Functional and structural demonstration of the presence of Ca-ATPase (PMCA) in both microvillous and basal plasma membranes from syncytiotrophoblast of human term placenta.
ABSTRACT It is known that human syncytiotrophoblast (hSCT) actively transports more than 80% of the Ca2+ that goes from maternal to fetal circulation. Transepithelial transport of Ca2+ is carried out through channels, transporters and exchangers located in both microvillous (MVM) and basal (BM) plasma membranes. The plasma membrane Ca-ATPase (PMCA) is the most important mechanism of Ca2+ homeostasis control in the human placenta. In this work, we reexamined the distribution of PMCA in isolated hSCT of term placenta. The PMCA activity was determined in isolated hSCT plasma membranes. A partial characterization of the PMCA activity was performed, including an evaluation of the sensitivity of this enzyme to an in vitro induced lipid peroxidation. Expression of the PMCA in hSCT plasma membranes and tissue sections was investigated using Western blots and immunohistochemistry, respectively. Our study demonstrates, for the first time, a correlation between the activity and structural distribution of PMCA in both MVM and BM of hSCT. It also demonstrates a higher PMCA activity and expression in MVM as compared to BM. Finally, PMCA4 seems to be preferentially distributed in both hSCT plasma membranes, while PMCA1 is shown to be present in the hSCT homogenate. However, the membrane fractions did not show any PMCA1 labeling. Our results must be taken into account in order to propose a new model for the transport of calcium across the hSCT.
- Nature 07/1995; 375(6533):634-5. · 38.60 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The plasma membrane calcium pump is the system that ejects Ca2+ out of eukaryotic cells: this is documented in all animal and plant cells, although knowledge on the latter is only now beginning to be established. Information on lower eukaryotic cells, e.g., yeast, is still scarce, but it also is beginning to develop. The pump shares the catalytic properties of ion-motive ATPases of the P-type family, but has distinctive regulation properties: it is modulated by calmodulin, acidic phospholipids, a number of protein kinases, possibly by the interaction of calcium with its COOH-terminal region, and by aggregation (dimerization) through the calmodulin binding domain. The latter acts as an endogenous inhibitor of pump activity, much as phospholamban does for the sarcoplasmic reticulum pump. The analogy of the regulation mechanisms of the two pumps is heightened by the finding that phosphorylation of the calmodulin binding domain by protein kinase C removes its autoinhibiting function, as other kinases do in the case of phospholamban. The pump is the product of a family of four genes located on different human chromosomes. The isoform diversity is dramatically enhanced by alternative splicing of the transcripts, occurring at "hot spot" A (NH2-terminal) and C (COOH-terminal). At present more than 20 different transcripts with striking tissue and developmental specificity have been detected.The FASEB Journal 11/1994; 8(13):993-1002. · 5.70 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The plasma membrane Ca2+ pump is responsible for the fine regulation of the intracellular Ca2+ level and is thus involved in the control of several cellular processes. The activity of the pump is regulated by a multiplicity of mechanisms, among which are calmodulin, acidic phospholipids, kinase-mediated phosphorylation, or an oligomerization process. The C-terminal part of the molecule interacts with the region of the pump close to the active site, leading to the decrease of the activity in the resting state. Four genes coding for different isoforms of the plasma membrane Ca2+ ATPase are known in humans. Isoform 1 and 4 represent housekeeping isoforms, whereas isoforms 2 and 3 are only present in specialized tissues. The variability of the protein is further increased by alternative RNA splicing at two sites (A, C). Alternative splicing occurs within (splice site C) or near (splice site A) regions coding for regulatory domains of the protein. In all isoforms a corresponding splice form exists at both splice sites. These common splice forms are present in all tissues, whereas isoform unique splice forms are normally only present in specialized tissues. In neuronal tissues all isoforms and almost the complete set of splice forms are found. The transcripts of the different isoforms are distributed in a region-specific manner in neuronal tissues.Journal of Neurobiology 04/1994; 25(3):312-24. · 3.05 Impact Factor
Functional and Structural Demonstration of the Presence of Ca-ATPase
(PMCA) in Both Microvillous and Basal Plasma Membranes from
Syncytiotrophoblast of Human Term Placenta
R. Marı ´na,*,1, G. Riquelmeb,1, V. Godoyb, P. Dı ´azb, C. Abada, R. Cairesa,
T. Proverbioa, S. Pi~ neroa, F. Proverbioa,1
aLaboratorio de Bioenerge ´tica Celular, Centro de Biofı ´sica y Bioquı ´mica, Instituto Venezolano de Investigaciones
Cientı ´ficas (IVIC), A.P. 21827, Caracas 1020A, Venezuela
bLaboratorio de Electrofisiologı ´a de Membranas, Fisiologı ´a y Biofı ´sica, Instituto de Ciencias Biome ´dicas (ICBM),
Facultad de Medicina, Universidad de Chile, Casilla 70005, Santiago 7, Chile
It is known that human syncytiotrophoblast (hSCT) actively transports more than 80% of the Ca2þthat goes from maternal to fetal circu-
lation. Transepithelial transport of Ca2þis carried out through channels, transporters and exchangers located in both microvillous (MVM) and
basal (BM) plasma membranes. The plasma membrane Ca-ATPase (PMCA) is the most important mechanism of Ca2þhomeostasis control in
the human placenta. In this work, we reexamined the distribution of PMCA in isolated hSCT of term placenta. The PMCA activity was deter-
mined in isolated hSCT plasma membranes. A partial characterization of the PMCA activity was performed, including an evaluation of the sen-
sitivity of this enzyme to an in vitro induced lipid peroxidation. Expression of the PMCA in hSCT plasma membranes and tissue sections was
investigated using Western blots and immunohistochemistry, respectively. Our study demonstrates, for the first time, a correlation between the
activity and structural distribution of PMCA in both MVM and BM of hSCT. It also demonstrates a higher PMCA activity and expression in
MVM as compared to BM. Finally, PMCA4 seems to be preferentially distributed in both hSCT plasma membranes, while PMCA1 is shown to
be present in the hSCT homogenate. However, the membrane fractions did not show any PMCA1 labeling. Our results must be taken into ac-
count in order to propose a new model for the transport of calcium across the hSCT.
Keywords: PMCA; Placenta; Syncytiotrophoblast; Microvillous plasma membranes; Basal plasma membranes
It is known that Ca2þplays a key role in intracellular sig-
naling in many cells. As a consequence, the intracellular
Ca2þconcentration should be tightly regulated . This con-
trol is exerted through various mechanisms, among which
the plasma membrane Ca-ATPase (PMCA) plays an important
role in the maintenance of low free Ca2þconcentration in the
cytoplasm, by extruding cytosolic Ca2þfrom the cells .
This enzyme is known to be coded by four separate genes
(PMCA 1e4), and due to alternative splicing, there are at least
20 variants designated by lowercase letters following the iso-
form number (e.g., PMCA1a, PMCA1b, etc.) [3,4]. PMCA 2
and 3 isoforms are found in specialized tissues while isoforms
1 and 4 are present in almost all tissues including placenta .
However, there is a high variability in the patterns of expres-
sion for the different splice variants of the PMCA isoforms.
For instance, in placental tissues only PMCA4a, PMCA4b
and PMCA1aare present . The different PMCA isoforms
differ primarily in their regulatory regions and the modulation
* Corresponding author. Tel.: þ58 2 504 1395; fax: þ58 2 504 1093.
E-mail addresses: firstname.lastname@example.org, email@example.com (R. Marı ´n).
1R. Marı ´n, F. Proverbio and G. Riquelme are co-senior authors of this
of their activities strongly depends on the membrane
The human placental syncytiotrophoblast (hSCT) is a polar-
and fetal circulations. This epithelium has a maternal-facing
microvillous membrane (MVM) and a fetal-facing basal mem-
brane (BM). Thus, any transport of ions and solutes to and
from the fetal compartment must be performed across both
blast actively transport more than 80% of the Ca2þthat goes
from maternal to fetal circulation . Transepithelial transport
of Ca2þis carried out by channels, transporters and exchangers
located in both MVM and BM .
Placental Ca2þtransport has been reported to be inhibited
by the addition of erythrosin B, a placental PMCA inhibitor
. In addition, Ca2þtransport in perfusion studies of
placenta, is minimally affected when the Naþof the incuba-
tion medium is substituted with choline, a clear indication
that the Naþ/Ca2þexchanger (NCX) has a minimal role in
the transplacental movement of Ca2þfrom mother to fetus
. Accordingly, it has been proposed that PMCA plays an
important role in Ca2þhomeostasis in the human placenta
. One of the very first studies looking for PMCA in
hSCT, considered the fact of a higher concentration of cal-
cium in the fetal circulation as compared to the maternal
one, suggesting a predominant presence of such an extrusion
mechanism in BM . This consideration led these authors
to demonstrate the presence of PMCA in membrane vesicles
of BM from hSCT, without paying attention to MVM. In an-
other set of studies, Borke et al.  and Strid and Powell
, by using immunohistochemistry techniques in human
placental tissue and with a monoclonal anti-PMCA antibody
5F10, found an apparent exclusive distribution of PMCA in
BM. In addition, Borke et al. , determined from the West-
ern blot analysis of MVM preparations that the antibody 5F10
binds poorly to the region around 140 kDa. These specific
findings have been used, since then, to conclude that the
PMCA, similarly to renal and intestinal epithelia, is located
exclusively in BM from hSCT . However, in preliminary
experiments we have found: (a) the presence of a Mg2þ-
dependent, thapsigargin-insensitive, Ca2þ-stimulated-ATPase
activity in MVM fractions isolated from human term placenta
; and (b) PMCA expression in MVM fractions of placen-
tal membranes by Western blot and immunohistochemistry
with specific antibodies [16,17]. These findings, together
with the fact that the PMCA and the Na,K-ATPase (NKA)
are colocalized in the basolateral plasma membranes of renal
and intestinal epithelia, while the NKA is present in both BM
and MVM from hSCT , led us to reexamine the distribu-
tion of PMCA in hSCT of term placenta. In this study, we
tested the hypothesis that PMCA is present in both MVM
and BM from hSCT. The biochemical expression of the
PMCA activity was assayed and partially characterized in
isolated preparations from both MVM and BM. Expression
of the PMCA in isolated hSCT plasma membranes and tissue
sections was also investigated using Western blots and immu-
2. Materials and methods
2.1. Placenta collection
Placentae obtained from normal pregnancies were collected immediately
after delivery from the Maternity Hospital ‘‘Concepcio ´n Palacios’’ in Caracas,
Venezuela and San Jose ´ Hospital Maternity Unit (monitored Unit Outpatient
clinic, healthy pregnancies) in Santiago, Chile and transported to the respec-
tive laboratories on ice. Any woman that, according to her medical history,
was under medical treatment to control blood pressure, or if she was taking
>1 g of elemental calcium per day during pregnancy, or if she had a history
of hypertension, diabetes, calcium metabolism disorders, or any other chronic
medical illness, was not considered for this study.
2.2. Preparation of syncytiotrophoblast plasma membranes
The human placental plasma membranes (MVM) and basal membranes
(BM) were prepared from fresh placentae following a previously described
method [19,20]. In brief: the maternal decidua was removed, and the central
portion between the maternal and fetal surfaces was used for the preparation.
Placental villous tissue (80e100 g) was chopped into small pieces, washed
with 0.9% NaCl to remove blood and filtered through gauze. The purification
method involved different steps: differential centrifugation, precipitation of
non-microvillous membranes with magnesium ions and a sucrose gradient
step. All solutions were buffered with 20 mM Tris-maleate, pH 7.4. Sucrose
gradient preparation: a portion (2e3 ml) of the microvillous-enriched prepa-
ration and the basal membrane-enriched preparation were overlayed on the
sucrose gradient. The band at the sucrose interface concentrations 37/45%
(w/v) corresponds to the apical fraction (MVM) and the band at the sucrose
interface concentrations 47/52% (w/v) corresponds to the basal fraction
(BM). The two fractions were collected, diluted 10-fold with the buffer
20 mM Tris-maleate, pH 7.4, and centrifuged at 110,000 ? g for 30 min.
The final pellets were resuspended in 300 mM sucrose, 20 mM Tris-maleate,
pH 7.4, and stored in liquid nitrogen or at ?50?C (freezer). The purity and
enrichment of the membrane fractions were determined routinely by assaying
alkaline phosphatase, an apical membrane marker; adenylate cyclase/
b-adrenergic receptor (by measuring
membrane markers; cytochrome c oxidase/succinate dehydrogenase, mito-
chondrial membrane markers and glucose-6-phosphatase, endoplasmic retic-
ulum marker [12,20].
3H-dihydroalprenolol binding), basal
2.3. PMCA activity
The PMCA activity was determined by measuring the quantity of inor-
ganic phosphate liberated from the hydrolysis of ATP, according to a modi-
fication of the method described elsewhere . Briefly, 180 ml of the
incubation medium were preincubated for 2 min at 37?C, and the reaction
was started by addition of 20 ml of membrane suspension. After 10 min
incubation, the reaction was stopped by addition of 300 ml of a cold solution
containing: 2.85% ascorbic acid; 1.76% HCl; 0.48% ammonium molybdate;
and 2.85% SDS. The samples were shaken and kept at 0?C for 10 min.
Then, 500 ml of 2% sodium citrate, 2% sodium arsenite and 2% glacial ace-
tic acid solution were added to each tube, which were then rewarmed, after
shaking, for 10 min at 37?C. The absorbance of each tube was determined
in a Milton Roy spectrophotometer at 705 nm. The ATPase activity is
expressed as nmol Pi/mg protein min, after subtraction of a blank run in par-
allel under the same conditions except for the membrane suspension, which
was added only after the addition of the ascorbic acid solution. The protein
concentration, in all the cases, was determined according to the method of
Bradford . The PMCA activity was calculated as the difference in the
phosphate liberated in a medium containing 250 mM sucrose; 5 mM ATP;
5 mM MgCl2; 1 mM ouabain; 2 mM EGTA; 2 mM EDTA; 30 mM Trise
HCl (pH 7.2 at 37?C); 55 mM KCl, 2 mg/ml calmodulin; 1 mM thapsigargin
and 2 mM free calcium, minus the one liberated in the same medium, but in
the absence of calcium. Purified bovine brain calmodulin was generously
supplied by Dr. Gustavo Benaim from the Universidad Central de
2.4. SDS pretreatment of the syncytiotrophoblast
In order to avoid the presence of membrane vesicles, the membrane frac-
tions were always pretreated before the assays with SDS, as previously
described . A 250 ml aliquot of the fraction (0.4 mg protein/ml) was pre-
treated with up to 8 ml of a solution containing 6.25 mg SDS/ml, 1% BSA,
25 mM Imidazole pH 7.2 at 37?C. The fractions were incubated for 20 min
at 37?C, and then immediately assayed for ATPase activity.
2.5. Ultraviolet irradiation of the syncytiotrophoblast
A 750 ml aliquot of the syncytiotrophoblast plasma membranes (1 mg
protein/ml) from normal pregnant women was poured into a glass vial, placed
on ice, and illuminated from approximately 4 cm distance by a mineral light
(wavelength 254 nm maximum, specified strength 280 mW/cm2at 15 cm dis-
tance) for different periods of time to induce lipid peroxidation . The UV
irradiation of the hSCT plasma membranes was carried out in the presence and
absence of 50 mM butylated hydroxytoluene (BHT).
2.6. Lipid peroxidation measurements
The amount of lipid peroxidation of the plasma membranes was estimated
by measuring the thiobarbituric acid-reactive substances (TBARS). The
TBARS were determined according to the method described by Feix et al.
. The absorbance was measured at 532 nm and the TBARS values were
calculated by using a malondialdehyde standard curve, prepared by acid
hydrolysis of 1,1,3,3-tetramethoxypropane. The values are expressed as
nmoles of malondialdehyde per milligram of protein.
Human placental tissue samples (0.5 cm3) from normal pregnancies were
placed in 0.9% NaCl and fixed in 3.7% buffered formaline at pH 7.4, for
24 h minimum. Subsequently, the tissue was rinsed five times in ice-cold phos-
phate buffered saline (PBS) and dehydrated through a graded series of ethanol
to xylene, embedded in paraffin, and cut into 5 mm-thick sections. Afterwards,
paraffin was removed in xylene and the sections were rehydrated by passage
through graded ethanol and, finally, distilled water. The sections were blocked
for 1 h at room temperature with 4% bovine serum albumin (BSA) in PBS.
Tissue was then incubated for 2 h at room temperature with a monoclonal
antibody raised against all isoforms of PMCA (clone 5F10) diluted 1:1000,
a monoclonal antibody against PMCA4 isoform diluted 1:1000 (clone JA9);
a rabbit polyclonal antibody against PMCA1 isoform diluted 1:500 (all from
Affinity Bioreagents, CO) and a mouse monoclonal antibody against human
placental alkaline phosphatase (anti-PLAP, clone 8B6; Sigma-Aldrich, Inc.)
in 2% BSA in PBS. Negative control sections were treated similarly, except
that primary antibodies were omitted. After rinsing the samples with PBS, tis-
sue sections were incubated for 1 h at room temperature with biotin conjugated
goat anti-mouse and biotin conjugated goat anti-rabbit antibodies, for mono-
clonal and polyclonal primary antibodies used respectively (Vector Laborato-
ries, Burlingame, CA). The secondary antibody was rinsed off with three
changes of PBS, and Horseradish peroxidase-streptavidin reagent was added
for 30 min at room temperature. Slides were finally treated with NOVAREDO`
(Vector Laboratories, Burlingame, CA) to visualize the antigen stain. After the
appearance of a red reaction product, slides were washed in PBS, dehydrated
in graded ethanol, cleared in xylene, and mounted in ENTELLAN?(Merck).
Before dehydration, the slides were counterstained using hematoxylin.
2.8. Electrophoresis, Western blotting and
Aliquots of homogenate (H), MVM and BM preparations (20 mg total pro-
tein) were separated on a 7.5% SDS-PAGE gel. Electrophoresis was performed
at 100 V and the gel was transferred to a nitrocellulose membrane (BioRad
162-0115) for 2 h at 100 V. The nitrocellulose membrane was blocked for
2 h at room temperature with 3% non-fat milk in saline buffer-Tween
(138 mM NaCl, 270 mM KCl and 0.05% Tween 20), and washed in saline
buffer-Tween. The membranes were incubated with a primary antibody for
2 h at room temperature, i.e. anti-PMCA monoclonal antibody (clone 5F10,
diluted 1:1000), anti-PMCA4 monoclonal antibody (clone JA9, diluted
1:1000) and rabbit polyclonal antibody against PMCA1 isoform, diluted
1:500 (all from Affinity Bioreagents, CO). After washing with saline buffer-
Tween, the membranes were incubated with the specific secondary antibody;
anti-mouse horseradish peroxidase-linked antibody (1:10,000), and incubated
for 1 h at room temperature. The final detection was done using the enhanced
chemiluminiscence Western blotting analysis system (ECL, Amersham, RPN
2106). Protein content was quantified using UN-SCAN-IT gel Automated Dig-
itizing System, version 4.1 (Silk Scientific Corporation).
2.9. Statistical analysis
Comparisons between the different conditions were assessed by one-way
ANOVA followed by post hoc analysis with the Student-NewmaneKeuls
test. All results are expressed as means ? SE and (n) represents the number
of experiments performed with different preparations. The PMCA activity
was calculated from paired data. A p-value ?0.05 was accepted as statistically
3.1. Purity of MVM and BM fractions
The plasma membrane preparations of hSCTwere carefully
evaluated for membrane purity and cross contamination.
Table 1 shows the enrichment factors, calculated as the ratio
of activity in membrane fractions to that in the homogenate
(H), of the different membrane markers. The MVM fraction
(PLAP) of around 21-fold, and low contamination with BM,
as indicated by the enrichment factor of the BM markers
(adenylate cyclase activity and3H-dihydroalprenolol binding).
The BM fraction also showed an important enrichment, as
judged by adenylate cyclase activity (around 12-fold when
compared to the activity in the microsomal fraction) and
3H-dihydroalprenolol binding (around 13-fold). Cross contam-
ination of the BM fraction with MVM was low, as indicated
by the 2-fold enrichment of PLAP in the BM fraction. The
ratio between the adenylate cyclase activity of MVM and
BM fractions (MVM/BM) was 0.09, indicating a very low
of alkaline phosphatase
Comparison of the enrichment factor of several membrane markers in basal
(BM) and microvillous (MVM) plasma membrane fractions from human
2.31 ? 0.37
13.20 ? 0.61
12.15 ? 0.20
1.63 ? 0.30
0.23 ? 0.02
21.90 ? 3.65
2.23 ? 0.22
0.71 ? 0.11
2.80 ? 0.44
1.09 ? 0.07
The enrichment factors were calculated as the ratio of activity in membrane
fractions to that in the homogenate (H). Values are means ? SE of eight deter-
minations carried out with different preparations. *The enrichment factor was
calculated versus the activity of the microsomal fraction from our
contamination of MVM by BM. Similarly, the same ratio, cal-
culated for the
low: 0.20 (data not shown). Both BM and MVM fractions
showed a poor contamination with mitochondrial membranes
and endoplasmic reticulum. The purity and cross contamina-
tion of the purified membranes were comparable to those
reported for single or paired apical and basal membrane
3H-dihydroalprenolol binding was also very
3.2. PMCA activity of MVM and BM fractions
Both membrane preparations (MVM and BM) were
assayed for PMCA activity. The biochemical criteria used to
identify the PMCA activity included the determination of an
Mg2þ-dependent ATPase activity, stimulated by free Ca2þin
the micromolar range and by calmodulin, a well-known mod-
ulator of this ATPase activity . PMCA activity is inhibited
by 30 mM vanadate (a potent inhibitor of P-type ATPases
such as PMCA ) and is insensitive to 2 mM ouabain
(Na,K-ATPase inhibitor) and 1 mM thapsigargin (SERCA
inhibitor). Before the ATPase assays, the membrane prepara-
tions were pretreated with SDS in order to avoid the presence
of membrane vesicles . The results of the pretreatment
with SDS of both BM and MVM on the PMCA activity are
shown in Fig. 1. For both membrane preparations, the maximal
ATPase activity was reached when the membranes were pre-
treated with 0.25 mg SDS/mg protein. Consequently, before
any PMCA assay, the membranes were always pretreated
with this SDS/protein ratio.
Table 2 shows the PMCA activity in H, MVM and BM frac-
tions. It can be seen that the PMCA activity is not only present
in both MVM and BM, but also that the activity in MVM is
about twice as much as that in BM.
3.3. Partial characterization of PMCA activity in hSCT
The PMCA activity in MVM and BM preparations was fur-
ther analyzed. A partial characterization of the activity of both
enzymes is presented in Table 3. The different parameters
analyzed were almost identical for both membrane prepara-
tions, which showed similar Kmfor free Ca2þstimulation,
similar calmodulin modulation, similar optimal pH and tem-
perature values of the assay medium, and insensitivity to
2 mM ouabain and 1 mM thapsigargin. In addition, we tested
the effect of 30 mM vanadate, a P-type ATPase inhibitor, on
the PMCA activity of both MVM and BM. At the indicated
concentration, vanadate produced almost complete inhibition
of the PMCA in both membrane fractions (Table 3), indicating
that these ATPase activities correspond to the P-type ATPases.
Considering the fact that the PMCA activity of human red
cell ghosts is quite sensitive to membrane lipid peroxidation
, we utilized this parameter to further characterize the
PMCA activity in MVM and BM. Lipid peroxidation was
induced by UV light irradiation at 4?C for different lengths
of time, and determination of TBARS was used as an indica-
tion of the level of lipid peroxidation of the membranes .
The treatment was carried out in the presence and absence of
the antioxidant butylated hydroxytoluene (BHT). As shown in
Fig. 2, for both preparations, the TBARS of the plasma mem-
branes increased proportionally as a function of the UV light
treatment time, while the PMCA activity decreased propor-
tionally to this increment. When the UV irradiation of the
hSCT plasma membranes of both samples was carried out in
the presence of 50 mM BHT, the two measured parameters
PMCA activity and level of lipid peroxidation, remained unaf-
fected (data not shown). This is a clear indication that the UV
0.00.10.2 0.3 0.4 0.5
nmoles Pi/mg prot min
0.00.10.2 0.3 0.40.5
nmoles Pi/mg prot min
g SDS/ g protein
g SDS/ g protein
Fig. 1. Effect of the treatment of BM (panel A) and MVM (panel B) from
hSCT with SDS on the PMCA activity. Values are expressed as means ? SE
of six determinations carried out with different preparations.
PMCA activity of the microvillous (MVM) and basal (BM) plasma membrane
fractions of human term placenta
FractionATPase activity: nmoles Pi/mg prot min
5 ? 2
51 ? 5
92 ? 4
Values are means ? SE of five determinations carried out with different
treatment of the plasma membranes, per se, does not affect the
structure of membrane proteins, such as the PMCA. Data from
Fig. 2 were utilized to perform a regression analysis of the
PMCA activity as a function of the TBARS, for both MVM
and BM. This plot is shown in Fig. 3. Although the sensitivity
of the PMCA activity to lipid peroxidation is different for both
membranes, as indicated by the different slopes of the regres-
sion lines, it is evident that in both cases, the PMCA activity is
quite sensitive to the level of lipid peroxidation of the plasma
membrane (coefficient of determination, r2¼ 0.924 and 0.945
for BM and MVM, respectively).
3.4. Expression of PMCA in plasma membranes from
TheWesternblotanalysisofthe membranefractions was car-
ried out with a monoclonal antibody against all isoforms of
PMCA (clone 5F10), a monoclonal against PMCA4 (clone
JA9) and a polyclonal antibody against PMCA1 isoform. The
antibody 5F10 detects the four PMCA isoforms, including their
respective splice variants, because it recognizes an epitope that
is common to all known human PMCA gene products and splic-
ing variants . Western blot analyses were performed in sam-
plesofpairedH,MVM and BM purifiedfractions froma total of
a main band around 140 kDa, which is in agreement with the
molecular weight of the PMCA . The second band, which
co-migrated with the 140 kDa band, has been previously seen
in red cell ghosts and BM from hSCT preparations [28e30],
and it has been suggested to be a splice variant of the active
PMCA. As shown by the relative density, the expression of
responds to less than 20% of the sum of apical and basal PMCA
densitometrically quantified bands.Comparable results wereob-
as shown in Fig. 4B. The Western blot analyses confirm our data
in both MVM and BM, with higher ATPase activity in MVM.
Additional experiments were performed in order to try to
identify the presence of the PMCA1 isoform in the hSCT
plasma membranes. When a rabbit polyclonal antibody against
PMCA1 isoform was used, there was a strong mark in H
(Fig. 4C); however, none of the purified fractions showed
a specific band for PMCA1.
Partial characterization of the PMCA activity of microvillous (MVM) and
basal (BM) plasma membrane fractions of human term placenta
Kmfor free Ca2þ(mmol/l)
% inhibition by vanadate (30 mM)
Ouabain sensitivity (2 mM)
Thapsigargin sensitivity (1 mM)
0.22 ? 0.03*
95 ? 2*
0.25 ? 0.02*
97 ? 3*
*Values are means ? SE of six determinations carried out with different
0 10 2030 405060
nmoles Pi/mg protein min
UV irradiation time, min
010 2030 405060
nmoles MDA/mg protein
UV irradiation time, min
Fig. 2. PMCA activity (panel A) and TBARS (panel B) as a function of the
irradiation time with UV light (254 nm) of both BM (filled symbols) and
MVM (empty symbols) from hSCT. Values are expressed as means ? SE of
six determinations carried out with different preparations.
TBARS (nmoles MDA/mg prot)
nmoles Pi/mg prot min
BM (r = 0.961, r2 = 0.924)
MVM (r = 0.972, r2 = 0.945)
Fig. 3. PMCA activity as a function of the level of lipid peroxidation (TBARS)
of BM and MVM from hSCT. Data were taken from the mean values of Fig. 2.
3.5. Immunohistochemistry of placental villous tissue
Further experiments were performed in order to try to iden-
tify the location of PMCA in placental villous tissue. The
tissue sections (5 mm) were counterstained with hematoxylin
to mark the syncytiotrophoblast nucleus. To control the
antigenicity and intactness of the placental epithelium, the
presence of PLAP, an epithelial apical membrane marker,
was confirmed. Fig. 5B shows an immunohistochemical stain-
ing of placental villous tissue section from a normal preg-
nancy, with anti-human PLAP (clone 8B6; Sigma-Aldrich,
Inc). It can be observed that this marker is immunolocalized
in the maternal-facing side of the hSCT, which is in agreement
with previous reports [19,20], and with our biochemical results
(Table 1). Subsequently, monoclonal antibodies against all
PMCA isoforms (clone 5F10), against PMCA4 isoform (clone
JA9) and against PMCA1 isoform were used in paraffin-
embedded sections of human placental villous tissue. As
shown in Fig. 5C,D, PMCA staining for all isoforms was pre-
dominantly seen in hSCT, with partial expression in the fetal
specific staining for PMCA of the fetal-facing and maternal-
facing sides, including a trophoblast mark. Similar distribution
for PMCA4 isoform was also observed (Fig. 5E,F) but appar-
ently weaker in the cytoplasm. In contrast, the staining of
PMCA1 isoform (Fig. 5H), seems to be mainly in the tropho-
blast in all the studied samples. The latter observations are in
agreement with the Western blot results mentioned above
(Fig. 4C), where there was only evidence of the PMCA1
isoform in the placental homogenate and no staining was
obtained in the purified hSCT plasma membranes. Negative
control sections, where primary antibodies were omitted,
showed no specific staining for the different monoclonal and
polyclonal antibodies used (Fig. 5A,G, respectively). Similar
results were obtained in four placentae.
This study describes, for the first time, a correlation be-
tween activity and structural distribution of the plasma mem-
brane Ca-ATPase (PMCA) in both, MVM (maternal-facing)
and BM (fetal-facing) plasma membranes from the syncytio-
trophoblast of human placenta. It also demonstrates higher
PMCA activity and expression in MVM as compared to
BM. Since the percentage of cross contamination of BM in
the MVM fraction is very low (Table 1) the probability that
the high expression of PMCA in MVM could be due to the
presence of BM in the MVM fractions seems to be very
unlikely. Specific PMCA activity was identified in both
MVM and BM membranes (Tables 2 and 3), which, similar
to red blood cell ghosts , is quite sensitive to the level
of lipid peroxidation of the membranes (Figs. 2 and 3).
PMCA activity of both MVM and BM, show a similar behav-
ior for the tested biochemical parameters, i.e. Kmfor free
Ca2þ; calmodulin stimulation; optimal pH and temperature
values; vanadate, ouabain and thapsigargin sensitivities
(Table 3) as well as its sensitivity to UV-induced lipid perox-
idation (Figs. 2 and 3).
The immunohistochemistry results confirm the presence of
PMCA in polarized plasma membranes of hSCT. Interestingly,
both the PMCA activity (Table 2) and the Western blot data
(Fig. 4) indicate that the PMCA in MVM is, at least, twice
as much as that in BM, probably due to a higher PMCA
PMCA Expression (AU)
PMCA-4 Expression (AU)
Fig. 4. Western blotting and densitometric analysis for PMCA and PMCA 4
isoform in MVM and BM syncytiotrophoblast purified membranes. (A) Histo-
grams of Western blot analysis and representative image (inset) show that total
PMCA expression is observed in both membranes and that the PMCA mark is
bigger in MVM, as compared to the BM fraction. (B) Histograms of Western
blot analysis and representative image (inset) for PMCA4 isoform. (C)
Western blot for PMCA1 isoform. The histograms summarize quantification
of densitometric analyses from six independent placentae, expressed as Arbi-
trary Units (AU) (*p < 0.05; ***p < 0.005).
Fig. 5. Immunohistochemical localization of PMCA in human placental tissue. Panels A through H are placental villous sections showing the hSCT. Scale bar for
all panels represents 10 mm. (A) Control tissue using only secondary goat anti-mouse antibody, without primary antibody (40?). (B) Tissue section stained with
antibody against PLAP (clone 8B6). The arrow indicates MVM specific staining (100?). (C) Immunostaining of tissue section with antibody against all PMCA
isoforms (clone 5F10). The arrow indicates specific PMCA staining, predominantly in hSCT (40?). (D) Detailed microphotography. PMCA immunostaining was
located in the microvillous (short arrow) and basal (long arrow) membranes and it was also distributed in hSCT cell cytoplasm. (E) Immunostaining with antibody
against PMCA4 isoform (clone JA9) revealed specific staining similar to that obtained with antibody against all PMCA isoforms (arrow, 40?). (F) Detailed
microphotography: PMCA4 is located in the microvillous (short arrow) and in the basal (long arrow) hSCT membranes (100?). (G) Negative control was
performed by omitting primary and using only secondary goat anti-rabbit antibody (100?). (H) Immunostaining with antibody against PMCA1 isoform.
PMCA1 staining is distributed in hSCT cell cytoplasm (100?). Nuclei in blue were hematoxylin-stained.
density in MVM. It is clear that, similar to the NKA in hSCT
 and in liver canaliculi epithelial cells , the distribution
of PMCA in hSCT does not follow the pattern of strict polar-
ization of this ATPase seen in classical transporting epithelia.
It is not localized only in the basal membrane, but in both
MVM and BM.
It is known that in placental tissues there are only PMCA4a,
PMCA4b and PMCA1a isoforms . On this regard, our
results are in agreement with these findings. The antibody
JA9 reacts specifically against both PMCA4aand PMCA4b
variants. Fig. 4B seems to show a doublet band (presumably
129 and 134 kDa) in all samples indicating the presence of
both PMCA4aand PMCA4bvariants. On the other hand, the
presence of the PMCA1 isoform in the hSCT plasma mem-
branes was detected with Western blotting of the homoge-
nates; however, none of the purified fractions showed
a specific band for PMCA1 (Fig. 4C). It is clear that this point
deserves further studies in order to identify the possible role of
PMCA1 in hSCT.
The cholesterol content and the phospholipids saturation in-
dex of MVM are known to be higher than those of BM ,
indicating a lower fluidity in the former. Accordingly, since
both PMCA and NKA activities are sensitive to high choles-
terol content and low fluidity of the plasma membranes
[33,34], it could be expected a lower activity of both enzymes
in MVM. However, PMCA and NKA show higher activity
values in MVM than in BM (ref.  and Table 2). This could
be interpreted as an indication of either a higher density of
these ATPase molecules in MVM than in BM, which is in
agreement with our Western blots results (ref.  and
Fig. 4), or a differential sensitivity of these ATPases to both
cholesterol/phospholipids ratio and fluidity of the plasma
In the literature, there are only two reports that failed to
show evidence for PMCA on MVM [13,14]. Accordingly, it
is important to discuss the possible causes of the differences
between the current report and the previous ones suggesting
or showing that PMCA is only present in the BM. In one of
them, Borke et al. , described by immunohistochemistry
and Western blotting techniques with the monoclonal antibody
5F10, the presence of PMCA in both MVM and BM from
hSCT, although the labeling of PMCA in MVM was really
weak. It is clear that Borke et al.  could not rule out the
presence of PMCA in MVM. Furthermore, none of them
[13,14] used morphologic controls in their immunohistochem-
istry studies, e.g. staining of the nuclei. Therefore, in these
studies it is not possible to affirm in a conclusive way, the pres-
ence of PMCA only in the fetal-facing side of hSCT.
Strid and Powell  showed, by means of immunocyto-
chemistry techniques, a clear distribution of PMCA in the
BM only. They also showed, by means of Western blot analy-
ses, the presence of both PMCA1 and PMCA4 in the BM frac-
tions only. However, they did not perform the same Western
blot analyses for the MVM fractions. In contrast, in our West-
ern blot studies using isoform-specific antibodies, PMCA4 but
not PMCA1 was shown to be present in both MVM and BM
(Fig. 4). Although it is difficult to try to explain the possible
sources of differences between the study of Strid and Powell
and our study, one can suggest that the source of the antibodies
used in the current study is clearly different from that used by
Strid and Powell, with the exception of the general PMCA,
clone 5F10. Strid and Powell used a set of antibodies that dif-
fers in their specificity to the N-terminal region of each protein
(first 80e90 amino acids). In our case, we used a polyclonal
antibody against PMCA1, which was made using a synthetic
peptide immunogen corresponding to the residues A(5) N N
S V A Y S G V K N S I K E A N(22) of the rat PMCA1.
This might account for some of the differences in distribution
found for PMCA1 and PMCA4 in hSCT.
A careful revision of previous studies led us to find reports
showing the presence of a high affinity Ca2þ-stimulated
ATPase activity in MVM of human term placenta . This
ATPase activity was shown to be able to transport Ca2þ
against its gradient in MVM vesicles , and to have bio-
chemical characteristics resembling those of PMCA, e.g.,
high affinity for free Ca2þ, vanadate-sensitivity, calmodulin-
stimulation, oxalate-insensitive, optimal pH around 7.2 .
Interestingly, these biochemical characteristics of the active
transport of Ca2þin MVM vesicles, are quite similar to those
shown in Table 3 . Furthermore, there is a report demon-
strating the ultracytochemical localization of a Ca-ATPase
activity in hSCT showing strong ATPase activity in MVM
and a weak activity in BM , which is in agreement with
the biochemical data (Table 3) and the Western blot analyses
(Fig. 4) shown in the current study.
Our results lead us to propose a new model for the transport
of calcium across the hSCT, considering the presence of an
important fraction of PMCA in MVM. All the Ca2þions
acquired by the fetus throughout gestation, must come neces-
sarily from the maternal circulation and, since the Ca2þcon-
centration in the fetal blood is higher than that in the mother
circulation , the ion must be actively transported across
the placenta from the maternal to the fetal circulation. In
fact, the syncytiotrophoblast actively transports close to 80%
of the Ca2þgoing from the maternal to fetal circulation, where
it is needed for fetal skeleton growth, especially during the
third trimester of pregnancy . The directionality of the net
transport of the ion across any of the individual plasma mem-
branes will depend on the balance between the passive influx
of Ca2þand the rate at which the ion is pumped out across that
membrane. Consequently, considering the fact that there is
a higher activity of PMCA in MVM than in BM, the only
way to have a net flux of the ion from the mother to the fetus
is through a higher Ca2þpermeability in MVM than in BM. In
fact, MVM possess more mechanisms of Ca2þentry than
those in BM, indicating a higher Ca2þpermeability in MVM
than BM . In this way, most of the Ca2þentering passively
to the syncytiotrophoblast across the MVM will be returned to
the mother by the PMCA from these membranes. PMCA of
BM, on the other hand, will be responsible for the quantity
of Ca2þextruded through BM, in order to get a net Ca2þtrans-
port from the mother to the fetus. NCX does not seem to play
an important role in the control of Ca2þmovement through
We are grateful to Dr. M. Pe ´rez and the staff of The San
Jose ´ Hospital Maternity Unit for assistance in obtaining the
biological material. Purified bovine brain calmodulin was gen-
erously supplied by Dr. Gustavo Benaim from the Universidad
Central de Venezuela. We thank Dr. Matthew Turvey (Univer-
sity of Birmingham) for his valuable corrections. This research
was supported in part by grant from Fondecyt e Chile No.
1070695, by grant No. H9/181/R427, Project 96350, from the
World Health Organization and by grant from FONACIT e
Venezuela No. F-2005000222.
 Clapham DE. Intracellular calcium. Replenishing the stores. Nature
 Carafoli E. Plasma membrane calcium ATPase: 15 years of work on the
purified enzyme. FASEB J 1994;8:993e1002.
 Carafoli E, Stauffer T. The plasma membrane calcium pump: functional
domains, regulation of the activity, and tissue specificity of isoform
expression. J Neurobiol 1994;25:312e24.
 Zylinska L, Kawecka I, Lachowicz L, Szemraj J. The isoform- and loca-
tion-dependence of the functioning of the plasma membrane calcium
pump. Cell Mol Biol Lett 2002;7:1037e45.
 Stauffer TP, Guerini D, Carafoli E. Tissue distribution of the four gene
products of the plasma membrane Ca2þpump. A study using specific
antibodies. J Biol Chem 1995;270:12184e90.
 Strehler EE, Zacharias DA. Role of alternative splicing in generating
isoform diversity among plasma membrane calcium pumps. Physiol
 Pitkin RM. Calcium metabolism in pregnancy and the perinatal period:
a review. Am J Obstet Gynecol 1985;151:99e109.
 Belkacemi L, Bedard I, Simoneau L, Lafond J. Calcium channels,
transporters and exchangers in placenta: a review. Cell Calcium
 Tuan R, Kushner T. Calcium-activated ATPase of the human placenta:
identification, characterization, and functional involvement in calcium
transport. Placenta 1987;8:53e64.
 Williams JM, Abramovich DR, Dacke CG, Mayhew TM, Page KR.
Inhibitor action on placental calcium transport. Calcif Tissue Int 1991;
 Moreau R, Daoud G, Masse A, Simoneau L, Lafond J. Expression and
role of calcium-ATPase pump and sodium-calcium exchanger in differ-
entiated trophoblasts from human term placenta. Mol Reprod Dev
 Fisher GJ, Kelley LK, Smith CH. ATP-dependent calcium transport
across basal plasma membranes of human placental trophoblast. Am J
 Borke JL, Caride A, Verma AK, Kelley LK, Smith CH, Penniston JT,
et al. Calcium pump epitopes in placental trophoblast basal plasma
membranes. Am J Physiol 1989;257:C341e6.
 Strid H, Powell TL. ATP-dependent Ca2þtransport is up-regulated
during third trimester in human syncytiotrophoblast basal membranes.
Pediatr Res 2000;48:58e63.
 Caires R, Proverbio F, Abad C, Pi~ nero S, Proverbio T, Varela H, et al.
Effect of hypoxia on the Ca-ATPase activity and lipid peroxidation level
in the syncytiotrophoblast plasma membranes. Argentina: III Latin-
American Symposium on Maternal-Fetal Interaction and Placenta: Basic
& Clinical Research. Co ´rdoba; 2007.
 Godoy V, Riquelme G. Calcium ATPase expression in human syncytio-
trophoblast membranes. Argentina: III Latin-American Symposium on
Maternal-Fetal Interaction and Placenta: Basic & Clinical Research.
Co ´rdoba; 2007.
 Navarrete D, Gonza ´lez A, Avellaira C, Riquelme G. Expresio ´n de calcio
ATPasa en membranas de sinciciotrofoblasto. XVII Reunio ´n Anual
Sociedad Chilena de Reproduccio ´n y Desarrollo. Re~ naca: Chile; 2006.
 Johansson M, Jansson T, Powell TL. Naþ-Kþ-ATPase is distributed to
microvillous and basal membrane of the syncytiotrophoblast in human
placenta. Am J Physiol 2000;279:R287e94.
 Illsley NP, Wang ZQ, Gray A, Sellers MC, Jacobs MM. Simultaneous
preparation of paired, syncytial, microvillous and basal membranes
from human placenta. Biochim Biophys Acta 1990;1029:218e26.
 JimenezV, Henriquez M, Llanos P, Riquelme G. Isolation and purification
of human placental plasma membranes from normal and pre-eclamptic
pregnancies. a comparative study. Placenta 2004;25:422e37.
 Nardulli G, Proverbio F, Limongi FG, Marı ´n R, Proverbio T. Preeclamp-
sia and calcium adenosine triphosphatase activity of red blood cell
ghosts. Am J Obstet Gynecol 1994;171:1361e5.
 Bradford MM. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye
binding. Analyt Biochem 1976;72:248e54.
 Proverbio F, Proverbio T, Marı ´n R. Naþ-ATPase is a different entity from
the (NaþþKþ)-ATPase in rat kidney basolateral plasma membranes.
Biochim Biophys Acta 1986;858:202e5.
 Marı ´n R, Rodrı ´guez AJ, Proverbio T. Partial characterization of the
inhibitory effect of lipid peroxidation on the ouabain-insensitive Na-
ATPase of rat kidney cortex plasma membranes. J Bioenerg Biomemb
 Feix JB, Bachowski GJ, Girotti AW. Photodynamic action of merocya-
nine 540 on erythrocyte membranes: structural perturbation of lipid
and protein constituents. Biochim Biophys Acta 1991;1075:28e35.
 Matteo R, Proverbio T, Co ´rdova K, Proverbio F, Marı ´n R. Preeclampsia,
lipid peroxidation, and calcium adenosine triphosphatase activity of red
blood cell ghost. Am J Obstet Gynecol 1998;178:402e8.
 Adamo HP, Caride AJ, Penniston JT. Use of expression mutants and
monoclonal antibodies to map the erythrocyte Ca2þpump. J Biol
 CavieresJD.Calmodulinandthetargetsizeofthe(Ca2þþ Mg2þ)-ATPase
of human red-cell ghosts. Biochim Biophys Acta 1984;771:241e4.
 Kelley LK, Borke JL, Verma AK, Kumar R, Penniston JT, Smith CH.
The calcium-transporting ATPase and the calcium- or magnesium-
dependent nucleotide phosphatase activities of human placental
trophoblast basal plasma membrane are separate enzyme activities.
J Biol Chem 1990;265:5453e9.
 Oviedo NJ, Benaim G, Cervino V, Proverbio T, Proverbio F, Marin R. The
plasma membrane Ca2þ-ATPase protein from red blood cells is not mod-
ified in preeclampsia. Biochim Biophys Acta 2006;1762:381e5.
 Takemura S, Omori K, Tanaka K, Matsuura S, Tashiro Y. Quantitative
immunoferritin localization of Naþ/KþATPase on canine hepatocyte
cell surface. J Cell Biol 1984;99:1502e10.
 Powell TL, Jansson T, Illsley NP, Wennergren M, Korotkova M,
Strandvik B. Composition and permeability of syncytiotrophoblast
plasma membranes in pregnancies complicated by intrauterine growth
restriction. Biochim Biophys Acta 1999;1420:86e94.
 Kimelberg HK, Papahadjopoulos D. Effects of phospholipid acyl chain
fluidity, phase transitions, and cholesterol on (NaþþKþ)-stimulated
adenosine triphosphatase. J Biol Chem 1974;249:1071e80.
 Ohta A, Mohri T, Ohyashiki T. Effect of lipid peroxidation on
membrane-bound Ca2þ-ATPase activity of the intestinal brush-border
membranes. Biochim Biophys Acta 1989;984:151e7.
border membranes of the human term placenta. Placenta 1986;7:365e73.
 Treinen KA, Kulkarni AP. ATP-dependent Ca2þuptake in brush border
membranes of human term placenta. Placenta 1987;8:477e86.
 Matsubara S, Tamada T, Saito T. Cytochemical study of the electron-
microscopical localization of CaþþATPase activity in the human
trophoblast. Nippon Sanka Fujinka Gakkai Zasshi 1987;39:1080e6.