 =7.5 ), uptake values were lower and were not different with statistical significance ( p>0.05 ). The results showed that F uptake is pH-dependent and is not Na+ or K +-dependent. 59
Figure 19. The effect of Na+ and K+ gradient on F uptake· in the absence ( A,
pHi=pH0 =7.~) and in the presence of a pH gradient (B, pHi=7.5, pH0 =5.5). Membrane vesicles prepared from small intestine were preloaded with 320 mM mannitol, 20 mM HEPES/Tris and pH 7.5. The uptake medium consisted of 10 mM NaF, 20 mM
HEPES/Tris (pH7.5), 300 mMmannitol or 150mM NaCl or 150 mM KCl in tlie absence of pH gradient and 10 mM NaF, 20 mM MES/Tris (pH 5.5), 150 mM NaCl or 150 mM
KCl in the presence ofpH gradient. F uptake (nmoles/mg protein/15 s)
0 I\)
)> z ll)
,3 ll) :::s :::s ;::. 9.
F uptake (nmoles/mg protein/15 s)
--'- --'- 0 (Jl 0 ' 0,
tJJ 60
Figure 20. The effect of Na+ and K+ gradient on F uptake in the absence ( A, pHi=pH0 =7.5) and in the presence of a pH gradient (B, pHi=7.5, pH0 =5.5). Membrane vesicles prepared from kidney were preloaded with 320 mM mannitol, 20 mM
HEPES/Tris and pH 7.5. The uptake medium consisted of 10 mM NaF, 20 inM
HEPES/Tris (pH7.5), 300 mMmannitol or 150 mM NaCl or 150 mM KCl in the absence of pH gradient and 10 mM NaF, 20 mM MES/Tris (pH 5.5), 150 mM NaCl or 150 mM
KCl in the presence ofpH gradient. F uptake (nmoles/mg protein/15 s)
0 I\.)
)>
z O>
3 O> ::J ::J ;:.: 0
F uptake (nmoles/mg protein/15 s)
-I. -I. 0 Ul 0 Ul
IJJ
3 O> ::J . ::J 61
12. Effect of N 0 3 - and CI· on fluoride uptake To further examine the possibility that chloride competes w_ith fluoride for the same transport system, chloride and nitrate were tested for the ability to interfere with the uptake of fluoride in intestinal BBMV. The experiments were done in the presence of an inward
directed proton gradient (pHi=7.5, pH0 =5.5). The uptake buffer contained 10 mM F- and different chloride concentrations (10, 50 and 150 mM CI-) or 150 mM N03-. Figure 21 shows that the uptake of fluoride was inhi~ited by 150 mM N03 - and by 50 and 150 mM CI- compared to mannitol or gluconate control (p<0.05). This suggested t~at N~- and chloride can compete with fluoride for the same transport system. It is possible, however,. that the rapid influx of these ions may have interfered with fluoride uptake via its conductance pathway.
When the experiments were carried out using renal BBMV in the presence of proton gradient, the results were similar to intestinal BBMV (figure 22). 62
Figure 21. Effect of cz- and NO 3- on F uptake in the presence of a pH gradient (pHi= 7.5, pH0 =5.5). Membrane vesicles prepared from small intestine were preloaded with 320 mM mannitol, 20 mM HEPES/Tris and pH 7.5. The uptake medium consisted of 10 mM NaF,
20 mM HEPES/Tris (pH 7.5), 300 mMmannitol, 150 mM potassium gluconate, 150 mM
KNO3 or 150, 50 or 10 mM KCl and isosmotically adding potassium gluconate. 16
14 -U) LC) ,:-...__ C 12 se C. 10 C) ...__E U) 8 Cl) 0 E C 6 -Cl) Jia:: 4 sC. :::> LL 2
0 Potassium 150mM 50mM 10mM Potassium Mannitol Nitrate KCI KCI KCI Gluconate 63
Figure 22. Effect of cz- and NO3- on F uptake in the presence of a pH gradient (pHi=7.5, pH0 =5.5). Membrane vesicles prepared from kidney were preloaded with 320 mM mannitol, 20 mM HEPES/Tris and pH 7.5. The uptake medium consisted of 10 mM NaF,
20 mM HEPES/Tris (pH 7.5), 300 mM.mannitol, -J50 mM potassium gluconate, 150 mM
KNO 3 or 150, 50 or 10 mM KCl and isosmotically adding potassium gluconate. 16
-U) 14 It) T- --C 12 se 0. 10 en ...... E U) .a C1) "'6 E 6 -C ~s 4 0. . ::, u. 2
0 Potassium 150 mM 50 mM KCI 10 mM KCI Potassium Mannitol Nitrate KCI Gluconate 64
13. Effect of pH on migration of fluoride using everted intestinal sacs
The experiment was carried out to examine migration of fluoride using everted sacs from the proximal small intestine of rats. Serosal samples were collected at 5, 10, 15, 30,
60, 90 and 120 minutes. Absorption rates of fluoride ( 1 mM) were measured in. the presence of an inward-directed H+ gradient (pH8= 7.5, pHm = 5.5), no H+ gradient (pH8=
pHm = 7.5) and an outward-directed H+ gradient (pH8 = 6.5, pHm = 8.0). The differences of fluoride absorption rates between H+ gradient and no H+ gradient were not statistically significant (table VI). However, there was a consistent time-dependent net flux of fluoride from the mucosa to the serosal surface in all groups which was statistically significant
(p<0.05 comparing 5 minutes to 15 or more minutes) suggesting that the sacs were functioning properly. 65
Table VI. Effect of pH on migration offluoride using everted intestinal sacs. Values are means±SE. The serosal and mucosal solutions consisted of Ringer's solution, 10 mM glucose, 20 mM HEPES (pH 7.5) or MES (pH 5.5)/Tris and gassed with 95% 02 and 5%
CO2. The mucosal solution contained 1 mM fluoride. When there was no pH gradient, both pHs and pHm were 7.5. When there was pH gradient, both pHs was 7.5 and pHm was 5.5. Everted sacs were filled with 2.5 mL of serosal solution. Collection H+ gradient (pH8 =6.5, pHm=8.0) No H+ gradient (pH8=pHm=7.5) H+ gradient (pH8 =7.5, pHm=5.5}
time n Serosal Fnet flux n Serosal Fnetflux n Serosal Fnetflux
(minute)· F(mM) µmol/min F(mM) µmol/min F(mM) µmol/min
5 11 0.06±0.00 · 0.029 8 0.08±0.01 0.040 15 0.06±0.01 0.032
10 11 · 0.09±0.01 0.015 8 0.13±0.01 0.022 16 0.10±0.01 0.018
15 11 0.11±0.01 0.011 8 0.16±0.02 0.007 15 0.12±0.01 0.008
30 9 0.20±0.01 0.011 7 0.32±0.05 0.021 14 0.21±0.02 0.013
60. 8 0.32±0.03 0.008 6 0.52±0.07 0.011 14 0.42±0.03 0.013
90 9 0.40±0.03 0.005 7 0.60±0.07 0.002 12 0.56±0.02 0.008·
120 9 0.43±0.03 0.001 6 0.76±0.05 0.006 12 0.67±0.03 0.006 DISCUSSION
These experiments were performed to characterize the fluoride transport properties of isolated BBMV from rabbit small intestine and kidney and everted intestinal sacs from the rat. The results showed that the uptake of fluoride by the rabbit small intestinal and re~al BBMV is strongly dependent on a pH gra~ient and is partially inhibited by DIDS and
DEP. PCMBS inhibited flu?ride uptake in· a dose-response manner and reached a maximum inhibition of 83% when· its concentration was 5.nmM or 10.0 mM. DiBAC had no effect on fluoride uptake. Fluoride _uptake; by the rabbit smail intestinal and renal BBMV· may involve one or more transport proteins. Fluoride transported into rabbit small intestinal and renal BBMV may be divided into that portion due to HF diffusion anc;t that portion due to F--H+ cotransporter (or F--OH~ exchanger). Chloride and nitrate appeared to share the H+-cotransport system (or OH--exchanger) 'with fluoride. We did not find pH-dependence on migration of fluoride using everted intestinal sacs.
The gastric absorption, renal clearance, and transm~mbrane distribution of fluoride
(F) in variety of cell types all show a marked dependence on the magnitude of the pH gradient (Whitford, 1996). The permeation of the weak acid HF (pKa = 3.4), rather than ionic F, appears to largely account for these findings. Messer and Ophaug (1993) and
Nopakun et al.(1989, 1990) reported that, in intact rats without any measures taken to influence gastric acidity, the rate of increase in plasma fluoride concentrations was dependent on the rate of gastric emptying into the small intestine. This suggested that the rate of absorption from the upper intestine was greater than from the stomach. They further investigated gastric fluoride absorption in relation to gastric acidity and, in agreement with earlier findings by Whitford and :co-workers, conclude~ that the rate and extent of absorption from the stomach was inversely related to pH. Absorption from the small
66 67
intestine, however, compensated for the low gastric absorption at high pH so that the
extent of absorption after an hour or more was independent of the pH of. the gastric
contents. More recently they studied fluoride transport in vitro by intestinal and renal
epithelial monolayers (Messer and Birner, 1995). They reported that renal fluoride
transport occurred largely as undissociated HF while intestinal fluo~de transport occurred
as ionic fluoride.
Our data show that a proton gradient, as a driving force, strongly enhances fluoride
transport into the rabbit small intestinarand. renal BBMV. In the presence of a proton
gradient (outsid~>inside), fluoride rapidly crossed the apical membranes and accumulated
inside the vesicles. _A proton gradient was shown to stimulate fluori~e uptake with a 2-fold
. . transient accumulation (overshoot) of fluoride above the equilibrium uptake value. The
reduced uptake after 15 seconds appeared to be due to the subsequent fluoride efflux from
the vesicles dueto'tlle:progressive collapse of the pH gradient. On the other hand, fluoride
uptake in the absence of a H+ gradient and in the presence of an outward-directed H+
gradient was very slow and did not demonstrate an overshoot. Instead, fluoride uptake
. increased slowly with time and reached what appeared to be an equilibrium value sometime
between 5 arid 60 minutes. There was no statistically significant difference in the vesicular
equilibt1um concentration of fluoride (f~gures 4, 5), indicating that the observed changes in
the initial uptake rates were not due to pH-induced alterations in the ve~icle volume and /or
integrity. We evaluated the role of H+ concentration (table II and III) on BBMV fluoride
uptake. In the absence of a proton gradient (PHi=pH0 ), the initial uptake rates of fluoride
in three different proton concentration groups (pH 5.5, 7.5 and 9.0) were not significantly
different. This indicated that it was the transmembrane proton gradient, not internal or
external H+ concenttationper se, that influenced fluoride transport.
DIDS has been found to covalently link to the band 3 protein, an anion transport
system (Cabantchik and Rothstein, 1974). Since DIDS was found to specifically bind and
inhibit the function of band 3 protein, DIDS became a standard reagent for studying anion 68
transport (Rothstein, 1989). DIDS can also inhibit Cl--Hco3- exchange (Knickelbein et
J al., 1985b; Lamprecht et al., 1993), so4--OH- (Schron et al., 1985), and Cl-oxalate exchange (Knickelbein etal., 1985a; McConnell and Aronson, 1994). Some studies have shown that DIDS has no effect on fluoride uptake/efflux by red blood cells (Whitford et al,,,
1985) nor on the reabsorption from the cortical collecting duct of the rabbit (Rouch et al.,
1992). Uptake by rabbit submandibular cells, however, was partially inhibited by DIDS
(Melvin et al., 1990). In the present study, DIDS inhibited fluoride uptake by intestinal and renal BBMV by 40-56% (table IV and V). This suggested that fluoride transport in these systems involves a transport protein which is sensitive to DIDS. DEP, a histidine-specific reactive reagent, is usually used to characterize H+ coupled transport systems. Histidyl residues have been shown to be essential for the function of many transport systems that are driven by H+ gradients. This includes the transport systems responsible for the uptake of lactose, lactate and proline into E. coli membrane vesicles (Padan etal., 1979); the uptake of Na+ into renal BBMV (Grillo and
Aronson, 1986) and E. coli membrane vesicles (Damiano et al., 1985); the uptake of peptide into renal BBMV (Miyamoto et al., 1986) and the uptake of cephradine via dipeptide carriers into small intestinal BBMV (Kato et al., 1989). In the case of the transport of lactose, lactate and proline into E. coli membrane vesicles, essential histidyl residues are located at the H+-binding site rather than the bin9ing site for_the organic solutes .
(Padan etal., 1979). Studies with renal BBMV have demonstrated.that histidyl residues ' . are located at the amiloride-binding site of the ~a+-H+ exchanger (Grillo and Aronson,
1986) which also accepts H+ as a substrate (Aronson et al., 1983). Sokol et al.(1988) reported that the-uptake of p"."aminohippurate.was inhibited by DEP in the presence of H+ gradient. Takano etal. (1994) found that DEP selectively inhibited membrane potential sensitive p-aminohippurate uptake but not the uptake by the anion exchanger. Modification of the Na+-K+-ATPase with DEP altered the properties of Na+-K+-ATPase (Skou, 1985).
The uptake of fluoride was inhibited by DEP in the presence of a proton gradient (table IV, 69
V). These results, like those from the studies with DIDS, may provide evidence that proton-gradient dependent fluoride transport needs a transport protein in which histidyl residues play an important role in the process of fluoride crossing the apical membrane.
Membrane water channels were first shown by Macey and Farmer ( 1970) to be inhibited by mercurial sulfbydryl reagents including HgC1 2 and PCMBS. Solomon et al. (1983) found that the sulfbydryl reagent, PCMBS, also modified the tryptophan residues and postulated that the band 3 anion exchanger of RBC functions as an aqueous pore permeated by water and µrea. More recent studies using the Xenopus oocyte expression system failed to demonstrate concordance of water transport and anion transport (Zhang et al., 1991). It is now clear that the plasma membranes of some epithelia exhibit an extremely high water permeability that is mediated by specialized protein pores called water channels, now referred as to "aquaporins" (Sabolic and Brown, 1995). Preston et al.
(1993) demonstrated that cysteine residue 189 is critical to the structure, function, and processing of the channel-forming integral membrane protein (CHIP28) water channel.
Permeability of the water pore is blocked by reaction of Hg2+ with cysteine residue 189 in the native molecule. .Moreover, cysteine residue 189 is apparently critical to the native folding of the CHIP28 protein. In addition to that , PCMBS also inhibits the uptake of p aminohippurate (Sokol etal., 1986), tetraethylammonium in renal brush border membrane vesicles (Hori etal., 1987) and the uptake of tetraethylammonium in rat renal basolateral membrane (Katsura ~tal., 1992). In the present studies; fluoride uptake was inhibited by
PCMBS and the inhibition was dose-dependent (figure 8). This indicated that PCMBS is a very important fluoride transport inhibitor and sulfbydryl groups of transport system.
Grillo and Aronson (1986), Sokol et al. (1988) and Kato et al.(1989) indirectly demonstrated thatDEP did not facilitate dissipation of the H+ gradient. Hori etal. (1987) studied the effect of PCMB·s using Acridine Orange on the dissipation rate of the H+ gradient in BBMV and concluded that PCMBS had no effect on the dissipation rate of the
H+ gradient. In their cephradine transport study, Kato etal.(1989) found that the uptake of. 70 cephradine was inhlbited by DEP, b~t was ~ot inhibited by PCMBS in the presence of a proton gradient. T~e.se results indicated that histidy~· groups, not sulfhydryl groups, are essential for the transport of cephradine and indirectly demonstrated that PCMBS did not facilitate dissipation of a H+ gradient
The results shown in figure 11 appeared to show that the inhibitors (DIDS, DEP,
PCMBS and diBAC) facilitated the dissipation of imposed pH gradients. However, the results shown· in figure 12 obtained-in ·the absence of vesicles indicated that the inhibitors may have reacted with BCECF during the ~reeze-thawing loading process. The inhibitors may have quenched the BCECF fluorescence signal when they reacted with BCECF. This may have been the reason that the fluorescent signals of BBMV pretreated with the inhibitors were lower than those of controls~ This hypothesis was partially confirmed by another experiment (figure 13). These results indicated that diBAC slightly reduced intravesicular pH prior to adding pH 5.0 uptake buffer and that DIDS, DEP and PCMBS had little or no effect on intravesicular pH prior to adding pH 5.0 uptake buffer and did not
I facilitate dissipation of the imposed pH gradient, consistent with previous reports. Thus, it is likely that the inhibitions of proton-gradient dependent fluoride transport by DIDS, DEP and PCMBS are due to specific interactions with one or several transporters, rather than due to the decrease of the intravesicular space or to the collapse of the of proton gradient.
Fluoride can be transported into membrane vesicles by HF diffusion. In the absence of transmembrane proton gradient but in the presence of a fluoride gradient, fluoride was not transported into the intravesicular space. In the presence of transmembrane proton gradient, however, fluoride transport was very rapid and it transiently accumulated in membrane vesicles. This suggested that F- itself could not enter membrane vesicles using aquaporins (or any other transport pathway), although fluoride readily forms hydrogen bonds with water.
Raha et al. ( 1993) and Knauf et al. ( 1995) investigated the effects of fluorescent oxonol dyes on chloride uptake by red blood cell. The diBAC .concentrations used in their 71 studies ranged from 50 to 400 nM. They found that if diBAC was allowed to equilibrate with band 3, the concentration that gave the half-maximal inhibition of chloride exchange
(ICso) was only 1.05 nM. They co~cluded that diBAC is the most potent known inhibitor of band 3-mediated ~nion exchange in the red blood cell and_ that its site of action is different from that of DIDS. Under our experimental conditions, the results.showed that fluoride transport was not inhibited by diBAC (figure 9, 10), although ~i might slightly decrease intravesicular pH (figure 13). We found that diBAC did not decrease fluoride l transport even at 50 µM. This concentration was 125 times greater than that of the previous report (Knauf ei'al., ·1995). These result~ suggest that DIDS and diBAC may modify different parts of the band 3 protein or modify two different anion exchanger proteins. The site of action of diBAC may not involve fluoride transport. Another interpretation is that the differences may be attributed to species differences (rabbit versus human) or to differences in the properties of erythrocytic anion exchanger versus the one in the brush border membranes used in the present studies. Further studies are needed to clarify this discrepancy.
Liedtke and Hopfer (1982) and· Knickelbein et a/,. (1985b) demonstrated the existence of a chloride conductance pathway by concentrative c1- uptake driven by a K+ diffusion potential in rabbit small intestinal BBMV. The stilbenes, 4-acetamido-4' isothiocyanostilbene-2, 2'-disulfonate (SITS) and DIDS strongly inhibited Cl--OH exchange but did not affect the chloride conductance pathway. Under our experimental conditions using small intestinal BBMV (figure 16), we found that the fluoride uptake rate in K+ diffusion potential groups (A', B') were slightly higher than occurred in the absence of K+ diffusion potential (groups A, B, C, C'). The increased uptake was not inhibited by
DIDS. An inside negative potential, due to outwardly-directed K+ diffusion potential, reduced fluoride uptake although it did not reach statistical significance (figure 18). This may indicate that an inwardly-directed K + diffusion potential in small intestinal BBMV provides the energy for fluoride uptake by establishing an inside-positive potential, which 72 depends ori the action of valinomycin. Valinomycin specifically increases the K+ conductance of many biological membranes, thus providing the pathway for K + diffusion.
The p- movement in response to an intravesicular positive potential is consistent with an p conductance pathway. This pathway appears to exist separately from the F--H+ cotransporter (or F--oH- exchangers), since this transporting system was proton-gradient dependent and was inactivated by the inhibitors of DIDS, DEP and PCMBS. We did not find evidence for a fluoride conductive pathway in the renal BBMV (figure 17). Taken together, all experimental evidence from our studies suggest that there is a pH gradient dependent, carrier-mediated mechanism for fluoride transport in the rabbit small intestinal and renal BBMV. Fluoride transport into the membrane vesicles may involve three mechanis'ms since DIDS, DEP and PCMBS only partially inactivated fluoride uptake. The first mechanism seems to be uninhibitable fluoride transport which is not inhibited by DIDS~ DEP and PCMBS 3:nd may enter the vesicles in the form of HF by diffusion. ' The second m_echanism appears to be an inhibitable_ fluoride transport which is
. ' ' sensiti~e to DIDS, DEP and PCMBS ·inhibitors, in which fluoride may cross the brush border membrane by. F--H+ cotransporter or F--QH- exchangers. Fluoride ions are translocated into membrane vesicles b~ this transport system only in the prese~ce of an· inward-directed proton gradient. It ca~ mediate fluoride permeation of apical membranes by cotransport with H+ or exchanging with OH-. Sulfhydryl and histidyl groups appear to play a very important role in this transport system or carrier. This transport protein is different from band 3 anion exchanger protein since DIDS, a specific band 3 anion exchanger inhibitor, only partially inhibited fluoride transport and diBAC had no effect on fluoride transport. Although PCMBS was found to be the most potent inhibitor of the fluoride transporter, it cannot be used as a selective inhibitor of this transport system, because it also inhibited the transport of p-aminohippurate (Sokol et al., 1986) and tetraethylammonium ion (Hori et al., 1987; Katsura et al., 1992). Whether the 73
pharmacological differences between the fluoride transporter and band 3 protein reflect
structural differences remains to be determined.
The third mechanism of fluoride transport is the m_ovement of fluoride ion
transported into membrane vesicles through a membrane potential-sensitive conductive
pathway in small intestinal BBMV but not renal BBMV.
Knickelbein et al. (1985b) demonstrated an c1--Hc0:3- exchanger in rabbit ileal.
BBMV which was inhibited by DIDS. In our study, an outward-directed c1- and HCO3- gradients (50 mM inside, 0 mM outside) were set up to exchange extravesicular fluoride
ion in the absence of a pH gradient. Neither c1- nor· HC03- preloaded into the
intravesicular fluid space accelerated p- influx (figures 14, 15). This indicates that there is
no F--c1- and/or F--Hc03- exchangers in the rabbit small intestinal and renal BBMV.
In the present study, we showed that fluoride transport into membrane vesicles was
dependent on an inward-directed proton gradient. In intestine, the pH in the close vicinity
of the brush border membrane is acidic compared to the pH of the bulk solution in the
lumen (Lucas et al., 1975, 1976, 1978, 1983; Hogerle and Winne, 1983). The pH in this
region was reported to be 5.5-6.0. The presence of an acidic microclimate on the surface of
intestinal ·brush· border membrane results in a H+ gradient across this membrane. This
· inward-directed H+ grad_ient plays an important role in the absorption of vital -nutritients
such as small peptides and folate (Ganapathy and Leibach, 1985; Said etal., 1987). There
exists a Na+-H+ antiport system in both intestinal and renal brush border membrane
\ vesicles (Murer et al., 1976) which mediates uphilLeffluxof protons coupled to the
downhill influx of sodium. This Na+-H+ antiport system is primarily responsible for
maintaining the proton gradient across the brush border membrane of these epithelial cells
(Tiruppathi et al., 1988; Murer et al,., 1976; Knickelbein et al., 1983; Miyamoto et al., 1988; Ramaswamy etal., 1989). Therefore, it is possible that the inward-directed H+
gradient in vivo and in vitro is the driving force for fluoride transport in small intestine as
well as kidney. This may be why the intestinal fluoride absorption studies using everted- 74 sacs did not show a pH-dependence (table VI). That is, the process is, in fact, pH dependent but becaus~ the pH at the absorptive surface 'r~mains constant within a narrow range (pH 5.5-6.0), regardless of changes in the pH of the bulk solution, the pH qependence has not been detected. This hypothesis could be tested by using agents
(amiloride and harmaline) that inhibit the Na+-H+ antiport system.
The study in which the effect of medium osmolarity on fluoride uptake was tested showed that the uptake was completely due to transport into intravesicular fluid space rather than binding to the membrane in small intestinal BBMV (Figure 6). However, in the case of renal BBMV (Figure 7), there appeared to be some residual uptake of fluoride even at infinite medium osmolarity. This suggested that a portion of the uptake may have been due to binding to the membrane surface and not transported into intravesicular fluid space.
This residual uptake (binding) was about 10% of the total uptake under isosmolar conditions (300 mosm). It was also noticed with intestinal as well as renal BBMV that uptake was inversely related to medium osmolarity only if osmolarity was above 300 ~osm due to membrane vesicle shrinking. On the other hand, with medium osmolarities below
300 mosm (hypotonic conditions) the uptake declined with decreasing osmolarity. This was probably due to osmotic swelling of the vesicles under these conditions, leading to increased efflux of fluoride from leaking or partially tom vesicles.
The transport systems for many amino acids and sugars are sodium dependent
(Ganapathy et al., 1994). The study of the effects of a sodium or potassium gradient on uptake of fluoride in the small intestinal and renal BBMV showed that there was no Na+ - or K+-dependent fluoride uptake (figure 19, 20). This suggested that fluoride transport is Na+- or K+-independent
Knickelbein et al. (1985b) studied t~e effect of various anions on pH and HC~ gradient-stimulated chloride, uptake using rabbit ileal BBMV. In that study, the pH
stimulated (pHi=7.7, pH0 =5.5) chloride (5 mM) uptake at 6 seconds was determined in presence of 50 mM potassium fluoride. They found that fluoride inhibit~d the uptake of 75 chloride . Karniski ( 1989) found similar results. Our results demonstrated that chloride and nitrate can compete with fluoride for the same transport system in both-small intestinal and renal BBMV (figure 21, 22). These data suggested that fluoride , chloride and nitrate can share the same transport system. SUMMARY
1. The uptake of fluoride by the rabbit small intestinal and renal BBMV was strongly dependent on an inward-directed pH gradient.
2. The uptake of fluoride by the rabbit small intestinal and renal BBMV was partially inhibited by DIDS and DEP. PCMBS inhibited fluoride uptake in a dose-response manner. The maximum inhibition was 83% when the PCMBS concentration was 5 mM or
10 mM. Fluoride transported into rabbit small intestinal and renal BBMV may be divided into HF diffusion, F--H+ cotransport or F--oH- exchange.
3. There may be a fluoride conductive pathway in small intestinal BBMV but it appears to play a minor role. There was no evidence for this pathway in renal BBMV.
4. There was no evidence for Cl-/F- and/or HCO3-/F- exchange in the small intestinal or renal brush-border membrane vesicles.
5. DiBAC might slightly reduce intra.vesicular pH prior to adding pH 5.0 uptake buffer. DIDS, DEP and PCMBS did not change intravesicular pH prior to adding pH 5.0 uptake buffer and did not facilitate the dissipation rate of imposed pH gradient.
6. Fluoride transport was independent of Na+ and K + gradients.
7. Chloride and nitrate can compete.with fluoride for the same transport system.
8. DiBAC had no effect on fluoride uptake by rabbit small intestinal and renal
BBMV.
9. We did not find pH-dependent migration of fluoride using rat everted intestinal sacs.
76 REFERENCES
Aroson, P. S., M.A. Suhm and J. Nee. Interaction of external H+with the Na+-H+
exchanger in renal microvillus memprane vesicles. J. Biol. Chem. 258: 6767-
6771 (19831·
Bhussry, B.R.. Chronic toxic effects on enamel organ.· In: Fluoride and human health.
WHO. Geneva. pp 230-238 (1970).
Cabantchik, Z. I. and A. Rothstein. Membrane proteins related to anion.permeability of
human red blood cells. I. Localization of disulfonic stilbene binding sites in
proteins involved in permeation. J. Membr .. Biol. 15: 207-226 (1974)
Crane, R. K., and T. H. Wilson. In vitro method for the study of the rate of intestinal
absorption of sugars. J. Appl. Phy~iol. 12: 145 (1958).
Cremer, H. D., and W. Buttner. Absorption of fluorides. In: Fluoride and human health.
WHO. Geneva. pp 75-91(1970).
Damiano, E., M. Bassilana, and G. Leblanc. Chemical modifications of the Na+-H+
antiport in Escherichia coli membrane vesicles. Eur. J. Bioc~em. 148: 183-188
(1985).
Gabler, W. L. Absorption of fluoride through the oral mucosa of rats. Archs. Oral.
Biol. 13: 619-623 (1968).
Ganapathy. V, M. Brandsch, and P.H. Leibach. Intestinal transport of amino acids and
peptides. In: Johnson LD ed. Physiology ofthe Gastrointestinal Tract. New York:
Raven Press: 1773-1794. 1994.
Ganapathy, V., and F. H. Leibach. Is intestinal peptide transport energized by a proton
gradient. Am. J. Physiol. 0153-0160. (1985).
77 78
Gasalla-Herraiz, J., S. Rhee, and C. M. Isales. Calcium-sensitive probes for the
measurement of intracellular calcium: Effects of buffer system and magnesium
concentration. Biochem. Biophys. Res. Commun. 214: 373-388 (1995).
Grillo, F. G., and P. S. Aronson. Inactivation of the renal microvillus membrane
Na+-H+ exchanger by histidine-specific reagents. J. Biol. Chem. 261: 1120-1125
(1986).
Gutknecht, J., and A. Walter. Hydrofluoric and nitric acid transport through lipid bilayer
membranes. Biochim. Biophys._ Acta. 644: 153-156 (1981).
Harrison, J. E., K. G. McNeill, W. C. Sturtridge, T. M. Bagley, M. C. Muray, C.
Tam, and V. Fomasier. Three-year changes in bone mineral mass of
postmenopausal patients based on neutron activation analysis of the central third of
the skeleton. J. Clin. Endocrinol. Metab. 52: 751-758 (1981).
Hodge, H. C., and F. A. Smith. Biologic properties of inorganic fluorides. In:
Fluorine Chemistry. Simons, JF ( ed) Vol. IV, Academic Press, New York.
London, pp 1-37 (1965).
Hogerle, M. L., and D. Winne. Drug absorption by the rat jejunum perfused in situ.
Dissociation from the pH-partition theory and role of microclimate-pH and unstirred
layer. Naunyn-Schmiedeberg's Arch. Pharmacol. 322: 249-255. 1983.
Hori, R., H. Maegawa, T. Okano, M. Takano, and K. Inui. Effect of sulfhydryl
reagents on tetraethylammonium transport in rat renal brush border membranes.
J. Pharmacol. Exp. Ther. 241: 1010-1016 (1987).
Kainiski,L. P._ Activation of q-/oH- exchange by parachloromercuribenzoic acid in
- rabbit renal brush border membranes. J. Membr. Biol.· 112: 59:66 (1989).
Kato, M., H. Maegawa, T. Okano, K.Jnui, and R. Hori. Effect of various chemical
modifiers on H+ coupled transport of cephradine via dipeptide carriers in rabbit
intestinal brush-border membrane: Role of histidine residues. J. Pharmacol. Exp.
Ther. 251: 745-749 (1989). 79
Katsura, T., M. Takano, Y. Tomita, M. Yasuhara, K. Inui and R. Hori. Characteristics of
organic cation transporter in rat renal basolateral membrane. Biochim. Biophys.
Acta. 1146: 197-202 (1993).
Knauf, P.A., F. Y. Law, and K. Hahn. An oxonol dye is the most potent known
inhibitor of band 3-mediated anion exchange. Am. J. Physiol. 269: C1073-C1077
(1995).
Kntckelbein, R., P. S. Aronson, W. Atherton, and J. W. Dobbins. Sodium and chloride
transport across rabbit ileal brush-border. I. Evidence for Na-H exchange. Am.
J. Physiol. 245: G504-G510 (1983).
Knickelbein, R., P. S. Aronson, and J. W: Dobbins. Substrate and inhibitor specificity of
anion exchanger on the brush border membrane of rabbit ileum. J. Membrane.
Biol. 88: 199-204 (1985a).
Knickelbein, R., P. S. Aronson, C. M. Schron, J. Seifter, and J. W. Dobbins.
Sodium and chloride transport across rabbit ileal brush-border. II. Evidence for
Cl-HC03 exchange and mechanism of coupling. Am. J. Physiol. 249: G236-
G245 (1985b).
Lamprecht, G., U. Seidler, and M. Classen. Intracellular pH-regulating ion transport
mechanis1J1S in parietal cell basolateral membrane vesicles. Am. J. Physiol. 265:
G903-G910 (1993).
Liedtke, C. M., and U. Hopfer. Mechanism of Cl- translocation across small intestinal
brush- border membrane. II. Demonstration of c1--OH- exchange and c1-
conductance. Am. J. Physiol. 242: 0272~0280. 1982..
Lowry, 0. H., N. J. Rosebrough, ·A. L Farr, and R. J. Randall.' Protein measurement
with the Folin'phenol reagent. J. Biol. Chem. 193: '265-275 (1951)~
Lucas, M. L. Determination of acid surface pH in vivo in rat proximal jejunum. Gut 24: . . 734-739. 1983.
Lucas, M. L., J. A. Blair, B. T. Cooper, and W. T. Cooke. Relationship of acid 80
microclimate in rat and human intestine to malabsorption. Biochem. Soc. Tr0:ns·. 4:
154-156. 1976.
Lucas, M. L., B. T. Cooper, F. H. Lei, I. T. Johnson, G. K. T. Holmes,). A. Blair, and
W. T. Cooke. Acid microclimate· in coeliac and Crohn's disease: A model for
folare ~alabsorpti~n. Gut 19: 735-742. 1978. Lucas, ¥--1::, W. Schneider, F. J. Haberich, ~nd J. A. Blair. Direct mec!Surement by pH ·microelectrode of the pHmicroclimate in rat proximal jejunum. Proc. R. Soc.
Lond. B. Biol. Sci. 192: 39-48. 1975.
Macey, R. I. and R. E. L. Farmer. Inhibition of water and solute permeability in human
red cells. Bioc~iin. Biop_hys. Aq~a. ~il:)0~-106-(1970).
Mazze, R. I., J. R. Trudell, and M. J. Cousins. Methyoxyflurane metabolism
and renal dysfunction: Clinical correlation in man. Anesthesiology 35: 247-252
(1971).
McConnell, K. R., and P. S. Aronson. Effects of inhibitors on anion exchangers in rabbit
renal brush border membrane vesicles. J. Biol. Chem. 269: 21489-21494 (1994).
Melvin, J. E., W; H. Bowen, and R. J. Turner. Fluoride movement occurs via chloride
pathways in rat submandibular cells. J. Dent. Res. 69 (special issue): Abst. 1. 824
(1990).
Messer, H. H., and R. Birner. Fluoride transport in vitro. by intestinal and renal
epithelial monolayers. J. Dent. Res. 74: 749 (1995).
Messer, H. H., and R.H. Ophaug. Influence of gastric acidity on fluoride absorption in
rats. J. Dent. Res. 72: 619-622 (1993) ..
Miyamoto, Y., D. F. Balkovetz, V. Ganapathy, T. lwatsubo, M. Hanano, and F. H.
Leibach. Effects of phenothiazines on the Na+-H+ exchanger of the brush border
membrane from the proximal small intestine of the rabbit. J. Pharmacol. Exp.
Ther: 245: 823-828. 1988.
Miyamoto,. Y., V. Ganapathy, and F. H. Leibach. Proton gradient-coupled uphill 81
transport of glycylsarcosine in rabbit renal brush-border membrane vesicles.
Biochem. Biophys. Res. Commun. 132: 946-953 (1985).
Miyamoto, Y.,_V. Ganapathy, and F. H. Leibach. Identification of histidyl and thiol
groups at the active site of rabbit renal dipeptide transporter. J. Biol. Chem. 261:
16133-16140 (1986).
Murer, H., U. Hopper, and R. Kinne. Sodium/proton antiport in brush border
membrane vesicles isolated from rat small intestine and kidney. Biochem. J. 154:
579-604' (1976).
Nopakun, J., H. H. Messer, and V. Voller. Fluoride absorption from the gastrointestinal
tract of rats. J. Nutr. 119: 1411-1417 (1989).
Nopakun, J., and H. _H. Messer. Mechanism of fluoride absorption from the rat small
intestine. Nutr~ Res. 10: 771-780 (1990)-
Padan, E., L. Patel, and H. R. Kaback. Effect of diethylpyrocarbonate on lactose/proton
symport in Escherichia coli membrane vesicles. Proc. Natl. Acad. Sci. 76: 6221-
6225 ( 1979).
Parkins, F. M. Active transport: species and age effects with rodent intestine in vitro.
Biochim. Biophys. Acta. 241: 507-512 (1971).
Parkins, F. M., J. W. Hollyfield, A. J. Mccaslin, S. M. L. Wu, and R,, G. Faust.
Active transport of fluoride by the rat intestine in vitro. Biochim. Biophys. Acta.
126: 513-518 (1966).
Patten, J. R., G. M. Whitford, G. I. Stringer, and D. H. Pashley. Oral absorption of \ radioactive fluoride and iodide in rats. Archs. Oral. Biol. 23: 215-217 (1978).
Preston, G. M., J. S. Jung, W. B. Guggino, and P. Agre. The mercury-sensitive
residue at cysteine-189 in the CHIP28 water channel. J. Biol. Chem. 268: 17-20
(1993).
Raha, N. M., L. J. Spinelli, and P.A. Knauf. WW-781, a potent reversible inhibitor of
red cell Cl- flux, binds to band 3 by a two-step mechanism. Am. J. Physiol. 82
265: C521-C532 (1993).
Ramaswamy K, J. M. Harig, J. G. Kleiman, M. S. Harris and J.A. Barry: Sodium~protoh
exchange in human ileal brush-border membrane vesicles. Biochim. Biophys. Acta
I 981: 193-199. 1989.
Ramamoorthy, S., F. H.Leibach, V. B. Mahesh, and V.Ganapathy. Active transport of,
dopamine in human placental brush border.membrane vesicles. Am. J. Physiol.
262: C1189-Cl196 (1992).
Ramamoorthy, S., P. D. Prasad, P. Kulanthaivel, F.H. Leibach, R. D.Blakely, and
V.Ganapathy. Expression of a cocaine-sensitive norepinephrine transporter in
the human placental syncytiotrophoblast. Biochemistry. 32: 1346-1353 (1993). ·
Riggs, B. L., E. Seeman, S.F. Hodgson, D.R. Taves, and W.M.O'Fallon. Effect of the
fluoride /cal~ium regimen on vertebral fracture occurrence in postmenopausal
osteoporosis. N. EngL J; Med. ·306:-446-450.(1982).
Rothstein, A. Anion exchanges and barid 3 protein. In Tosteson D~ (ed): Membrane ' . . -··.transport: People and/de~. American Physiological Society, Bethesda, Maryland.
pp203-235( 1989).
Rouch, A. J., G. M. Whitford, and H. T. Campbell. Fluoride flux.in the rabbit CCD:
A pH- ·dependent event Kidney International. 41: 342-349 (1992).
Sabolic, I. and D. Brown. Water channels in renal and normal tissues. NIPS. 10: 12-17
(1995).
Said, H. M., F. K. Ghishan, and R. Redha. Folate transport by human intestinal brush
border membrane vesicles. Am. J. Physiol. 252: 0229-0236. 1987.
. Schron, C. M., R. G. Knickelbein and P. S. Aronson. pH gradient-stimulated sulfate ·
transport by rabbit ileal brush-border membrane vesicles: evidence for S04--0H
exchange. Am. J. Physiol. 249: 0607-0613 (1985).
Singh, A., and S. S Jolly. Chronic toxic effects on the skeletal system. In: Fluoride
and human health. WHO.· Geneva. pp 238-249 (1970). 83
Skou, J.C .. Modification of the (Na+-K+)-ATPase with 1- ethyl-3- (3-
dimethylaminopropyl) carbodiimide. and with diethylpyrocarbonate. : Effect on
the conformational transition. Biochim. Biophys. Acta 819: 119-130 (1985).
Smith, F. A. Metabolism of· inorganic fluoride: In Smith (ed), Handbook of
Experimental Pharmacology. Vol:- XX, part.I. Spri1:1ger, New York. pp 53-140
(1966).
Sokol, P. P., P. D. Holohan, and C. R. Ross. Sulfhydryl groups are essential for
organic anion exchange in canine renal brush border membranes. Biochim.
Biophys. Acta. 862: 335-342. (1986).
Sokol, P. P., P. D. Holohan, and C. R. Ross. Arginyl · and histidyl groups are
essential for organic anion exchange in renal brush:-border membrane
vesicles. J. Biol. Chem. 263: 7118-7123 ( 1988).
Solomon, A. K., B. Chasan, J. A. Dix, M. F. Lukacovic, M. R. Toon, and A. S.
Verkman. The aqueous pore in red cell membrane: Band 3 as a channel
for• anions , cations, nonelectrolytes, and water. Ann. NY Acad. Sci. 414: 97-
124 (1983).
Stookey, G. K., D. B. Crane, and J.C. Muhler. Role of skeleton. and kidney in
fluoride absorption in the rats. Proc. Soc. Exp. Biol. Med. 113: 366-370 (1963).
Stookey, G. K., D. B. Crane, and J. C. Muhler. Further studies on fluoride
absorption. Proc. Soc. Exp. Biol. Med. 115: 295-298 (1964a).
Stookey, G. K., D. B. Crane, and J. C. Muhler. In vitro studies concerning fluoride
absorption. Proc. Soc. Exp. Biol. Med. 115: 298-301 (1964b).
Takano~ M, J. Nagai, T. Okano, K. Hirozane, M. Yasuhara and R. Hori. Differential
sensitivity of organic anion transporters in rat renal brush-border membrane to
diethylpyrocarbonate. Biochim. Biophys. Acta 1192: 190-196 (1994).
Takuwa, N., T. Shimada, H. Matsumoto, and T. Hoshi. Proton gradient-coupled
transport of glycylglycine in rabbit renal brush-border membnµ1e vesicles. 84
Biochim. Biophys. Acta. 814: 186-190 ( 1985).
Taves, D. R. . Detennination of submicromolar co.ncentrations of fluoride in biological
samples.Talanta 15: 1015-1023 (1968).
Tiruppathi, C., Y. Miyamoto, V. Ganapathy, and.F. H. Leibach. Fatty acid-induced
alterations in transport systems of the small intestinal brush-border
membrane. Biochem.Pharmacol. 37: 1399-1405 (1988).
USPHS Ad Hoc Subcommittee on Fluoride: Review of fluoride: benefits and risk
(U.S. Publ. Hlth Serv., Dept. of Hlth. Human Serv., Bethesda, Md., 1991)
Whitford, G. M. The metabolism and toxicity offluoride. In: Myers HM (ed) Monographs
in Oral Science Vol. 16, 2nd edition. S Karger AG, Basel, Switzerland. pp 1-21
(1996).
Whitford, G. M., N. B. Allison, and D. E. Pearson. RBC fluorid~ conductance in the
presence of anion transport inhibitor. J. Dent. Res. 64 (special issue): Abst. 400,
219 (1985).
Whitford, G. M., R. S. Callan, and H. S_. Wang. Fluoride absorption through the
hamster cheek pouch: A pH-dependent event. J. Appl. Toxicol. 2: 303-306
(1982).
Whitford, G. M., and· D. H. Pashley. Fluoride absorption: the influence of gastric acidity.
Calcif. Tissue Int. 36: 032-307 (1984).
Whitford, G. M., a~d K. E. Reynolds. Plasma and developing enamel fluoride
concentrations during chronic acid-base disturbances. J. Dent. Res. 58: 2058-2065
(1979).
Whitford, G. M., and D. R. Taves. Fluoride-induced diuresis: renal tissue solute
concentration, functional, hemodynamic and histologic correlates in the rats.
Anesthesiol9gy. 39: 416-427. (1973).
Whitford, G. M., andJ. L. Williams. Fluoride absorption:. Independence from plasma
_fluoride levels. Proc. Soc~ Exp. Biol. Med.· 181 (4) :550-554 (1986). 85
Wiseman, G. Sac of everted intestinal technique for study __ of intestinal absorption in
vitro. Methods Med. Res. 9: '2:i!,7-292 (1961).
Zhang, R., S. A. Alper, B. Thorens, and A. S. Verkman. Evidence from oocyte
expression that the erythrocyte water channel is distinct from band 3 and the
glucose transporter. J. Clin. Invest. 88: 1553-1558 (1991).