High proton flux through membranes antidiuretic hormone water channels
containing
H. WILLIAM HARRIS, JR., DEEPAK KIKERI, AGNES JANOSHAZI, A. K. SOLOMON, AND MARK L. ZEIDEL Renal Divisions, Brocktonl West Roxbury Veterans Administration Medical Center, West Roxbury The Children’s Hospital, Brigham and Women’s Hospitals, and Biophysical Laboratory, Harvard Medical School, Boston, Massachusetts 02115
HARRIS, H. WILLIAM, AZI, A. K. SOLOMON,
JR., DEEPAK KIKERI, AND MARK L. ZEIDEL.
AGNES
JANOSH-
High proton flux through membranes containing antidiuretic hormone water channels. Am. J. Physiol. 259 (Renal Fluid Electrolyte Physiol. 28): F366-F371, 1990.-Antidiuretic hormone (ADH) stimulation of toad urinary bladder granular cells causes simultaneous increases in transepithelial water and H’ permeabilities (& and PH+, respectively), suggesting that ADH-elicited water channels inserted into granular cell apical membranes might be permeable to both water and H+. We have previously used self-quenching fluorophores entrapped within endocytic vesicles selectively retrieved from water-permeable apical membranes to measure vesicle PF. The membranes of these vesicles possess an extremely high PF such that our measurements provide only minimum estimates of vesicle PF and have limited our ability to quantitate the properties of ADH water channels. We therefore quantitated vesicle PH+ using similar rapid mixing techniques. Vesicle P H+ was 5.1 t 0.5 X lo-" cm/s. Activation energy of this process was 3.6 t 0.6 kcal/mol, indicative of H+ flux through an aqueous channel. The mercurial reagent, parachloromercuribenzenesulfonate (PCMBS), which inhibits ADH-stimulated transepithelial PF in intact bladders by 50and 60%, inhibited vesicle PH+ by 55%. N-Ethylmaleimide phloretin, which do not alter ADH-stimulated PF, did not affect vesicle PH+. We conclude that membranes containing ADH water channels possess substantial PH+ that likely reflects proton flux through water channels. The apparent high PH+ of the ADH water channel may have important implications for intracellular trafficking of these water channels in ADH-responsive epithelial cells. water permeability; bules; endosome
proton
permeability;
kidney;
collecting
tu-
HORMONE (ADH) markedly increases apical membrane water permeability (PF) of certain tight epithelial cells by vesicle-mediated insertion of unique, highly selective water channels (21, 23). These water channels allow passage of water, and possibly protons, while excluding larger ions and nonelectrolytes (4, 23). Efforts to isolate and characterize ADH water channels have been hampered by the relatively high PF of synthetic and biological membranes (10s3 cm/s) and by the lack of a specific assay that readily discriminates between ADH water channels and other membrane transporters (4). In toad bladder granular cells, water channels are stored in large tubular cytoplasmic vesicles called agANTIDIURETIC
02132;
grephores (9, 18). ADH stimulation causes aggrephore fusion with the granular cell apical plasma membrane and a dramatic increase in pF and proton permeability (Pn+) (6, 8). Termination of ADH stimulation causes apical membrane endocytosis and a fall in & and Pn+ to their low basal levels (8). We have previously demonstrated that fluorescent markers are selectively entrapped in the lumens of these endocytic vesicles that can be isolated from granular cell homogenates (2, 8-10, 11). The pF of these vesicles is greater than 4.5 x lo-’ cm/s at %“C, one of the highest known & values observed in biological membranes (9,22). Because the time resolution of existing instrumentation currently allows measurement of only minimum pF values in these vesicles (9), it is difficult to characterize the properties of water flow across the ADH water channel and to use water flow as an assay for successful reconstitution of water channel activity. We have therefore developed methods to measure proton flux across the membranes of these vesicles. Using a pH-sensitive fluorophore trapped within water channel-containing vesicles (9, 11)) we have examined the Pn+ properties of these vesicles and provide evidence that the ADH water channel mediates proton flux across the limiting membranes of these vesicles. Measurement of proton flux may be useful both in the characterization of ADH water channels and, ultimately, in their purification. MATERIALS
AND
METHODS
Dominican toads (Bufo marinus) were obtained from National Reagents, Bridgeport, CT. Urinary bladders of Bufo marinus were prepared and mounted on cannulas as small sacs.After a ZO-min interval of ADH stimulation (50 mu/ml), the apical surfaces of bladders were exposed to solutions containing 50 mg/ml of fluorescein dextran (F-dextran; av mol mass of 70,000 Daltons) and O-20 mM NaCl, 30 mM KCl, and l-20 mM N-2-hydroxyethylpiperazine-2\r’-2-ethanesulfonic acid (HEPES), pH 8.0, for an interval of 5 min followed by termination of ADH stimulation and incubation for an additional 10 min. The solution containing the F-dextran was then removed, the bladders were thoroughly rinsed to remove all extracellular F-dextran, and scraped epithelial cells were homogenized in a Polytron (Brinkmann Instruments, Westbury, NY) for 5 s at a setting of 7. The supernatant of a
F366
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
WATER
CHANNEL
PROTON
low-speed spin (700 g for 15 min) was centrifuged at 10,000 g for 10 min, and the pellet (“intermediate pellet”) used for all subsequent studies (9, 10) unless specified otherwise. In the absence of ADH, no detectable fluorescence was present in this vesicle fraction despite apical membrane exposure to F-dextran. In some experiments, bladders were incubated for 60 min after withdrawal of ADH, and the trapped F-dextran-containing vesicles were isolated as described above. Previous studies have demonstrated that this procedure results in entrapment of F-dextran within a morphologically distinct endosoma1 compartment, the multivesicular body (MVB) (2, 11). Fluorescence studies with MVB preparations were performed using the exact same protocols as those with the standard preparation. Standard cuvette fluorescence measurements were performed by use of an SLM-Aminco 500C spectrofluorometer (excitation wavelength 499 nm, emission wavelength 520 nm, slits 2 nm). Unless stated otherwise, all experiments were performed at 25OC. Stopped-flow measurements were performed in a High Tech stop-flow device (20-ms dead time) connected to a Photon Technologies Alphascan fluorometer. One syringe contain’& a concentrated vesicle suspension and anti-fluorescein antibody, whereas the other contained buffer with sufficient HCl or NaOH to change the final extravesicular pH as indicated. Data from 3-5 determinations performed on a single day were averaged. Results were fitted to single exponentials and the resulting time constant, 7, was converted to permeability by use of the following equation JH+
= I’H+( AC)A
where JH+ is the flux rate of protons in mol/s, Pn+ is the permeability coefficient of protons in cm/s, AC is the concentration gradient for protons across the vesicle membrane at the start of the experiment in mol/cm3, and A is the surface area of a single F-dextran-containing vesicle (7.1 X 10-l’ cm2; Refs. 9 and 10). JH+ was calculated by multiplying l/7 by the amount of buffer (in moles) in an individual vesicle (single vesicle volume, 1.77 x lo-l5 cm3). Estimates of vesicle volume and surface area were calculated from dimensions of vesicles loaded with horseradish peroxidase (HRP) instead of Fdextran. Previous studies (9-11) have verified these two markers have identical vesicle-loading characteristics. Electron microscopic examinations of HRP-loaded vesicles show that 89% are spherical in shape with a radius of 7.5 x lo-” cm and a volume-to-surface area ratio of 2.5 X 10m6cm. When vesicles were treated with the sulfhydryl-reactive reagents, para-chloromercuribenzenesulfonate (PCMBS) and N-ethylmaleimide (NEM), they were exposed to the compound (1 mM PCMBS or 2 mM NEM) or vehicle (buffer) for 30 min on ice in HEPES buffer before JH+ measurements. Vesicles were then diluted into buffer in the absence or presence of inhibitors and assayed as described. Phloretin experiments were performed after addition of 0.1 mM phloretin to both the intravesicular contents and extracellular solution of Fdextran-containing vesicles. All chemicals were pur-
PERMEABILITY
F367
chased from Sigma (St. Louis, MO) and were of analytical grade. RESULTS
To assure that our measurements truly reflected changes in intravesicular pH (pHi), all experiments were performed after the addition of anti-fluorescein antibody. As shown in Fig. lA, this antibody bound to fluorescein and quenched virtually all extravesicular fluorescence. Subsequent addition of Triton X-100 lysed the vesicles and quenched F-dextran fluorescence such that, in the presence of antibody, nearly all of the fluorescence signal could be accounted for as intravesicular fluorescein. Thus all subsequent experiments were performed in the presence of the antibody to eliminate extravesicular fluorescence. Intravesicular fluorescence correlated linearly with pH over the pH range of 6.5-8.0, permitting estimation of pHi. As shown in Fig. lB, addition of ATP (tracing A) or Na+ (tracing B) to the extravesicular medium did not alter pHi at 25OC. The failure of ATP or Na+ to change pHi may result from either 1) the absence of vesicle Na+H+ exchange or H+-adenosinetriphosphatase (ATPase) activities or 2) vesicle membrane Pi--i+ that is too high (relative to the rates of H+ transport pathways) to permit generation of measurable pH gradients. To test this latter hypothesis, extravesicular pH was abruptly acidified from pH 8.0 to 6.0 at 25OC. (Fig. lC), resulting in rapid acidification of pHi. As shown in Fig. 1B (tracing C) the inward flux of protons was so rapid that extravesicular addition of NH&l at 25°C produced no change in pHi, presumably because the rate of JH+ balanced the rate of NH3 influx. JH+ was markedly slowed when extravesicular pH was abruptly acidified at 5°C rather than 25°C (Fig. 1C). Temperature reduction slowed JH+ and allowed detection of intravesicular alkalization after addition of NH&l (Fig. lB, tracing D). Vesicle JH+ was symmetrical, regardless of whether the interior or exterior was initially more acidic (see below). Thus any pH gradients generated by vesicle proton transport pathways such as Na+H+ exchange or a H+-ATPase might have been rapidly dissipated. It is also possible that NH: itself entered the vesicles as rapidly as NH,; we have recently demonstrated this phenomenon in mouse thick ascending limb of Henle (15). Figure 2 displays measurements of vesicle JH+ under stopped-flow conditions. As shown in Fig. 2A, the rate of decay of the imposed pH gradient varied with changes in the intravesicular buffer concentration. When the lumens of F-dextran-loaded vesicles were buffered with 20 mM HEPES, the time constant of the decline in pHi was given by a 7 of 7.2 s and yielded a Pn+ of 2.5 X 10m3 cm/s. When vesicles were loaded and maintained in 2 mM HEPES, the 7 was 0.8 s, giving a Pn+ of 2.2 X 10m3 cm/s. The close agreement of &-i+ values under conditions of low and high intravesicular HEPES concentrations indicates that, in this range of added buffer, the added buffer itself (and not other vesicle components) dominates vesicle buffer capacity. The change in 7 and the constancy of PH+ as a function of intravesicular buffer
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
F368
WATER
CHANNEL
PROTON
PERMEABILITY
A
2*4T
-
Low Buffer
- -
High
Buffer
1.8
0.6
.
0
1
.
2
Add NH4+
3
l
25 OC
4
5
6
7
8
Control - -. Gramicidin
C
Add
Na+ or ATP
L
0 c 0
C
tI
II
II
II
I
10
20
30
40
50
I
. ..
..
l
e
lb
C
25'C 5OC
.
l -
.
. l
l
.
. l
l
.
. “.***
.
.
.
.
. .
.
.
.
. .
20
ii
. l
10
4-
. l
0
i
30 Time,
40
50
.
60
3ec
FIG. 1. Measurement of intravesicular fluorescence of apical membrane vesicles loaded with fluorescein dextran (F-dextran). A: effect of addition of anti-fluorescein antibody and Triton X-100 on fluorescence of vesicles loaded with F-dextran. In 5 experiments, antibody addition reduced fluorescence of vesicle suspension by 22 -+ 7% (n = 5; SD); addition of Triton X-100 left 18 t 12% of fluorescence at 1 min and 4 t 0% at 12 min. Fluorescence is displayed in arbitrary units. Antibodymediated quenching was unaffected by changes in pH in the range of 5.5-9.0 or additions of Triton X-100 and NaCl to final concentrations of 4% and 2.0 M, respectively. B: effect of extravesicular addition (downward pointing arrows) of 9 mM ATP (tracing A), 20 mM NaCl (tracing B), 20 mM NH&l at 25°C (tracing C), or at 5°C (tracing D), on fluorescence of intravesicular F-dextran. Curves were displaced along y-axis for visual clarity. C: effect of changing extravesicular pH from 8.0 to 6.0 (at the vertical arrow) on intravesicular fluorescence at 25°C and 5°C. Extravesicular pH was lowered by abrupt addition of a small volume of 2 N HCl.
FIG. 2. Stopped-flow measurement of the rate of change of intravesicular pH in apical membrane vesicles. A: effect of buffer concentration on rate of pH change in F-dextran-loaded vesicles. Vesicles were loaded, prepared, and maintained in 20 mM (“high buffer”) or 2 mM (“low buffer”) HEPES buffer at pH 8.0; at start of experiment, pH outside was abruptly lowered to 6.0. Curves were displaced along yaxis for visual clarity. B: effect of gramicidin on proton flux (&+). Vesicles were preincubated for ~2 min in presence or absence of 5 PM gramicidin before assay of Jn+. These results are representative of identical experiments performed on 2 separate preparations. C: symmetry of inward and outward Jn+s in F-dextran-loaded vesicles. Vesicles were preincubated at pH 8.0 and diluted into pH 6.0 (tracing A), or preincubated at pH 6.0 and diluted to pH 8.0 (tracing B).
concentration indicate that our fluorescence measurements reflect actual Jn+ across the vesicle membrane. We had previously shown that vesicles labeled with carboxyfluorescein exhibited very high PF and that the entire fluorescent vesicle population exhibited this high permeability (9). Thus virtually all of the fluorescently
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
WATER
CHANNEL
PROTON
labeled intermediate pellet vesicles contained water channels. The ionophore gramicidin forms an aqueous channel that permits single-file diffusion of water molecules across lipid bilayers and markedly increases membrane Pn+ in a wide variety of biological and artificial membranes (3,16). To determine whether the F-dextran, like carboxyfluorescein, was entrapped in an apparently homogeneous population of vesicles, gramicidin was used to maximize Pn+. As shown in Fig. ZB, gramicidin markedly increased the rate of dissipation of the proton gradient (7 = 25 and 42 ms for two preparations with corresponding control values of 1.8 and 2.5 s) but did not alter the total change in fluorescence. Thus all intravesicular fluorescence that was sensitive to pH was measured in our stop-flow determinations and could be fitted by a single process. It is therefore unlikely that a significant population of fluorescently labeled vesicles of different Pn+ was present in our preparation. Since we have previously shown that vesicles prepared in this manner contain a very high & (9), it is highly likely that the Jn+ values measured in the present study occurred across membranes containing water channels. As shown in Fig. ZC, the rate of Jn+ was the same regardless of whether the lumen content of the vesicles or the outside medium was more acidic at the outset of the experiment (Pn+ for pH 8.0-6.0, 3.9 t 0.5 X 10s3 cm/ s; PH+ for pH 6.0-8.0, 3.9 t 0.2 X 10v3 cm/s; n = 2). Like gramicidin, the ADH water channel is thought to mediate single-file diffusion of water molecules. Because ADH water channels are abundant in membranes of these apical membrane vesicles, we tested whether ADH water channels, like gramicidin, might mediate the high vesicle Pn+. Consistent with this hypothesis, previous work by Gluck and Al-Awqati (6) showed that ADH stimulation, while simultaneously increasing transepithelial &, increased transepithelial Pn+ in toad urinary bladder by 300%. We therefore examined the activation energy of Pnf, as well as its response to inhibitors of water flow through the ADH water channel. Preincubation of toad urinary bladder with PCMBS reduces transepithelial & by 60% (12); similar effects are observed in erythrocytes and renal proximal tubule (19, 28). As shown in Fig. 3A, preincubation of vesicles with PCMBS also inhibited 55% of Pn+ in F-dextranloaded vesicles. Control Pn+ was 5.1 t 0.5 X 10e3 cm/s (n = 6 vesicle preparations), whereas that in vesicles treated with 1 mM PCMBS was 2.2 t 0.3 X 10B3cm/s (n = 5; P < 0.05 compared with control; unpaired t test). The inhibitory effect of PCMBS was not shared by another sulfhydryl reagent, NEM (2 mM; vesicles were pretreated for 1 h with NEM in HEPES buffer on ice). In paired experiments, P H+ was 5.6 t 0.2 X low3 in the absence of reagent and 6.2 t 0.9 x 10m3cm/s (NS compared with control; n = 3) after pretreatment with NEM. Unlike PCMBS, NEM pretreatment has no inhibitory effect on water flux across the toad bladder or human erythrocyte membrane (1, 12, 19, 24, 28). Phloretin, which inhibits both ADH-stimulated urea flux in toad bladder and &+ across artificial planar black lipid bilayers (7, 17), was also without effect on apical membrane vesicle &+. In paired studies, PH+ was 2.4 t 0.4 x
F369
PERMEABILITY
A I :
-
Time,
Control
set
Ea=
3.6 Kcal/mol,
-7 + aI 5 -8
3.0
3.2 l/OK
3.4 (x 1 o-3)
3.6
3.8
FIG. 3. Effects of PCMBS and temperature on proton flux across vesicle membranes. A: effect of PCMBS on rate of pH change in Fdextran-loaded vesicles. In vesicles containing 17.6 mM HEPES (pH 8.0), extravesicular pH was abruptly lowered to 6.0 after 1 h preincubation at 4°C in absence (“Control”) or presence (“PCMBS”) of 1 mM PCMBS. B: Arrhenius plot for proton transport (Pn+) in toad bladder vesicles containing ADH water channels. Measurements were performed and data was analyzed as in Fig. 2 at varying temperatures. Each point represents mean t SE of 3-4 different experiments; correlation coefficient for the fitted line was 0.979. E,, activation energy.
10B3and 2.6 t 0.4 X 10B3cm/s (n = 3) for control and phloretin-treated vesicles, respectively. Vesicles were exposed to phloretin for 30 min before assay; stopped-flow studies were performed in the continued presence of phloretin. Figure 3B shows the effect of varying temperature on vesicle PH+, charted in an Arrhenius plot. The calculated activation energy of 3.6 kcal/mol indicates &+ via channels containing water; values X0 were obtained when protons diffused across lipid bilayers (7, 13). To determine whether these PH+ properties were unique to the intermediate pellet fraction that was derived from F-dextran-loaded aggrephores, we prepared a labeled fraction derived primarily from MVB. This fraction was obtained by allowing 60 min to pass after removal of ADH and subjecting the resulting vesicles to differential centrifugation and fluorescence measurements identical to those performed in standard vesicles. In MVB, PCMBS did not alter PH+ (7 was 1.4 and 2.1 for two control preparations and 1.1 and 2.1 s for the
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
F370
WATER
CHANNEL
PROTON
same two preparations, respectively, treated with PCMBS). The activation energies for Jn+ across two MVB preparations were 11.5 and 12.4 kcal/mol, values far different from the value of 3.6 for intermediate pellet fractions. DISCUSSION
We have selectively labeled with F-dextran a population of vesicles that is known to contain water channels (g-11). Jn+ across the limiting membranes of these vesicles is inhibited by PCMBS, but not by NEM or phloretin, and exhibits a low activation energy. These properties are specific to a population of vesicles known to be rich in water channels, because Jn+ across MVB was not inhibitable by PCMBS and exhibited an activation energy similar to that observed when protons cross lipid bilayers. Because the intermediate pellet fraction includes vesicles that do not contain water channels (such as mitochondria and lysosomes) as well as vesicles derived from apical membrane of ADH-stimulated granular cells (9II), it is critical to examine the evidence that the fluorescence signal measured in the present studies is, in fact, derived from the vesicles that contain the water channel. First, exposure of bladders to F-dextran in the absence of ADH results in no measurable uptake into cells and yields an intermediate pellet vesicle fraction with no detectable fluorescein fluorescence. Thus detectable fluorescence is present only when the bladders are exposed to ADH stimulation and subsequent removal in the presence of a large transepithelial osmotic gradient. Second, the F-dextran was entrapped in vesicles under conditions that selectively place fluid-phase markers in retrieved apical membrane vesicles, many of which have morphology identical to aggrephores (11) and have been demonstrated to contain particle aggregates in intact bladders (2). Third, the vesicles containing entrapped fluorophore in the intermediate pellet fraction have been shown to have an extremely high & (4.5 X low2 cm/s, 9). Thus, although the intermediate pellet fraction contains many types of vesicles, the fluorescent label is present in vesicles that contain water channels by morphological and transport criteria. There is also evidence that all of the F-dextran-containing vesicles contain ADH water channels. We have previously demonstrated that the entire fluorescence quenching of carboxyfluorescein entrapped in the intermediate pellet occurs within 5 ms, giving a very high minimum value for pF‘ (9). The fact that all of fluorescence quenching occurred at or above the very high minimum permeability value obtained demonstrates that all of the intravesicular fluorescence was entrapped in vesicles containing water channels. In the present studies, F-dextran was entrapped in a manner identical to the methods used previously (g-11). As shown in Fig. 2B, the protonophore gramicidin increased the rate of fall in fluorescence but did not alter the total change in fluorescence, indicating that, with respect to fluorescence, the vesicle population was homogeneous. Taken together, these data indicate that the fluorescence changes observed in the present studies were occurring
PERMEABILITY
in vesicles containing water channels and not in some separate fraction. Our data suggest that protons cross the membrane of the endocytic vesicle via the ADH water channel. Vesicle pretreatment with PCMBS reduced Jn+, whereas NEM and phloretin were without effect. Vesicle Pn+ was symmetrical and could be distinguished from the diffusion of NH3 across the membrane of these vesicles. Our measured activation energy for vesicle &+ was the same as that measured for the apparent activation energy of water flow through the ADH water channel in intact toad bladders (3.0 t 2.3 kcal/mol) (13, 27) and papillary endosomes obtained from ADH-stimulated Brattleboro rats (3.8 kcal/mo1)(26). Although our data do not rule out the presence of an additional proton-permeant membrane protein in endocytic vesicles containing water channels, the movement of protons across the water channels themselves provides the simplest explanation of our results. Definitive proof of Jn+ via water channels must await the availability for study of purified reconstituted water channels in artificial membranes. Inhibition of both PF and Pn+ by PCMBS may provide a convenient assay to distinguish the ADH water channel from other membrane transporters and allow further characterization of its structural-functional relationships. Because the intrinsic Pn+ of liposome bilayers is -10Y4 cm/s (3), these studies may be extended to measurements of ADH water channel Pn+ in liposomes and artificial bilayer membranes, thereby permitting isolation and characterization of this important membrane channel. Our results raise two questions relating to effects of ADH on proton transport in the intact bladder. First, why does ADH induce a 4O-fold increase in & but only a three- to fourfold increase in Pn+ (6)? Second, if protons rapidly permeate membranes containing ADH water channels, then how can the ADH-stimulated bladder or collecting duct maintain the large proton gradients (lumen pH of 5, basolateral pH of 7.4) observed in vivo? The answers to both questions involve factors present in the bladder but not necessarily in our vesicles. In the bladder, Pn+ through mitochondria-rich cells and between cells may be higher than that for water, resulting in a more modest increment in proton flux compared with water flux upon exposure to ADH. Furthermore, the ADH water channel itself may be somewhat charge selective, limiting the flux of protons relative to water. Finally, Jn+ in cells, unlike that of water, is governed as well by charge, buffer capacity, and mechanisms responsible for the regulation of intracellular pH. All of these factors may limit the Pu+ of intact bladder granular cells and collecting duct principal cells during stimulation with ADH. The high Pn+ of these vesicles that are retrieved from the water-permeable apical membrane of ADH-stimulated granular cells has implications for the endosomal processing of membranes containing ADH water channels. Evidence from many cell types demonstrates that pH gradients play a critical role in the sorting of proteins in endosomes and the Golgi apparatus (14, 20, 25). Our results provide an explanation for data from tight epi-
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
WATER
CHANNEL
PROTON
thelial cells in toad urinary bladders as well as the mammalian papilla, where endocytic vesicles containing ADH water channels do not appear to possess acidic lumens (5, 16). To prevent the collapse of vital intracellular pH gradients during its synthesis and processing in intracellular organelles, the Pn+ of the ADH water channel might be regulated. Alternatively, the pat king density of ADH water channels might remain low until final assembly in aggrephore vesicles. Low packing densities of ADH water channels would limit localized increases in membrane P n+ and enable existing intracellular acidification mechanisms to maintain organelle pH gradients. Retrieval and endosomal processing of ADH water channel membrane might then occur through specialized pathways not requiring pH gradients. We thank C. Hosselet for technical assistance and Drs. James Turner, Kevin Strange, Joseph Handler, Barry M. Brenner, and Steven Hebert for helpful discussions. H. W. Harris is a Clinician-Scientist and Established Investigator awardee of the American Heart Association. M. L. Zeidel is the recipient of research career development and merit review awards from the Department of Veterans Affairs of the United States. D. Kikeri is the recipient of an individual National Institutes of Health National Research Service Award. A. Janoshazi is a fellow of the American Heart Association, Massachusetts Affiliate. This work was supported by National Institute of Diabetes and Digestive and Kidney Disease Grants RO-1 DK-38874 to H. W. Harris and RO-1 DK-706882 to M. L. Zeidel, and was also supported by the Harvard Center for the Study of Kidney Diseases, and grants-in-aid from the American Heart Association, and the Squibb Institute for Medical Research to A. K. Solomon. Address for reprint requests: H. W. Harris, Division of Nephrology, The Children’s Hospital, Boston, MA 02115.
10.
11.
12.
13.
14. 15. 16.
17.
18.
19.
20. Received
30 November
1989; accepted
in final
form
20 March
1990.
REFERENCES
21.
1. CHEVALIER, J., N. ADRAGNA, J. BOURGUET, AND R. GOBIN. Fine structure of intramembranous particle aggregates in ADH-treated frog urinary bladder and skin: influence of glutaraldehyde and NEM. CeLL Tissue Res. 218: 595-603, 1981. 2. COLEMAN, R. C., H. W. HARRIS, AND J. B. WADE. Visualization of endocytosed markers in freeze-fracture studies of toad urinary bladder. J. Histochem. Cytochem. 35: 1405-1414, 1987. 3. DEAMER, D. W., AND J. W. NICHOLS. Proton flux mechanisms in model and biological membranes. J. Membr. Biol. 107: 91-103, 1989. A. Water Movement Through Lipid Bilayers, Pores 4. FINKELSTEIN, and Plasma Membranes, Theory and Reality. New York: Wiley and Sons, 1986. 5. FRANKI, N., G. DING, AND R. M. HAYS. The rate and pH of the aggrephore cycling system (Abstract). Kidney Int. 35: 187, 1989. 6. GLUCK, S., AND Q. AL-AWQATI. Vasopressin increases water permeability by inducing pores. Nature Lond. 284: 631-632, 1980. 7. GUTKNECHT, J. Proton/hydroxide conductance and permeability through phospholipid bilayers membranes. Proc. Natl. Acad. Sci. USA 84: 6443-6446, 1987. 8. HARRIS, H. W., AND J. S. HANDLER. The role of membrane turnover in the water permeability response to antidiuretic hormone. J. Membr. Biol. 103: 207-215, 1988. 9. HARRIS, H. W., J. S. HANDLER, AND R. BLUMENTHAL. Apical
22.
23. 24.
25. 26.
27,
28.
PERMEABILITY
F371
membrane vesicles of ADH-stimulated toad bladder are highly water permeable. Am. J. Physiol. 258 (Renal Fluid Electrolyte Physiol. 27): F237-F243, 1990. HARRIS, H. W., H. R. MURPHY, M. C. WILLINGHAM, AND J. S. HANDLER. Isolation and characterization of specialized regions of toad urinary bladder apical plasma membrane involved in the water permeability response to antidiuretic hormone. J. Membr. Biol. 96: 175-186, 1987. HARRIS, H. W., J. B. WADE, AND J. S. HANDLER. Fluorescent markers to study membrane retrieval in antidiuretic hormonetreated toad urinary bladder. Am. J. Physiol. 251 (Cell Physiol. 20): C274-C284,1986. HOCH, B., P. C. GORFEIN, D. LINZER, M. J. Fusco, AND S. D. LEVINE. Mercurial reagents inhibit flow through ADH-induced water channels in toad bladder. Am. J. Physiol. 256 (Renal Fluid Electrolyte Physiol. 25): F948-F953, 1989. KACHADORIAN, W. A., J. MULLER, S. W. RUDICH, AND V. A. DISCALA. Temperature dependence of ADH-induced water flow and intramembranous particle aggregates in toad bladder. Science Wash. DC 205: 910-913, 1979. KELLY, R. B. Pathways of protein secretion in eukaryotes. Nature Lond. 326: 14-16, 1987, KIKERI, D., A. SUN, M. L. ZEIDEL, AND S. C. HEBERT. Cell membranes impermeable to NH,. Nature Lond. 339: 478-480,1989. LENCER, W. I., D. BROWN, M. A. ARNOUT, D. A. AUSIELLO, AND A. S. VERKMAN. Endocytic vesicles from renal papilla which cycle the vasopressin-sensitive water channel do not contain H-ATPase (Abstract). J. Cell Biol. 107: 810, 1988. LEVINE, S.D., N. FRANKI, AND R. M. HAYS. Effect of phloretin on water and solute movement in the toad urinary bladder. J. Clin. Invest. 52: 1435-1442, 1973. MULLER, J., W. A. KACHADORIAN, AND V. A. DISCALA. Evidence that ADH stimulated intramembranous particle aggregates are transferred from cytoplasmic to luminal membranes in toad bladder epithelial cells. J. Cell Biol. 85: 83-95, 1980. OJCIUS, D. M., AND A. K. SOLOMON. Sites ofp-chloromercuribenzenesulfonate inhibition of red cell urea and water transport. Biochim. Biophys. Acta 942: 73-82, 1988. ORCI, L., M. RAVAZZOLA, M.-J. STORCH, R. G. W. ANDERSON’ J.D. VASSALLI, AND A. PARRELET. Proteolytic maturation of insulin is a post-golgi event which occurs in acidifying clathrin-coated secretory vesicles. Cell 49: 865-868, 1987. PARISI, M., J. BOURGUET, P. RIPOCHE, AND J. CHEVALIER. Simultaneous minute to minute determination of unidirectional and net water fluxes in frog urinary bladder. A reexamination of the two barriers in series hypothesis. Biochim. Biophys. Acta 556: 509523, 1979. SHI, L.-B., AND A. S. VERKMAN. Very high water permeability in vasopressin-induced endocytic vesicles from toad urinary bladders. J. Gen. Physiol. 94: 1101-1115, 1989. STRANGE, K., AND K. SPRING. Cell membrane water permeability of rabbit cortical collecting duct. J. Membr. Biol. 96: 27-43, 1987. TOON, M. R., AND A. K. SOLOMON. Control of red cell urea and water permeability by sulfhydryl reagents. Biochim. Biophys. Acta 860: 361-375,1986. TYCKO, B., AND F. R. MAXFIELD. Rapid acidification of endocytic vesicles containing cy2 macroglobulin. Cell 28: 643-651, 1982. VERKMAN, A. S., W. I. LENCER, D. BROWN, AND D. A. AUSIELLO. Endosomes from kidney collecting tubule cells contain the vasopressin-sensitive water channel. Nature Lond. 333: 268-269, 1988. WANG, J. H., C. V. ROBINSON’ AND I. S. EDELMAN. Self-diffusion and structure of liquid water. III. Measurement of self-diffusion of liquid water with 2H7 “H, and “0 as tracers. J. Am. Chem. Sot. 75: 466-470, 1955. WHITTEMBURY, G., P. CARPI-MEDINA, E. GONZALEZ, AND H. LINARES. Effect of para-chloromercuribenzenesulfonic acid and temperature on cell water osmotic permeability of proximal straight tubules. Biochim. Biowhvs. Acta 775: 365-373, 1984.
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
containing
H. WILLIAM HARRIS, JR., DEEPAK KIKERI, AGNES JANOSHAZI, A. K. SOLOMON, AND MARK L. ZEIDEL Renal Divisions, Brocktonl West Roxbury Veterans Administration Medical Center, West Roxbury The Children’s Hospital, Brigham and Women’s Hospitals, and Biophysical Laboratory, Harvard Medical School, Boston, Massachusetts 02115
HARRIS, H. WILLIAM, AZI, A. K. SOLOMON,
JR., DEEPAK KIKERI, AND MARK L. ZEIDEL.
AGNES
JANOSH-
High proton flux through membranes containing antidiuretic hormone water channels. Am. J. Physiol. 259 (Renal Fluid Electrolyte Physiol. 28): F366-F371, 1990.-Antidiuretic hormone (ADH) stimulation of toad urinary bladder granular cells causes simultaneous increases in transepithelial water and H’ permeabilities (& and PH+, respectively), suggesting that ADH-elicited water channels inserted into granular cell apical membranes might be permeable to both water and H+. We have previously used self-quenching fluorophores entrapped within endocytic vesicles selectively retrieved from water-permeable apical membranes to measure vesicle PF. The membranes of these vesicles possess an extremely high PF such that our measurements provide only minimum estimates of vesicle PF and have limited our ability to quantitate the properties of ADH water channels. We therefore quantitated vesicle PH+ using similar rapid mixing techniques. Vesicle P H+ was 5.1 t 0.5 X lo-" cm/s. Activation energy of this process was 3.6 t 0.6 kcal/mol, indicative of H+ flux through an aqueous channel. The mercurial reagent, parachloromercuribenzenesulfonate (PCMBS), which inhibits ADH-stimulated transepithelial PF in intact bladders by 50and 60%, inhibited vesicle PH+ by 55%. N-Ethylmaleimide phloretin, which do not alter ADH-stimulated PF, did not affect vesicle PH+. We conclude that membranes containing ADH water channels possess substantial PH+ that likely reflects proton flux through water channels. The apparent high PH+ of the ADH water channel may have important implications for intracellular trafficking of these water channels in ADH-responsive epithelial cells. water permeability; bules; endosome
proton
permeability;
kidney;
collecting
tu-
HORMONE (ADH) markedly increases apical membrane water permeability (PF) of certain tight epithelial cells by vesicle-mediated insertion of unique, highly selective water channels (21, 23). These water channels allow passage of water, and possibly protons, while excluding larger ions and nonelectrolytes (4, 23). Efforts to isolate and characterize ADH water channels have been hampered by the relatively high PF of synthetic and biological membranes (10s3 cm/s) and by the lack of a specific assay that readily discriminates between ADH water channels and other membrane transporters (4). In toad bladder granular cells, water channels are stored in large tubular cytoplasmic vesicles called agANTIDIURETIC
02132;
grephores (9, 18). ADH stimulation causes aggrephore fusion with the granular cell apical plasma membrane and a dramatic increase in pF and proton permeability (Pn+) (6, 8). Termination of ADH stimulation causes apical membrane endocytosis and a fall in & and Pn+ to their low basal levels (8). We have previously demonstrated that fluorescent markers are selectively entrapped in the lumens of these endocytic vesicles that can be isolated from granular cell homogenates (2, 8-10, 11). The pF of these vesicles is greater than 4.5 x lo-’ cm/s at %“C, one of the highest known & values observed in biological membranes (9,22). Because the time resolution of existing instrumentation currently allows measurement of only minimum pF values in these vesicles (9), it is difficult to characterize the properties of water flow across the ADH water channel and to use water flow as an assay for successful reconstitution of water channel activity. We have therefore developed methods to measure proton flux across the membranes of these vesicles. Using a pH-sensitive fluorophore trapped within water channel-containing vesicles (9, 11)) we have examined the Pn+ properties of these vesicles and provide evidence that the ADH water channel mediates proton flux across the limiting membranes of these vesicles. Measurement of proton flux may be useful both in the characterization of ADH water channels and, ultimately, in their purification. MATERIALS
AND
METHODS
Dominican toads (Bufo marinus) were obtained from National Reagents, Bridgeport, CT. Urinary bladders of Bufo marinus were prepared and mounted on cannulas as small sacs.After a ZO-min interval of ADH stimulation (50 mu/ml), the apical surfaces of bladders were exposed to solutions containing 50 mg/ml of fluorescein dextran (F-dextran; av mol mass of 70,000 Daltons) and O-20 mM NaCl, 30 mM KCl, and l-20 mM N-2-hydroxyethylpiperazine-2\r’-2-ethanesulfonic acid (HEPES), pH 8.0, for an interval of 5 min followed by termination of ADH stimulation and incubation for an additional 10 min. The solution containing the F-dextran was then removed, the bladders were thoroughly rinsed to remove all extracellular F-dextran, and scraped epithelial cells were homogenized in a Polytron (Brinkmann Instruments, Westbury, NY) for 5 s at a setting of 7. The supernatant of a
F366
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
WATER
CHANNEL
PROTON
low-speed spin (700 g for 15 min) was centrifuged at 10,000 g for 10 min, and the pellet (“intermediate pellet”) used for all subsequent studies (9, 10) unless specified otherwise. In the absence of ADH, no detectable fluorescence was present in this vesicle fraction despite apical membrane exposure to F-dextran. In some experiments, bladders were incubated for 60 min after withdrawal of ADH, and the trapped F-dextran-containing vesicles were isolated as described above. Previous studies have demonstrated that this procedure results in entrapment of F-dextran within a morphologically distinct endosoma1 compartment, the multivesicular body (MVB) (2, 11). Fluorescence studies with MVB preparations were performed using the exact same protocols as those with the standard preparation. Standard cuvette fluorescence measurements were performed by use of an SLM-Aminco 500C spectrofluorometer (excitation wavelength 499 nm, emission wavelength 520 nm, slits 2 nm). Unless stated otherwise, all experiments were performed at 25OC. Stopped-flow measurements were performed in a High Tech stop-flow device (20-ms dead time) connected to a Photon Technologies Alphascan fluorometer. One syringe contain’& a concentrated vesicle suspension and anti-fluorescein antibody, whereas the other contained buffer with sufficient HCl or NaOH to change the final extravesicular pH as indicated. Data from 3-5 determinations performed on a single day were averaged. Results were fitted to single exponentials and the resulting time constant, 7, was converted to permeability by use of the following equation JH+
= I’H+( AC)A
where JH+ is the flux rate of protons in mol/s, Pn+ is the permeability coefficient of protons in cm/s, AC is the concentration gradient for protons across the vesicle membrane at the start of the experiment in mol/cm3, and A is the surface area of a single F-dextran-containing vesicle (7.1 X 10-l’ cm2; Refs. 9 and 10). JH+ was calculated by multiplying l/7 by the amount of buffer (in moles) in an individual vesicle (single vesicle volume, 1.77 x lo-l5 cm3). Estimates of vesicle volume and surface area were calculated from dimensions of vesicles loaded with horseradish peroxidase (HRP) instead of Fdextran. Previous studies (9-11) have verified these two markers have identical vesicle-loading characteristics. Electron microscopic examinations of HRP-loaded vesicles show that 89% are spherical in shape with a radius of 7.5 x lo-” cm and a volume-to-surface area ratio of 2.5 X 10m6cm. When vesicles were treated with the sulfhydryl-reactive reagents, para-chloromercuribenzenesulfonate (PCMBS) and N-ethylmaleimide (NEM), they were exposed to the compound (1 mM PCMBS or 2 mM NEM) or vehicle (buffer) for 30 min on ice in HEPES buffer before JH+ measurements. Vesicles were then diluted into buffer in the absence or presence of inhibitors and assayed as described. Phloretin experiments were performed after addition of 0.1 mM phloretin to both the intravesicular contents and extracellular solution of Fdextran-containing vesicles. All chemicals were pur-
PERMEABILITY
F367
chased from Sigma (St. Louis, MO) and were of analytical grade. RESULTS
To assure that our measurements truly reflected changes in intravesicular pH (pHi), all experiments were performed after the addition of anti-fluorescein antibody. As shown in Fig. lA, this antibody bound to fluorescein and quenched virtually all extravesicular fluorescence. Subsequent addition of Triton X-100 lysed the vesicles and quenched F-dextran fluorescence such that, in the presence of antibody, nearly all of the fluorescence signal could be accounted for as intravesicular fluorescein. Thus all subsequent experiments were performed in the presence of the antibody to eliminate extravesicular fluorescence. Intravesicular fluorescence correlated linearly with pH over the pH range of 6.5-8.0, permitting estimation of pHi. As shown in Fig. lB, addition of ATP (tracing A) or Na+ (tracing B) to the extravesicular medium did not alter pHi at 25OC. The failure of ATP or Na+ to change pHi may result from either 1) the absence of vesicle Na+H+ exchange or H+-adenosinetriphosphatase (ATPase) activities or 2) vesicle membrane Pi--i+ that is too high (relative to the rates of H+ transport pathways) to permit generation of measurable pH gradients. To test this latter hypothesis, extravesicular pH was abruptly acidified from pH 8.0 to 6.0 at 25OC. (Fig. lC), resulting in rapid acidification of pHi. As shown in Fig. 1B (tracing C) the inward flux of protons was so rapid that extravesicular addition of NH&l at 25°C produced no change in pHi, presumably because the rate of JH+ balanced the rate of NH3 influx. JH+ was markedly slowed when extravesicular pH was abruptly acidified at 5°C rather than 25°C (Fig. 1C). Temperature reduction slowed JH+ and allowed detection of intravesicular alkalization after addition of NH&l (Fig. lB, tracing D). Vesicle JH+ was symmetrical, regardless of whether the interior or exterior was initially more acidic (see below). Thus any pH gradients generated by vesicle proton transport pathways such as Na+H+ exchange or a H+-ATPase might have been rapidly dissipated. It is also possible that NH: itself entered the vesicles as rapidly as NH,; we have recently demonstrated this phenomenon in mouse thick ascending limb of Henle (15). Figure 2 displays measurements of vesicle JH+ under stopped-flow conditions. As shown in Fig. 2A, the rate of decay of the imposed pH gradient varied with changes in the intravesicular buffer concentration. When the lumens of F-dextran-loaded vesicles were buffered with 20 mM HEPES, the time constant of the decline in pHi was given by a 7 of 7.2 s and yielded a Pn+ of 2.5 X 10m3 cm/s. When vesicles were loaded and maintained in 2 mM HEPES, the 7 was 0.8 s, giving a Pn+ of 2.2 X 10m3 cm/s. The close agreement of &-i+ values under conditions of low and high intravesicular HEPES concentrations indicates that, in this range of added buffer, the added buffer itself (and not other vesicle components) dominates vesicle buffer capacity. The change in 7 and the constancy of PH+ as a function of intravesicular buffer
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
F368
WATER
CHANNEL
PROTON
PERMEABILITY
A
2*4T
-
Low Buffer
- -
High
Buffer
1.8
0.6
.
0
1
.
2
Add NH4+
3
l
25 OC
4
5
6
7
8
Control - -. Gramicidin
C
Add
Na+ or ATP
L
0 c 0
C
tI
II
II
II
I
10
20
30
40
50
I
. ..
..
l
e
lb
C
25'C 5OC
.
l -
.
. l
l
.
. l
l
.
. “.***
.
.
.
.
. .
.
.
.
. .
20
ii
. l
10
4-
. l
0
i
30 Time,
40
50
.
60
3ec
FIG. 1. Measurement of intravesicular fluorescence of apical membrane vesicles loaded with fluorescein dextran (F-dextran). A: effect of addition of anti-fluorescein antibody and Triton X-100 on fluorescence of vesicles loaded with F-dextran. In 5 experiments, antibody addition reduced fluorescence of vesicle suspension by 22 -+ 7% (n = 5; SD); addition of Triton X-100 left 18 t 12% of fluorescence at 1 min and 4 t 0% at 12 min. Fluorescence is displayed in arbitrary units. Antibodymediated quenching was unaffected by changes in pH in the range of 5.5-9.0 or additions of Triton X-100 and NaCl to final concentrations of 4% and 2.0 M, respectively. B: effect of extravesicular addition (downward pointing arrows) of 9 mM ATP (tracing A), 20 mM NaCl (tracing B), 20 mM NH&l at 25°C (tracing C), or at 5°C (tracing D), on fluorescence of intravesicular F-dextran. Curves were displaced along y-axis for visual clarity. C: effect of changing extravesicular pH from 8.0 to 6.0 (at the vertical arrow) on intravesicular fluorescence at 25°C and 5°C. Extravesicular pH was lowered by abrupt addition of a small volume of 2 N HCl.
FIG. 2. Stopped-flow measurement of the rate of change of intravesicular pH in apical membrane vesicles. A: effect of buffer concentration on rate of pH change in F-dextran-loaded vesicles. Vesicles were loaded, prepared, and maintained in 20 mM (“high buffer”) or 2 mM (“low buffer”) HEPES buffer at pH 8.0; at start of experiment, pH outside was abruptly lowered to 6.0. Curves were displaced along yaxis for visual clarity. B: effect of gramicidin on proton flux (&+). Vesicles were preincubated for ~2 min in presence or absence of 5 PM gramicidin before assay of Jn+. These results are representative of identical experiments performed on 2 separate preparations. C: symmetry of inward and outward Jn+s in F-dextran-loaded vesicles. Vesicles were preincubated at pH 8.0 and diluted into pH 6.0 (tracing A), or preincubated at pH 6.0 and diluted to pH 8.0 (tracing B).
concentration indicate that our fluorescence measurements reflect actual Jn+ across the vesicle membrane. We had previously shown that vesicles labeled with carboxyfluorescein exhibited very high PF and that the entire fluorescent vesicle population exhibited this high permeability (9). Thus virtually all of the fluorescently
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
WATER
CHANNEL
PROTON
labeled intermediate pellet vesicles contained water channels. The ionophore gramicidin forms an aqueous channel that permits single-file diffusion of water molecules across lipid bilayers and markedly increases membrane Pn+ in a wide variety of biological and artificial membranes (3,16). To determine whether the F-dextran, like carboxyfluorescein, was entrapped in an apparently homogeneous population of vesicles, gramicidin was used to maximize Pn+. As shown in Fig. ZB, gramicidin markedly increased the rate of dissipation of the proton gradient (7 = 25 and 42 ms for two preparations with corresponding control values of 1.8 and 2.5 s) but did not alter the total change in fluorescence. Thus all intravesicular fluorescence that was sensitive to pH was measured in our stop-flow determinations and could be fitted by a single process. It is therefore unlikely that a significant population of fluorescently labeled vesicles of different Pn+ was present in our preparation. Since we have previously shown that vesicles prepared in this manner contain a very high & (9), it is highly likely that the Jn+ values measured in the present study occurred across membranes containing water channels. As shown in Fig. ZC, the rate of Jn+ was the same regardless of whether the lumen content of the vesicles or the outside medium was more acidic at the outset of the experiment (Pn+ for pH 8.0-6.0, 3.9 t 0.5 X 10s3 cm/ s; PH+ for pH 6.0-8.0, 3.9 t 0.2 X 10v3 cm/s; n = 2). Like gramicidin, the ADH water channel is thought to mediate single-file diffusion of water molecules. Because ADH water channels are abundant in membranes of these apical membrane vesicles, we tested whether ADH water channels, like gramicidin, might mediate the high vesicle Pn+. Consistent with this hypothesis, previous work by Gluck and Al-Awqati (6) showed that ADH stimulation, while simultaneously increasing transepithelial &, increased transepithelial Pn+ in toad urinary bladder by 300%. We therefore examined the activation energy of Pnf, as well as its response to inhibitors of water flow through the ADH water channel. Preincubation of toad urinary bladder with PCMBS reduces transepithelial & by 60% (12); similar effects are observed in erythrocytes and renal proximal tubule (19, 28). As shown in Fig. 3A, preincubation of vesicles with PCMBS also inhibited 55% of Pn+ in F-dextranloaded vesicles. Control Pn+ was 5.1 t 0.5 X 10e3 cm/s (n = 6 vesicle preparations), whereas that in vesicles treated with 1 mM PCMBS was 2.2 t 0.3 X 10B3cm/s (n = 5; P < 0.05 compared with control; unpaired t test). The inhibitory effect of PCMBS was not shared by another sulfhydryl reagent, NEM (2 mM; vesicles were pretreated for 1 h with NEM in HEPES buffer on ice). In paired experiments, P H+ was 5.6 t 0.2 X low3 in the absence of reagent and 6.2 t 0.9 x 10m3cm/s (NS compared with control; n = 3) after pretreatment with NEM. Unlike PCMBS, NEM pretreatment has no inhibitory effect on water flux across the toad bladder or human erythrocyte membrane (1, 12, 19, 24, 28). Phloretin, which inhibits both ADH-stimulated urea flux in toad bladder and &+ across artificial planar black lipid bilayers (7, 17), was also without effect on apical membrane vesicle &+. In paired studies, PH+ was 2.4 t 0.4 x
F369
PERMEABILITY
A I :
-
Time,
Control
set
Ea=
3.6 Kcal/mol,
-7 + aI 5 -8
3.0
3.2 l/OK
3.4 (x 1 o-3)
3.6
3.8
FIG. 3. Effects of PCMBS and temperature on proton flux across vesicle membranes. A: effect of PCMBS on rate of pH change in Fdextran-loaded vesicles. In vesicles containing 17.6 mM HEPES (pH 8.0), extravesicular pH was abruptly lowered to 6.0 after 1 h preincubation at 4°C in absence (“Control”) or presence (“PCMBS”) of 1 mM PCMBS. B: Arrhenius plot for proton transport (Pn+) in toad bladder vesicles containing ADH water channels. Measurements were performed and data was analyzed as in Fig. 2 at varying temperatures. Each point represents mean t SE of 3-4 different experiments; correlation coefficient for the fitted line was 0.979. E,, activation energy.
10B3and 2.6 t 0.4 X 10B3cm/s (n = 3) for control and phloretin-treated vesicles, respectively. Vesicles were exposed to phloretin for 30 min before assay; stopped-flow studies were performed in the continued presence of phloretin. Figure 3B shows the effect of varying temperature on vesicle PH+, charted in an Arrhenius plot. The calculated activation energy of 3.6 kcal/mol indicates &+ via channels containing water; values X0 were obtained when protons diffused across lipid bilayers (7, 13). To determine whether these PH+ properties were unique to the intermediate pellet fraction that was derived from F-dextran-loaded aggrephores, we prepared a labeled fraction derived primarily from MVB. This fraction was obtained by allowing 60 min to pass after removal of ADH and subjecting the resulting vesicles to differential centrifugation and fluorescence measurements identical to those performed in standard vesicles. In MVB, PCMBS did not alter PH+ (7 was 1.4 and 2.1 for two control preparations and 1.1 and 2.1 s for the
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
F370
WATER
CHANNEL
PROTON
same two preparations, respectively, treated with PCMBS). The activation energies for Jn+ across two MVB preparations were 11.5 and 12.4 kcal/mol, values far different from the value of 3.6 for intermediate pellet fractions. DISCUSSION
We have selectively labeled with F-dextran a population of vesicles that is known to contain water channels (g-11). Jn+ across the limiting membranes of these vesicles is inhibited by PCMBS, but not by NEM or phloretin, and exhibits a low activation energy. These properties are specific to a population of vesicles known to be rich in water channels, because Jn+ across MVB was not inhibitable by PCMBS and exhibited an activation energy similar to that observed when protons cross lipid bilayers. Because the intermediate pellet fraction includes vesicles that do not contain water channels (such as mitochondria and lysosomes) as well as vesicles derived from apical membrane of ADH-stimulated granular cells (9II), it is critical to examine the evidence that the fluorescence signal measured in the present studies is, in fact, derived from the vesicles that contain the water channel. First, exposure of bladders to F-dextran in the absence of ADH results in no measurable uptake into cells and yields an intermediate pellet vesicle fraction with no detectable fluorescein fluorescence. Thus detectable fluorescence is present only when the bladders are exposed to ADH stimulation and subsequent removal in the presence of a large transepithelial osmotic gradient. Second, the F-dextran was entrapped in vesicles under conditions that selectively place fluid-phase markers in retrieved apical membrane vesicles, many of which have morphology identical to aggrephores (11) and have been demonstrated to contain particle aggregates in intact bladders (2). Third, the vesicles containing entrapped fluorophore in the intermediate pellet fraction have been shown to have an extremely high & (4.5 X low2 cm/s, 9). Thus, although the intermediate pellet fraction contains many types of vesicles, the fluorescent label is present in vesicles that contain water channels by morphological and transport criteria. There is also evidence that all of the F-dextran-containing vesicles contain ADH water channels. We have previously demonstrated that the entire fluorescence quenching of carboxyfluorescein entrapped in the intermediate pellet occurs within 5 ms, giving a very high minimum value for pF‘ (9). The fact that all of fluorescence quenching occurred at or above the very high minimum permeability value obtained demonstrates that all of the intravesicular fluorescence was entrapped in vesicles containing water channels. In the present studies, F-dextran was entrapped in a manner identical to the methods used previously (g-11). As shown in Fig. 2B, the protonophore gramicidin increased the rate of fall in fluorescence but did not alter the total change in fluorescence, indicating that, with respect to fluorescence, the vesicle population was homogeneous. Taken together, these data indicate that the fluorescence changes observed in the present studies were occurring
PERMEABILITY
in vesicles containing water channels and not in some separate fraction. Our data suggest that protons cross the membrane of the endocytic vesicle via the ADH water channel. Vesicle pretreatment with PCMBS reduced Jn+, whereas NEM and phloretin were without effect. Vesicle Pn+ was symmetrical and could be distinguished from the diffusion of NH3 across the membrane of these vesicles. Our measured activation energy for vesicle &+ was the same as that measured for the apparent activation energy of water flow through the ADH water channel in intact toad bladders (3.0 t 2.3 kcal/mol) (13, 27) and papillary endosomes obtained from ADH-stimulated Brattleboro rats (3.8 kcal/mo1)(26). Although our data do not rule out the presence of an additional proton-permeant membrane protein in endocytic vesicles containing water channels, the movement of protons across the water channels themselves provides the simplest explanation of our results. Definitive proof of Jn+ via water channels must await the availability for study of purified reconstituted water channels in artificial membranes. Inhibition of both PF and Pn+ by PCMBS may provide a convenient assay to distinguish the ADH water channel from other membrane transporters and allow further characterization of its structural-functional relationships. Because the intrinsic Pn+ of liposome bilayers is -10Y4 cm/s (3), these studies may be extended to measurements of ADH water channel Pn+ in liposomes and artificial bilayer membranes, thereby permitting isolation and characterization of this important membrane channel. Our results raise two questions relating to effects of ADH on proton transport in the intact bladder. First, why does ADH induce a 4O-fold increase in & but only a three- to fourfold increase in Pn+ (6)? Second, if protons rapidly permeate membranes containing ADH water channels, then how can the ADH-stimulated bladder or collecting duct maintain the large proton gradients (lumen pH of 5, basolateral pH of 7.4) observed in vivo? The answers to both questions involve factors present in the bladder but not necessarily in our vesicles. In the bladder, Pn+ through mitochondria-rich cells and between cells may be higher than that for water, resulting in a more modest increment in proton flux compared with water flux upon exposure to ADH. Furthermore, the ADH water channel itself may be somewhat charge selective, limiting the flux of protons relative to water. Finally, Jn+ in cells, unlike that of water, is governed as well by charge, buffer capacity, and mechanisms responsible for the regulation of intracellular pH. All of these factors may limit the Pu+ of intact bladder granular cells and collecting duct principal cells during stimulation with ADH. The high Pn+ of these vesicles that are retrieved from the water-permeable apical membrane of ADH-stimulated granular cells has implications for the endosomal processing of membranes containing ADH water channels. Evidence from many cell types demonstrates that pH gradients play a critical role in the sorting of proteins in endosomes and the Golgi apparatus (14, 20, 25). Our results provide an explanation for data from tight epi-
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
WATER
CHANNEL
PROTON
thelial cells in toad urinary bladders as well as the mammalian papilla, where endocytic vesicles containing ADH water channels do not appear to possess acidic lumens (5, 16). To prevent the collapse of vital intracellular pH gradients during its synthesis and processing in intracellular organelles, the Pn+ of the ADH water channel might be regulated. Alternatively, the pat king density of ADH water channels might remain low until final assembly in aggrephore vesicles. Low packing densities of ADH water channels would limit localized increases in membrane P n+ and enable existing intracellular acidification mechanisms to maintain organelle pH gradients. Retrieval and endosomal processing of ADH water channel membrane might then occur through specialized pathways not requiring pH gradients. We thank C. Hosselet for technical assistance and Drs. James Turner, Kevin Strange, Joseph Handler, Barry M. Brenner, and Steven Hebert for helpful discussions. H. W. Harris is a Clinician-Scientist and Established Investigator awardee of the American Heart Association. M. L. Zeidel is the recipient of research career development and merit review awards from the Department of Veterans Affairs of the United States. D. Kikeri is the recipient of an individual National Institutes of Health National Research Service Award. A. Janoshazi is a fellow of the American Heart Association, Massachusetts Affiliate. This work was supported by National Institute of Diabetes and Digestive and Kidney Disease Grants RO-1 DK-38874 to H. W. Harris and RO-1 DK-706882 to M. L. Zeidel, and was also supported by the Harvard Center for the Study of Kidney Diseases, and grants-in-aid from the American Heart Association, and the Squibb Institute for Medical Research to A. K. Solomon. Address for reprint requests: H. W. Harris, Division of Nephrology, The Children’s Hospital, Boston, MA 02115.
10.
11.
12.
13.
14. 15. 16.
17.
18.
19.
20. Received
30 November
1989; accepted
in final
form
20 March
1990.
REFERENCES
21.
1. CHEVALIER, J., N. ADRAGNA, J. BOURGUET, AND R. GOBIN. Fine structure of intramembranous particle aggregates in ADH-treated frog urinary bladder and skin: influence of glutaraldehyde and NEM. CeLL Tissue Res. 218: 595-603, 1981. 2. COLEMAN, R. C., H. W. HARRIS, AND J. B. WADE. Visualization of endocytosed markers in freeze-fracture studies of toad urinary bladder. J. Histochem. Cytochem. 35: 1405-1414, 1987. 3. DEAMER, D. W., AND J. W. NICHOLS. Proton flux mechanisms in model and biological membranes. J. Membr. Biol. 107: 91-103, 1989. A. Water Movement Through Lipid Bilayers, Pores 4. FINKELSTEIN, and Plasma Membranes, Theory and Reality. New York: Wiley and Sons, 1986. 5. FRANKI, N., G. DING, AND R. M. HAYS. The rate and pH of the aggrephore cycling system (Abstract). Kidney Int. 35: 187, 1989. 6. GLUCK, S., AND Q. AL-AWQATI. Vasopressin increases water permeability by inducing pores. Nature Lond. 284: 631-632, 1980. 7. GUTKNECHT, J. Proton/hydroxide conductance and permeability through phospholipid bilayers membranes. Proc. Natl. Acad. Sci. USA 84: 6443-6446, 1987. 8. HARRIS, H. W., AND J. S. HANDLER. The role of membrane turnover in the water permeability response to antidiuretic hormone. J. Membr. Biol. 103: 207-215, 1988. 9. HARRIS, H. W., J. S. HANDLER, AND R. BLUMENTHAL. Apical
22.
23. 24.
25. 26.
27,
28.
PERMEABILITY
F371
membrane vesicles of ADH-stimulated toad bladder are highly water permeable. Am. J. Physiol. 258 (Renal Fluid Electrolyte Physiol. 27): F237-F243, 1990. HARRIS, H. W., H. R. MURPHY, M. C. WILLINGHAM, AND J. S. HANDLER. Isolation and characterization of specialized regions of toad urinary bladder apical plasma membrane involved in the water permeability response to antidiuretic hormone. J. Membr. Biol. 96: 175-186, 1987. HARRIS, H. W., J. B. WADE, AND J. S. HANDLER. Fluorescent markers to study membrane retrieval in antidiuretic hormonetreated toad urinary bladder. Am. J. Physiol. 251 (Cell Physiol. 20): C274-C284,1986. HOCH, B., P. C. GORFEIN, D. LINZER, M. J. Fusco, AND S. D. LEVINE. Mercurial reagents inhibit flow through ADH-induced water channels in toad bladder. Am. J. Physiol. 256 (Renal Fluid Electrolyte Physiol. 25): F948-F953, 1989. KACHADORIAN, W. A., J. MULLER, S. W. RUDICH, AND V. A. DISCALA. Temperature dependence of ADH-induced water flow and intramembranous particle aggregates in toad bladder. Science Wash. DC 205: 910-913, 1979. KELLY, R. B. Pathways of protein secretion in eukaryotes. Nature Lond. 326: 14-16, 1987, KIKERI, D., A. SUN, M. L. ZEIDEL, AND S. C. HEBERT. Cell membranes impermeable to NH,. Nature Lond. 339: 478-480,1989. LENCER, W. I., D. BROWN, M. A. ARNOUT, D. A. AUSIELLO, AND A. S. VERKMAN. Endocytic vesicles from renal papilla which cycle the vasopressin-sensitive water channel do not contain H-ATPase (Abstract). J. Cell Biol. 107: 810, 1988. LEVINE, S.D., N. FRANKI, AND R. M. HAYS. Effect of phloretin on water and solute movement in the toad urinary bladder. J. Clin. Invest. 52: 1435-1442, 1973. MULLER, J., W. A. KACHADORIAN, AND V. A. DISCALA. Evidence that ADH stimulated intramembranous particle aggregates are transferred from cytoplasmic to luminal membranes in toad bladder epithelial cells. J. Cell Biol. 85: 83-95, 1980. OJCIUS, D. M., AND A. K. SOLOMON. Sites ofp-chloromercuribenzenesulfonate inhibition of red cell urea and water transport. Biochim. Biophys. Acta 942: 73-82, 1988. ORCI, L., M. RAVAZZOLA, M.-J. STORCH, R. G. W. ANDERSON’ J.D. VASSALLI, AND A. PARRELET. Proteolytic maturation of insulin is a post-golgi event which occurs in acidifying clathrin-coated secretory vesicles. Cell 49: 865-868, 1987. PARISI, M., J. BOURGUET, P. RIPOCHE, AND J. CHEVALIER. Simultaneous minute to minute determination of unidirectional and net water fluxes in frog urinary bladder. A reexamination of the two barriers in series hypothesis. Biochim. Biophys. Acta 556: 509523, 1979. SHI, L.-B., AND A. S. VERKMAN. Very high water permeability in vasopressin-induced endocytic vesicles from toad urinary bladders. J. Gen. Physiol. 94: 1101-1115, 1989. STRANGE, K., AND K. SPRING. Cell membrane water permeability of rabbit cortical collecting duct. J. Membr. Biol. 96: 27-43, 1987. TOON, M. R., AND A. K. SOLOMON. Control of red cell urea and water permeability by sulfhydryl reagents. Biochim. Biophys. Acta 860: 361-375,1986. TYCKO, B., AND F. R. MAXFIELD. Rapid acidification of endocytic vesicles containing cy2 macroglobulin. Cell 28: 643-651, 1982. VERKMAN, A. S., W. I. LENCER, D. BROWN, AND D. A. AUSIELLO. Endosomes from kidney collecting tubule cells contain the vasopressin-sensitive water channel. Nature Lond. 333: 268-269, 1988. WANG, J. H., C. V. ROBINSON’ AND I. S. EDELMAN. Self-diffusion and structure of liquid water. III. Measurement of self-diffusion of liquid water with 2H7 “H, and “0 as tracers. J. Am. Chem. Sot. 75: 466-470, 1955. WHITTEMBURY, G., P. CARPI-MEDINA, E. GONZALEZ, AND H. LINARES. Effect of para-chloromercuribenzenesulfonic acid and temperature on cell water osmotic permeability of proximal straight tubules. Biochim. Biowhvs. Acta 775: 365-373, 1984.
Downloaded from www.physiology.org/journal/ajprenal at Glasgow Univ Lib (130.209.006.061) on February 12, 2019.
