Vol. 264,No. 19,Issue of July 5,pp. 11103-11106,1989
Printed in U.S.A.
OF BIOLOGICAL
CHEMISTRY
THEJOURNAL
0 1989 by The American Society for Biochemistry and Molecular Biology, Inc.
C1/HC03 Exchange in Human Placental Brush Border
Membrane Vesicles*
(Received for publication, January 31, 1989)
Steven M. Grass1
From the Department of Pharmacology, State University of New York Health Science Center, Syracuse, New York 13210
The human placenta performs an important function in
normal fetal development by serving as an interface for net
transfer of carbon dioxide from fetal to maternal blood supplies. Inthis capacity the placenta may also function to
* This work was supported in part by grants-in-aid from the Cystic
Fibrosis Foundation and the American Heart Association. The costs
of publication of this article were defrayed in part by the payment of
page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solelyto indicate
this fact.
isolated from human term placenta by divalent cation aggregation
and differential centrifugation as described previously (5). Briefly,
the villous tissue of placenta obtained within 15 min of elective
caesarean section was quickly dissected and minced into small (cm) fragments at 4 “C. The tissue fragments were rinsed three times
in 300 mM mannitol, 10 mM HEPES/TMA, pH 7, and gently stirred
for approximately 30 min using a Teflon spatula. The tissue suspension was filtered through cotton gauze, and phenylmethylsulfonyl
fluoride was added to a final concentration of 0.2 mM. The filtrate
The abbreviations used are: DIDS, 4,4’-diisothiocyanostilbene2,2‘-disulfonic acid; HEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; TMA, tetramethylammonium; MES,4-morpholineethanesulfonic acid; TAPS, N-tris[hydroxymethyl]metbyl-3-aminopropanesulfonic acid.
11103
Downloaded from www.jbc.org by on January 21, 2007
maintain normal fetal acid/base statuswhen confronted with
Membrane transportpathwaysfortransplacental
transfer of COz/HC03 were investigated by assessing changes in maternal blood pH or pCOZ levels. Both of these
the possible presence of a cl/HCos exchange mecha- processes require transfer of COz, as a gas or related species
nism in the maternal-facing membraneof human pla- (HCO;, HZCO& across the fetal-facing and/ormaternalcental epithelial cells. Cl/HC03 exchange was tested
facing membrane of syncytiotrophoblast cells. In its simplest
for in preparations
of purified brush border membraneform CO, transport across the syncytiotrophoblast may occur
vesicles by 36Cl tracer flux measurements and deter- entirely asthe highly permeable gaseous species moving down
minations of acridineorangefluorescence
changes. its fetal-maternal concentration gradient. Alternatively, gasUnder 10% COz/90%Nz the imposition of an outwardly eous CO, may enter the syncytiotrophoblast cell across its
directed HCO; concentration gradient (pH. 6/pHi 7.5) basal membrane where HCO; is formed for export by mestimulated C1- uptake to levels approximately 2-fold diated transferacross the maternal-facing brushborder memgreater than observed at equilibrium. Maneuvers designed to offset the development of ion gradient-in- brane. Intracellular conversion of CO, to HCO, may be facilduced diffusion potentials (valinomycin,KO= Ki) sig- itated by base equivalents generated from a Na/H exchange
nificantly reduced HCO; gradient-driven C1- uptake mechanism present in the brush border membrane (1).Albut concentrative accumulationof C1- persisted. Early though as yet unknown in humans, placental carbonic anhytime point determinations performed in the presumed drase activity has been measured in all species examined (rat,
hamster, guinea pig, sheep, pig), which may further suggest
absence of membrane potential suggests the reduced
the formation ofHCO; from CO, (2). Investigations ofC1level ofHCO;
gradient-driven C1- uptakeresulted
from a more rapid dissipation of the HCO; concentra- transport pathways in human placental brush border memtion gradient. Concentrative accumulation ofC1- was branes indicate the presence of a DIDS-sensitive C1- uptake
not observed in the presence of a pH gradient alone and a DIDS-insensitive voltage-dependent C1- uptake which
under 100% Nz,suggesting a preference of HCO; over suggests the existence of both an electroneutral anion exchanOH- as a substrate for transport. As monitored by ger and a Cl--conductive pathway (3, 4). These observations
acridine orange fluorescence the
C1- gradient-depend- indicate the possible presence of a syncytiotrophoblast anion
ent collapse of an imposed pH gradient (pH, 8.5/pHi 6) exchange mechanism at the maternal-facing membrane that
was accelerated in the presence of C0z/HC03 when may couple the favorable transmembrane C1- gradient to the
compared withits absence,indicating coupling of efflux of intracellular HCO;. The presence of a C1”conductive
HCO; influx to C1- efflux. Increasing concentrations pathway may serve to recycle C1- back across the brush border
of the anion exchange inhibitor 4,4’-diisothiocyano- membrane thus preventing an increase in intracellular C1-.
stilbene-Z,Z’-disulfonicacid were observed to cause a
In an effort to determine a possible role for ion-coupled
stepwise reduction in HCO; gradient-driven C1- uptake (Iso-25 p ~ further
)
suggesting the presence of a HCO, transport pathways in transplacental CO, transfer
and/or placental maintenance of fetal acid/base balance, the
cl/Hco3exchange mechanism. The resultsof this study presence
of a C1/HC03 exchange mechanism was assessed
provideevidence
for a 4,4’-diisothiocyanostilbene2,2’-disulfonic acid-sensitive Cl/HCOs exchange mech- using preparations of purified brush border membrane vesianism inthe maternal-facing membrane
of human pla- cles. Evidence supporting the existence of a C1/HCO3 excental epithelial cells. The identification of an ion- change mechanism was obtained from 36Cl traceruptake
studies and acridine orange fluorescence measurements.
coupled HCO; transport pathway in placental epithelia
may suggest functional roles in mediating transplacenEXPERIMENTALPROCEDURES
tal transfer of COz as well as maintenance of fetal acid/
base balance.
Membrane Preparations-Brush border membrane vesicles were
11104
C1/HC03Exchange in Human Placental Brush Border Membrane Vesicles
of ethanol were added to control aliquots of membrane. All solutions
were prepared with distilled-deionized water and passed through 0.22pm Millipore filters.
RESULTS AND DISCUSSION
HCO, Gradient-driven Cl- Influx-The presence of a C1/
HCO, exchange mechanism in human placentalbrush border
membrane wouldbe suggested by the ability of a HCO;
concentration gradient to serve as a driving force for intravesicular concentrative C1- accumulation. The time course of
intravesicular c1- accumulation (c1; = 4.15 mM) is illustrated
0
0
0'
0
'
'
15s 30s
60s
120s
"
Time
FIG. 1. HCOi gradient-driven C1- influx. Brush border membrane vesicles were pre-equilibrated under 10% CO*/90% NP with
(pH, 6/pHi 6 + COZ/HCO3) 110 mM TMA gluconate, 57.3 mM
potassium gluconate, 52 mM MES, 45.3 mM HEPES, 25 mM TMA
(pH, 6/pH; 7.5
COS/HCO3) 110 mM TMA
(OH-), 2 mMHCO;;
gluconate, 57.3 mM KHCO,, 52 mM mannitol, 45.3 mM HEPES, 23
mM TMA (OH-). Uptake of %C1(4.15mM) occurred from extravesicular solutions pre-equilibrated under 10% C02/90% N2 and containing (pH, 6/pH; 6 C02/HC03) 108 mM TMA gluconate, 57.3 mM
potassium gluconate, 52 mM MES, 44.3 mM HEPES, 26 mM TMA
(pH, 6/pHi 7.5
COJHCO,) 108 mM TMA
(OH-), 2 mMHCO;;
gluconate, 57.3 mM K+, 59 mM gluconate, 47 mM MES, 9 mM HEPES,
26mM TMA (OH-), 29 mM mannitol, 2 mMHCO;. A representative
experiment of three independent observations is illustrated.
+
+
+
Downloaded from www.jbc.org by on January 21, 2007
was centrifuged at 8,100 rpm for 15 min using an SS-34 rotor
(Sorvall). The low speed pellet was discarded and the supernatant
was centrifuged at 19,000 rprn for 40 min. The high speed pellet was
gently resuspended and MgCl, was added to a final concentration of
1 2 mM. After incubating for 10 min the membrane suspension was
centrifuged a t 5,000 rpm for 15 min to pellet the M$+-induced
aggregates. The low speed supernatant was centrifuged at 19,000 rpm
for 40 min and theresulting pellet (brush border membrane vesicles)
resuspended and washed twice in buffers designated for each experiment. Membranes were stored frozen (-70 "C) and used within 2
weeks of preparation. The isolated membrane vesicles were enriched
25.4 f 1.3(S.E., n = ?')-fold in alkaline phosphatase activity (6)
compared with homogenates of villous tissue. Membrane marker
enzyme enrichments for the basal membrane (Na/K-ATPase), mitochondria (succinic dehydrogenase), and endoplasmic reticulum
(NADH dehydrogenase) were 0.68 f 0.05 (S.E., n = 7), 0.43 f 0.02
(S.E., n = 7), and 0.34 f 0.03 (S.E., n = 7), respectively (7-9). Protein
was determined by a sodium dodecyl sulfate-Lowry assay using bovine
serum albumin as the standard (10).
Isotopic Flux Measurements-Frozen (-70 "C) aliquots of membrane vesicles werethawed at room temperature and isosmotic solutions of appropriate ionic composition were added to obtain the
desired intravesicular solution described for each experiment in the
figure legends. The membrane suspension was incubated for 90 min
at room temperature to attain transmembrane equilibration of the
added media. During the pre-equilibration period the membranes
were gassed continuously with humidified 100% Nz or 90% N2, 10%
COS.The extravesicular media wereprepared similarly, and thefinal
composition for each experiment is given in the figure legends.
Intravesicular 32Clcontent was assayed in triplicate at 37 "C in the
continued presence of either 100%NSor 90% N2, 10%COz bya rapid
filtration technique previously described (11).A metronome was used
to determine C1- uptake values a t 4 s or less (12). The uptake reaction
was quenched by the rapid addition of 210 mM potassium gluconate,
20 mM HEPES/TMA, pH 7.5, kept a t 4 "C. The diluted membrane
suspension was passed through a 0.65-pm Millipore filter (DAWP)
and washed with an additional 9 ml of the quench buffer. The process
of quenching, filtration, and washing occurred routinely within a 15s period. The filters were dissolved in3 mlof
Ready-Soh H P
(Beckman) and counted by scintillation spectroscopy. The timed
uptake values obtained were corrected for the nonspecific retention
of isotope by the filters. While absolute C1- uptake values expressed
per mg of membrane protein varied from membrane preparation to
membrane preparation, relative changes resulting from experimental
manipulations were highly reproducible.
Fluorescence Determinations of ApH-Changes in intravesicular
pH in response to an imposed pH gradient (pH, 8.5/pH; 6) and the
experimental conditions described were monitored by acridine orange
fluorescence using a SPEX DM 3000 fluorometer (13). Fluorescence
was measured a t excitation and emission wavelengths of 496 and 530
nm, respectively, with a bandpass of 3.6 nm. Equal aliquots of thawed
membranes washed into either 165 mM KCI, 10 mM MES/TMA, pH
6, or 165 mM potassium gluconate, 10 mM MES/TMA, pH 6, were
diluted with solutions of appropriate ionic composition to obtain the
desired intravesicular solution as described in the figure legends. The
membrane suspensions were incubated for 90 min at room temperature under continuous gassing with either humidified 100% NZ or
99% NP, 1%COS. The extravesicular or cuvette solutions were prepared similarly, and their ionic composition isgiven in the figure
legends. Experiments were initiated by the rapid addition of 10 r l of
membranes (30-60 pg) protein) to 2.5 ml of stirred cuvette solutions
thermostatically maintained at 25 "C. Upon addition of the vesicles
to the cuvette buffer an immediate "pH jump" or quench in fluorescence was noted in response to the imposed pH gradient. No significant difference in the magnitude of the pH jump was observed among
the various transmembrane ionic conditions tested, which indicates
the dye response and pH gradient were unaffected by manipulating
the intra-and extravesicular ionic composition. Time-dependent
changes in fluorescence occurring immediately after the pH jump
were determined in the absence and presence of COz/HCO3. For each
of the two conditions tested the fluorescenceresponse to anoutwardly
directed C1- gradient was corrected for changes in fluorescence measured in the absence ofC1- to obtain the C1- gradient-dependent
change in fluorescence.
Materials-Valinomycin, DIDS, and acridine orange were purchased from Sigma. 36Clwas obtained from Du Pont-New England
Nuclear. Valinomycin was dissolved in 95% ethanol and was added
to the membrane suspension in a 1:300 dilution. Equivalent volumes
pHo6/pH 7.5 + C 0 2 / H C 0 ,
"G /
.-c
B
PH,, WPH
7.5 +co2mco3
5
1
00
15s 30spHoG/pH
60s6 +CO,/HCO,
120s
90 rnin
Time
FIG. 2. Effect of valinomycin (VAL)on HCO; gradientdriven C1- influx. Brush border membrane vesicles were pre-equilibrated as described in the legend to Fig. 1. Uptake of %C1(4.15 mM)
occurred from extravesicular solutions described in the legend to Fig.
1. Where indicated (+ VAL) membrane vesicles were preincubated
with valinomycin (0.17 mg/ml) for a minimum of 30 min. A representative experiment of three independent observations is shown.
11105
Cl/HC03Exchange in HumanPlacental Brush BorderMembrane Vesicles
20
0
40
60
Time (seconds)
0'
0.0
1 .o
2.0
3.0
4.0
Time (seconds)
0
15s 30s
60s
120s
-/
90 min
Time
FIG. 4. OH- gradient-driven C1- influx. Brush border membrane vesicles were pre-equilibrated under 10% C02/90% N2 as described in the legend to Fig. l for (pH. 6/pHi 6 + COJHCOJ and
(pH. 6/pHi 7.5 + COZ/HCO3)or under 100% Nz with (pH, 6/pH, 7.5
- CO,/HCO,) 110 mM TMA gluconate, 57.3 mM potassium gluconate,
52 mM mannitol, 45.3 mM HEPES, 23 mM TMA (OH-). Uptake of
36Cl(4.15mM) occurred from extravesicular solutions under 10%
CO2/90% Nz as described in the legend to Fig. 1 for (pHo6/pHi 6 +
COz/HCO3) and (pH, 6/pHi 7.5 + COz/HC03) or under 100% N2:
(pH, 6/pHi 7.5 - C02/HC03) 108 mM TMA gluconate, 57.3 mM
potassium gluconate, 47 mM MES, 9 mM HEPES, 26 mM TMA
(OH-), 42 mM mannitol. Membrane vesicles were preincubated with
valinomycin (0.17 mg/ml) for a minimum of 30 min. A representative
experiment of three independent observations is shown.
in Fig. 1 as a function of an imposed transmembrane pH and
HCO; gradient. In theabsence of an imposed HCO; concentration gradient (pH, 6/pHi 6) C1- uptake was low and slowly
approached an equilibrium value at 1.5 h. The imposition of
an outwardly directed HCO, gradient (pH, 6/pHi 7.5) resulted
in a marked stimulation of C1- uptake accumulating to levels
approximately 2-fold greater than observed at equilibrium.
The concentrative accumulation of C1- noted in the presence
of an outwardly directed HCO; gradient suggests an anion
I
0
25
50
75 125 100
150200 175
[DIDSI M
FIG. 6. Effects of DIDS on HCOS gradient-driven C1- influx.
Brush border membrane vesicles were pre-equilibrated under 10%
COz/90% N2as described in the legend to Fig. 1 for (pH. 6/pHi 7.5 +
COz/HC03). The 15-5 uptake of %c1(4.15 mM) occurred from an
extravesicular solution described in the legend to Fig. 1 for (pH, 6/
pHi 7.5 + COJHCO,) and containing DIDS a t the concentrations
shown. Membranes were preincubated with valinomycin (0.17 mg/
ml) for a minimum of 30 min. The data are illustrated as the mean
& S.E. of three experiments, each performed using a different membrane preparation.
flux coupling consistent with the operation of a Cl/HCOJ
exchange mechanism. However, as the HCO, gradient-induced stimulation of C1- uptake may have resulted from
indirect electrical coupling to an ion gradient (H+, OH-,
HC0;)-induced inside positive diffusion potential, the nature
of HCO; coupling to C1- uptake was examined by determining
the effect of maneuvers designed to minimize membrane
potential development. To the extent that C1- uptake was
electrostatically coupled to ion gradient (H+, OH-, HC0;)induced diffusion potentials, a reduced level ofC1- uptake
Downloaded from www.jbc.org by on January 21, 2007
FIG. 3. Early time point determinations of HCOS gradieutdriven C1- influx. Brush border membrane vesicles werepre-equilibrated as described in the legend to Fig. 1. Uptake of 36Cl(4.15 mM)
occurred from extravesicular solutions described in the legend to Fig.
1. Where indicated (+ VAL) membrane vesicles were preincubated
with valinomycin (VAL) (0.17 mg/ml) for a minimum of 30 min. A
representative experiment of three independent observations is illustrated.
FIG. 5.C1- gradient-dependent intravesicular alkalinization. Brush border membranes were pre-equilibrated under 100% NZ
with 164 mM potassium gluconate, or 160 mMKC1, 4 mM potassium
gluconate, 10 mM MES/TMA (OH-), pH 6, and 6 p M acridine orange.
Membranes gassed with 1%COz/99% NZ were similarly pre-equilibrated. The extravesicular cuvette solutions were gassed in parallel
with membrane vesicles and consisted of 100% NZ- 164 mM potassium gluconate, 10 mM TAPS/TMA (OH-), pH 8.5, 6 p M acridine
orange; 1%CO2/99% Nz - 57.3 mM KHCO;, 106.7 mM potassium
gluconate, 10 mM TAPS/TMA (OH-), pH 8.5, 6 p M acridine orange.
Membranes were preincubated with valinomycin (0.17 mg/ml) for a
minimum of 30 min. The C1- gradient-dependent change in intravesicular alkalinization is illustrated as thedifference between fluorescence responses in the absence and presence ofC1-. The results
shown are representative of four experiments each performed using
a different membrane preparation.
11106
Cl/HC03Exchange in HumanPlacental Brush BorderMembrane Vesicles
imposed pH gradient (pH, 8.5/pHi 6) was determined in the
absence and presence of CO,/HCO3. To the extent thata C1/
HCO3 exchange mechanism was operative in these membrane
vesicles then theC1- gradient-dependent rate of intravesicular
alkalinization would be greater in the presence than absence
of C02/HC03. The marked stimulation ofC1- gradient-dependent alkalinization conferred by the presence ofCO,/
HC03 as shown in Fig. 5 suggests C1- efflux is coupled to
HCOT influx consistent with the presence of a C1/HC03
exchanger.
Effect of DIDS on HCO, Gradient-driven GI- Influx-Finally, the concentration-dependent inhibition of HCO; gradient-driven C1- uptake by the ClJHCO, exchange inhibitor
DIDS was assessed to verify further itsexistence in placental
brush border membrane (14). Shown in Fig. 6 are 15-s C1uptake values measured inthe presence of an outwardly
directed HCO; gradient and extravesicular DIDS concentrations ranging from 10 to 250 PM. As would be expected for
the presence of a C1/HC03 exchange mechanism HCO, gradient-driven Cl- uptake was decreased as a function of inhibitor concentration (I5,, -25 PM).
In conclusion, the existence of a DIDS-sensitive, C1/HCO3
exchange mechanism has been demonstrated in brush border
membrane vesicles isolated from the maternal surface of
human placental epithelial cells. The presence of this ioncoupled membrane transport pathway for HCO; suggests a
possible role in mediating transplacental transferof CO, from
fetus to mother. The identified HCOF transport pathway may
be postulated to serve an additional function in maintaining
fetal acid/base balance when the placenta is exposed to
changes in maternal blood pH and pCOz levels.
Acknowledgments-The excellent technical assistance of Chris
Oginand Stacey Hulbert is gratefullyacknowledge as well as the
excellent secretarial assistance of Dorothy Stechyshyn.
REFERENCES
1. Balkovetz, D. F., Leibach, F. H., Mahesh, V. B., Devoe, L. D.,
Cragoe, E. J., and Ganapathy, V. (1986) Am. J. Physiol. 2 5 1 ,
C852-C860
2. Lutwak-Mann, C. (1955) J. Endocrinol. 1 3 , 26-38
3. Shennan, D. B., Davis, B., and Boyd, C. A. R. (1986) Pfluegers
Arch. Eur. J. Physiol. 406, 60-64
4. Illsley, N.P., and Verkman, A. S. (1987) Biochemistry 2 6 , 12151219
5. Ogin, C., and Grassl, S. M. (1989) Biochim. Biophys. Acta 9 8 0 ,
248-254
6. Bowers, G. N., and McComb, R. B. (1966) Clin.Chem. 12, 7089
7. Jbrgensen, P. L. (1968) Biochim. Biophys. Acta 1 5 1 , 212-224
8. Earl, D. C. N., and Korner, A. (1965) Biochem. J. 9 4 , 721-734
9. Wallach, D. F. H., and Kamat, V. B. (1966) Methods Enzymol.
8, 165-166
10. Peterson, G. L. (1983) Methods Enzymol. 91,95-115
11. Grassl, S. M., Holohan, P. D., and Ross, C. R. (1987) J. Biol.
Chem. 262,2682-2687
12. Wright, S. H., Kippen, I., and Wright, E. M. (1982) J. Biol. Chem.
2 5 7 , 1773-1778
13. Schuldiner, S., Rottenberg, H., and Avron, M. (1971) Eur. J.
Biochem. 25,64-70
14. Lowe, A. G., and Lambert, A. (1983) Biochim. Biophys. Acta6 9 4 ,
353-374
Downloaded from www.jbc.org by on January 21, 2007
would be expected in the presence of charge compensating
movements of K+ across valinomycin-treated membranes.
Although, as shown in Fig. 2, the stimulation of C1- uptake
measured in the presence of an outwardly directed HCO;
gradient was reduced in membranes incubated with valinomycin, the concentrative accumulation of C1- persisted. While
these results may indicate the presence of a Cl--conductive
pathway also identified in previous studies of placental brush
border membranes (3, 4), C1- uptake was observed to exceed
equilibrium levels when measured in thepresumed absence of
membrane potential differences. This finding suggests a direct
chemical coupling ofHCO; efflux to C1- influx consistent
with the existence of a Cl/HC03exchange mechanism.
The effect of short-circuiting membrane potential development to reduce HCO, gradient-driven C1- uptake may have
also occurred in part as a result of promoting a more rapid
dissipation of the imposed HCOF concentration gradient. To
test this possibility measurements of HCO, gradient-driven
C1- uptake were performed at early time points in vesicles
treated with and without valinomycin. As shown in Fig. 3 C1uptake was essentially identical in the presence and absence
of valinomycin at 1s but thereafter was progressively reduced
in short-circuited membranes. These data suggest that after
only 1 s the imposed HCO; gradient had not dissipated to
levels where differences in C1- uptake may be distinguished
between membranes treated with and without valinomycin.
The similarity of 1-s C1- uptake values in the presence and
absence of valinomycin would not be expected if the effect of
blunting membrane potential development was entirely due
to inhibition of conductive C1- uptake. We submit thisobservation as further evidence for the presence of Cl/HCO3 exchange mechanism by suggesting the decreased C1- uptake
measured in short-circuited membrane vesicles resulted a t
least partly from a more rapid decay of the driving force for
exchange.
In an effort to distinguish between HCOF and OH- as the
preferred driving force for exchange with C1-, measurements
ofC1- uptake by valinomycin-pretreated membranes were
performed in the presence of the same pH gradient (pH, 6/
pHi 7.5) gassed with and without CO,. As illustrated in Fig. 4
the imposition of an inside alkaline pH gradient in the nominal absence of COz induced only a small increase in C1uptake compared with control values determined in the absence of a pHgradient (pH, 6/pHi 6). Notably, a concentrative
accumulation ofC1- was only observed in the presence of a
pH gradient and C02/HC03 which strongly suggests HCO;
as thepreferred anion for exchange with C1-.
Fluorescence Studies of C1- Gradient-driven HCO; InfluxThe presence of a placental brush
border membrane C1/HC03
exchange mechanism was investigated further by examining
the reciprocal role relationship between C1- and HCO; as the
driving force for anion exchange. Not only should gradients
of HCO; drive C1- uptake as previously shown but a C1gradient-dependent accumulation of intravesicular HCO;
should also be observed to indicate the presence of Cl/HC03
exchange. Using the fluorescent pH probe acridine orange,
the rate of intravesicular alkalinization in response to an