Activity and Protein Expression of Na (cid:1) /K (cid:1) ATPase Are Reduced in Microvillous Syncytiotrophoblast Plasma Membranes Isolated from Pregnancies Complicated by Intrauterine Growth Restriction

In contrast to classical transporting epithelia, the Na (cid:1) /K (cid:1) AT- Pase is distributed to both the microvillous membrane (MVM) and the basal membrane (BM) of the placental syncytiotro- phoblast.Na (cid:1) /K (cid:1) ATPaseisimportantinmaintainingtheelec-trochemical gradient for Na (cid:1) , which represents the driving forceforNa (cid:1) -coupledtransportofnutrients.Wehypothesized that syncytiotrophoblast Na (cid:1) /K (cid:1) -ATPase activity is reduced in intrauterine growth restriction (IUGR). We isolated MVM and BM from control (n (cid:2) 10) and IUGR placentas (n (cid:2) 11). The protein expression of Na (cid:1) /K (cid:1) -ATPase (cid:3) 1 -subunit was deter-mined by Western blotting and found to be slightly reduced in MVM isolated from IUGR ( (cid:4) 10%; P < 0.05) placentas. Na (cid:1) /K (cid:1) ATPase activity was measured as the ouabain-sensitive, K (cid:1) dependent cleavage of the fluorescent pseudosubstrate 3- O -methylfluoresceinphosphateandwasreducedby35%inMVMobtainedfromIUGRplacentas( P < 0.02). To assess the tran- scriptionallevelsofNa (cid:1) /K (cid:1) -ATPasemRNA,realtimePCRwas used. No significant changes in steady state mRNA levels for Na (cid:1) /K (cid:1) -ATPase were detected. The expression of the Na (cid:1) /K (cid:1) ATPase (cid:3) 1 -subunit and Na (cid:1) /K (cid:1) -ATPase activity in the BM were unaffected in cases of IUGR. These data suggest that Na (cid:1) /K (cid:1) -ATPase activity is reduced in the MVM of placentas from IUGR pregnancies. These changes might impair the function of Na (cid:1) -coupled transporters and contribute to the reduced growth of these fetuses. ( J Clin Endocrinol Metab 88: 2831–2837, 2003)

I NTRAUTERINE GROWTH RESTRICTION (IUGR) is the second most important cause of perinatal and infant mortality and morbidity (1). The adverse effects of restricted intrauterine growth reach beyond infancy because IUGR is associated with neurological sequelae (2) as well as with an increased risk for illnesses as diverse as cardiovascular disease (3), diabetes (4), and schizophrenia in adult life (5). The pathophysiological mechanisms underlying IUGR are not well established. Impaired nutrient and oxygen supply to the developing fetus due to compromised uteroplacental circulation and/or placental transport is likely to represent a common denominator for many cases of IUGR. Evidence has accumulated that IUGR is associated with specific alterations in transport systems in the syncytiotrophoblast, the transporting epithelium of the human placenta. For example, reduced activities of several amino acid transporters in the syncytiotrophoblast microvillous membrane (MVM) isolated from IUGR placentas have been reported (6,7,8), and the reduced levels of certain amino acids in fetal cord blood in IUGR (9, 10) may be a physiological correlate to these findings. Many transport processes across the MVM are sodium-coupled and consequently are dependent on the maintenance of a low intrasyncytial Na ϩ concentration. Sodiumcoupled counter transport is the driving force for the Na ϩ /H ϩ exchanger present in the MVM and represents an important mechanism for the extrusion of protons from the syncytiotrophoblast (11). Amino acid and phosphate uptake are other examples of sodium-coupled cotransport of nutrients.
As in all eukaryotic cells, a low intracellular Na ϩ concentration is maintained by the Na ϩ /K ϩ ATPase that exports three sodium ions for every two potassium ions imported. The establishment of an electrochemical gradient for Na ϩ across the plasma membrane is vital for cell functions as diverse as the propagation of nerve signals, volume regulation, nutrient absorption, and pH regulation. In most cells, it is diffusely distributed to the entire plasma membrane, whereas in epithelia the distribution of Na ϩ /K ϩ ATPase is polarized in patterns specific for the type of epithelium. We have previously shown that in the placental syncytiotrophoblast, the Na ϩ /K ϩ ATPase is distributed to both MVM and BM and that the activity and protein expression of the transporter are higher in MVM (12). This is in contrast to the basolateral localization in classical transporting epithelia, such as in the kidney and intestine.
An alteration in the capacity of the syncytiotrophoblast Na ϩ /K ϩ ATPase to maintain the Na ϩ gradient may indirectly affect all sodium-dependent transport systems. In this study, we hypothesized that the activity of syncytial Na ϩ /K ϩ ATPase is reduced in pregnancies complicated by IUGR, with the possible implication that this may contribute to the impairment of placental transport in this condition. To this end, we isolated MVM and BM from placentas obtained from IUGR as well as control placentas. Using these purified membranes we assessed the expression of Na ϩ /K ϩ ATPase by Western blotting and carried out activity studies using 3-O-methylfluorescein phosphate. To investigate the transcriptional levels of Na ϩ /K ϩ ATPase, we performed reverse transcription followed by real time PCR.

Patients and Methods Patients
Collection of placental tissue was carried out with informed consent, according to protocols approved by the Committee for Research Ethics at Gö teborg University. Gestational age was estimated from the last menstrual period and confirmed by ultrasound at 16 -18 wk gestation. Control placentas [appropriately grown for gestational age (AGA)] were obtained from uncomplicated term pregnancies (n ϭ 6) and from preterm deliveries (n ϭ 4; range, 28 -36 wk gestation) with birth weights that were appropriate for gestational age. These babies were delivered vaginally or by means of cesarean section due to maternal indication. IUGR was defined as a birth weight more than 2 sd below mean birth weight for gestational age (13). Placentas from IUGR pregnancies were obtained from preterm (n ϭ 5; range, 29 -36 wk gestation) and term pregnancies (n ϭ 6). In one of the preterm IUGR pregnancies, an additional pregnancy complication was diagnosed (gestational diabetes), whereas no other complication was present in the majority of IUGR cases (10 of 11). IUGR babies were delivered vaginally or by cesarean section due to indications of fetal distress. All preterm IUGR pregnancies were associated with increased pulsatility index in the umbilical artery as determined using Doppler ultrasound. In the term IUGR group, umbilical artery pulsatility index was moderately elevated in three cases and normal in three pregnancies. Other indications of fetal compromise were, however, present in the term IUGR group, including oligohydramnios (two cases) and signs of fetal distress during labor (decreased cardiotocography variability and tachycardia) in three cases.

Isolation of membrane vesicles
The preparation of BM and MVM vesicles was performed as described previously (14 -16). All procedures were conducted on ice, and the centrifugation steps were conducted at 4 C. In short, after delivery, placentas were immediately placed on ice, and preparation was initiated within 30 min. Placentas were dissected, and the decidua, chorionic plate, and amniotic sac were removed. Subsequently, approximately 100 g of villous tissue was cut into small pieces and rinsed with ice-cold physiological saline. Tissue was placed in buffer D [250 mm sucrose, 0.7 m pepstatin A, 1.1 m leupeptin, 0.8 m antipain, 80 nm aprotinin, 10 mm HEPES-Tris (pH 7.4)] at 4 C and homogenized using a polytron (Kinematika AG, Littau, Switzerland). The homogenate was centrifuged twice at 10,000 ϫ g for 15 min, and the resulting supernatant was centrifuged at 125,000 ϫ g for 30 min. The pelleted crude membrane fraction [postnuclear membrane pellet (P2)] was resuspended in buffer D, followed by the addition of 12 mm MgCl 2 . The resulting suspension was stirred slowly on ice. Subsequently, the suspension was centrifuged for 10 min at 2,500 ϫ g. The supernatant, which contained the MVM, was centrifuged for 30 min at 125,000 ϫ g, and the pellet containing the BM was further purified by means of a sucrose step gradient centrifugation. Finally, BM and MVM were centrifuged at 125,000 ϫ g for 30 min and resuspended in a volume of buffer D appropriate to give a final protein concentration of 5-10 mg/ml. Vesicles were aliquoted, snap-frozen in liquid nitrogen, and stored at Ϫ80 C until further use.

Assessment of vesicle purity
The commonly used MVM marker, alkaline phosphatase, is highly abundant in the syncytiotrophoblast microvillous membrane, whereas the activity is low or absent in other cell membranes in the human placenta (17). Alkaline phosphatase activity was measured according to standard protocols (18), and enzyme enrichments were calculated as the activity in the MVM and BM fractions relative to the activity in homogenate or P2. Adenylate cyclase is almost exclusively polarized to the BM of the syncytiotrophoblast cell (19 -21), and adenylate cyclase activity served in this study as a BM marker. Forskolin-induced cAMP production was measured using well established protocols (22) by RIA (NEN Life Science Products, Boston, MA). Tissue homogenates contain various factors of cytosolic origin that activate adenylate cyclase (14). Therefore, the adenylate cyclase activity of the P2 was used as the denominator in calculations of enrichments of adenylate cyclase.

Isoform specific antibodies
The monoclonal ␣ 1 -specific antibody used in this study, developed by Dr. Douglas Fambrough of Johns Hopkins University (Baltimore, MD), was obtained from the Developmental Studies Hybridoma Bank, developed under the auspices of the National Institute of Child Health and Human Development, and maintained by the University of Iowa, Department of Biological Sciences (Iowa City, IA; www.uiowa.edu/ϳdshbwww/). The antibody was made using the ␣ 1 -subunit of chicken Na ϩ / K ϩ -ATPase, and the specificity and other information on this antibody is given elsewhere (23). The antibody is cross-reactive with human and rat ␣ 1 -subunits.

Western blotting
Membrane proteins from MVM and BM were separated by SDS-PAGE as follows. Vesicle suspension was thawed on ice and diluted with buffer D so that the appropriate protein concentration was obtained. One volume of 3ϫ sample buffer [8 m urea, 170 mm (5%) SDS, 0.04 U bromphenol blue, 455 mm dithiotreitol, 50 mm Tris (pH 6.8 adjusted with HCl)] was then added to two volumes of this diluted vesicle suspension, followed by vigorous mixing. For electrophoresis, the Bio-Rad mini Protean II electrophoresis system (Bio-Rad Laboratories, Inc., Hemel Hempstead, Herts, UK), was used; 10 g vesicle protein were loaded on a 7.5% SDS polyacrylamide gel, which was mounted into a holding cassette filled and surrounded by electrophoresis buffer (25 mm Trisbase, 192 mm glycine, and 3.5 mm SDS). Appropriate molecular weight markers were also loaded. Electrophoresis was performed at 200 V during 45 min. Gels were thereafter equilibrated with transfer buffer [25 mm Tris-base, 192 mm glycine, and 20% (vol/vol) methanol in deionized water], by gentle agitation for 30 min. The gel was covered with a nitrocellulose transfer membrane (Hybond-ECL, Amersham International, Buckinghamshire, UK), and equilibrated with transfer buffer. The gel was placed in a Bio-Rad mini trans-blot electrophoresis transfer cell, covered with buffer, and transferred overnight at 30 V. The membrane was then blocked for 1 h in 5% Blotto buffer [0.1 m PBS, 0.1% Tween 20 (vol/vol), and 5% (wt/vol) nonfat dry milk]. After the membrane was washed in PBS/Tween (0.1 m PBS and 0.1% Tween 20), the primary antibody (diluted in PBS/Tween with 1.2 mm thimerosal) was added at a dilution of 1:1000 and incubated for 1 h at room temperature. After washing in PBS/Tween, the secondary antibody labeled with horseradish peroxidase was added and allowed to incubate for 1 h, followed by rinsing in PBS/Tween. The final detection was done using the ECL Western blot detection system (Amersham International). Relative density of the bands was evaluated by densitometry, using the Imaging Processing lab gel system (Signal Analytics Corp., Vienna, VA). Three control samples were repeatedly run on each gel to allow for equalization between gels. This was done by calculating the average of the three control samples on each gel and then a mean value for all control samples. This value was used to determine the equalization factor for each gel, and all densitometries values were subsequently corrected. AGA samples from all gels were then averaged, and samples were divided by the mean AGA value to generate the relative density used in subsequent statistical analysis.

Na ϩ /K ϩ -ATPase assay
The activity of Na ϩ /K ϩ -ATPase in MVM and BM fractions was measured by a fluorometric method (24,25) with some modifications (12). The method is based on the on-line determination of change in fluorescence due to the formation of the fluorescent compound 3-Omethylfluorescein (3-O-MF) from the parent compound 3-O-methylfluorescein phosphate. Fluorescence was assayed in a fluorometer (Photon Technology International, Lawrenceville, NJ) with the reaction mixture in a thermostatically controlled 2-ml cuvette. The excitation wavelength was 470 nm, and the emission wavelength was 510 nm. The amplitude of the emission was shown to be proportional to the concentration of 3-O-MF, and a standard curve was created using varying concentrations of 3-O-MF. The assay medium contained 50 m 3-Omethylfluorescein phosphate, 5 mm creatinine phosphate, 4 mm MgCl 2 , 0.5 mm EGTA, and 80 mm Tris HCl (pH 7.2). All assays were performed under continuous stirring at 37 C. To establish the background fluorescence, a 30-sec baseline period was recorded. Subsequently, 10 l of vesicles (Ϸ100 g protein) were added, and fluorescence was recorded during 90 sec. To activate the Na ϩ /K ϩ -ATPase, 10 l of 2 m KCl was added, giving a final concentration of 10 mm KCl, and fluorescence was recorded during an additional 90 sec. To assess any volume effects of KCl addition on fluorescence, KCl was replaced by equimolar choline chloride in control experiments. In addition, similar experiments were performed with vesicles added to assay medium containing 3 mm ouabain. The assays were performed on frozen samples; however, we have shown previously that freezing does not affect Na ϩ /K ϩ ATPase activity significantly (12). The activity of Na ϩ /K ϩ -ATPase was calculated as the slope difference of fluorescence recordings, before and after addition of potassium, with and without ouabain in the buffer. These values were related to the protein content of the vesicle preparations, giving the final expression of activity as nanomoles of liberated phosphate (Pi) per minute Ϫ1 per milligram protein Ϫ1 .

Isolation of RNA and reverse transcription
Immediately upon delivery, the placenta was placed on ice, and the decidual layer was removed. Pieces of trophoblast tissue were immersed in RNA isolation solution RNA stat 60 (Tel-Test, Friendswood, TX) and snap-frozen in liquid nitrogen. RNA was isolated according to the manufacturer's protocol. Briefly, the tissue was homogenized on ice using a polytron and allowed to stand for 5 min. Chloroform at 0.2 ml per initial milliliter RNA stat 60 was added, and the sample was mixed and centrifuged. The supernatant was discarded, and the pellet was washed in 75% ethanol followed by centrifugation. The pellet was allowed to dry to near completeness and dissolved in RNAse free water. The RNA concentration was determined using absorbance at 260 nm, and the integrity of all samples was determined by 1.2% agarose gel electrophoresis during denaturing conditions.
Reverse transcription was performed using the Superscript II kit according to the manufacturer's instructions (Life Technologies, Inc., Täby, Sweden). Briefly, the reaction was initiated by the addition of 2 g total RNA to random primers and nuclease free H 2 O, yielding a final volume of 24 l. The mixture was heated to 70 C during 10 min and placed on ice. To this, 8 l of 5ϫ first strand buffer, 4 l of 0.1 m dithiotreitol, and 2 l of 10 mm deoxynucleoside triphosphate was added, followed by vortexing and incubation for 10 min at 25 C and 2 min at 42 C. At this stage, 2 l of reverse transcriptase was added, and the resulting reaction was heated to 42 C for 60 min and 70 C for 15 min. The resulting cDNA was placed on ice, and 60 l H 2 O was added to each tube.

Real-time PCR
The real-time PCR method in this study is based on TaqMan chemistry. A fluorogenic probe is 5Ј labeled with the reporter dye FAM (6-carboxy-fluorescein). The 3Ј end of the probe is labeled with the quencher dye TAMRA (6-carboxy-tetramethyl-rhodamine). During the PCR process, this probe anneals to a base stretch situated between the forward and reverse primers. The DNA polymerase has a 5Ј-nuclease activity that by enzymatic cleavage separates the reporter dye from the quencher dye. This allows the on-line detection of the fluorescent probe. The exponential increase of fluorescence per PCR cycle is plotted against cycle number, thus generating an amplification plot of sigmoidal shape. The term threshold cycle (C T ) denotes the fractional cycle at which the reporter fluorescence reaches a certain level, in this case 10 times the sd of the background fluorescence that makes up the baseline. Higuchi et al. (26) have shown a linear relationship between C T and the log of initial target copy number. The C T value was therefore used to calculate the initial amounts of Na ϩ /K ϩ ATPase mRNA.

Oligonucleotides for TaqMan PCR assay
The GenBank database (http://www.ncbi.nlm.nih.gov/) was used to obtain the cDNA sequences for the genes assayed in this study. The nucleotide sequences for primers and FAM-TAMRA labeled probe are given in Table 1. All primer pairs were constructed so that the resulting amplicon spanned an exon junction to avoid the detection of genomic DNA. The TaqMan probes and oligonucleotide primers were designed with the software construction program Primer Express, version 1.0 (Perkin-Elmer Corp., Wellesley, MA). As endogenous control, the amplification of 18 S rRNA was used. Primers and probes for 18S rRNA were obtained in a Pre-Developed TaqMan Assay Reagent kit. All probes, primers, and universal master mix was purchased from Applied Biosystems (Stockholm, Sweden).

The PCR amplification
For PCR, the ABI PRISM 7700 Sequence Detector was used (Perkin-Elmer Corp.), equipped with a Gene-amp PCR system 9600 and a chargecoupled device camera set to monitor fluorescent emission spectra between the wavelengths 500 and 650 nm. For amplification of the reference genes, a 1:8 dilution of cDNA was added to the PCR consisting of 1ϫ TaqMan Buffer A, 5 mm MgCl 2 , 0.2 mm deoxynucleoside triphosphate (20 mm dUTP; 10 mm of dATP, dCTP, and dGTP; 1.25 U Taq Gold polymerase; 0.5 U AmpErase UNG; 15 pmol of each primer; and 5 pmol probe in a final volume of 50 l. Thermal cycling began by heating to 52 C for 2 min, followed by heating to 95 C for 10 min. Hereafter, a 50-cycle two-step PCR was run with the following parameters: 15 sec at 95 C and 1 min at 60 C. Samples were run in triplicates and in one assay.

Data presentation and statistical analysis
Data are given as mean Ϯ sem. Differences in immunoblots, activity data, and RNA quantification between AGA and IUGR groups were compared using t test. The relationship between gestational age and expression or activity was tested by linear regression analysis. Significance was defined at the P value less than 0.05 level.

Clinical data
Selected clinical data for control and IUGR groups are given in Table 2. Gestational age was not significantly different between groups. In the IUGR group, birth weight was 41% lower and placental weight was 40% lower as compared with the AGA group. Ponderal index was significantly lower in the IUGR group, consistent with asymmetric growth restriction. Therefore, the presence of clear signs of fetal compromise (such as increased pulsatility index in the umbilical artery, oligohydramnios, asphyxia, etc.), together with the low ponderal index, strongly suggest that IUGR babies were truly growth restricted rather than genetically small.

Purity of membrane vesicles
The marker enzyme activities for placental homogenate, MVM and BM fractions, are shown in Table 3. The marker for BM, adenylate cyclase activity, showed a 27-fold enrichment in the BM fraction from control placentas and a 29-fold enrichment in IUGR samples. The marker enzyme for MVM, alkaline phosphatase activity, was enriched 18-fold in the MVM fraction isolated from control placentas and 17-fold in IUGR samples. Marker enzyme activities in vesicles prepared from IUGR placentas were not significantly different from control placentas. It has been shown previously that this preparation is characterized by a low degree of cross contamination between syncytial plasma membranes as well as an insignificant contamination of intracellular or nonepithelial membranes (14 -16).

Immunoblotting
As shown previously, the density of the ␣ 1 -subunit of Na ϩ /K ϩ -ATPase was higher in MVM than in BM (data not shown). Figure 1 shows a representative Western blot of syncytiotrophoblast MVM from AGA and IUGR placentas. Figure 2 shows that the density of the 96-kDa band representing the ␣ 1 -subunit of Na ϩ /K ϩ -ATPase was reduced by 10% (n ϭ 10, control; n ϭ 11, IUGR; P Ͻ 0.02) in MVM isolated from IUGR placentas compared with controls, whereas there was no change in BM in association with IUGR. The samples used in the study covered the gestational period from 28 -40 wk in both AGA and IUGR. The expression of Na ϩ /K ϩ ATPase was analyzed by linear regression with respect to gestational age in both MVM and BM. No significant relationship between gestational age and expression of Na ϩ /K ϩ ATPase existed in the IUGR, AGA, or combined groups (data not shown).
Na ϩ /K ϩ -ATPase activity Figure 3 demonstrates the activity of the Na ϩ /K ϩ -ATPase in MVM and BM from IUGR and control placentas. The activity in IUGR MVM (2.5 Ϯ 0.47 nmol P i ⅐mg protein Ϫ1 ⅐min Ϫ1 ) was decreased by 35%, compared with control vesicles (3.9 Ϯ 0.35 nmol P i ⅐mg protein Ϫ1 ⅐min Ϫ1 ; P Ͻ 0.05). No differences between IUGR and control were observed in the BM. Analysis of Na ϩ /K ϩ ATPase activity over FIG. 2. Densitometry analysis of Western blots probed for the ␣ 1isoform of the Na ϩ /K ϩ ATPase. MVM and BM isolated from AGA (n ϭ 10) and IUGR (n ϭ 11) placentas were analyzed. Control MVM was arbitrarily assigned the value of 1, and the other categories normalized to that. Data are shown as mean Ϯ SEM. *, P Ͻ 0.05, t test.
FIG. 3. Na ϩ /K ϩ -ATPase activity in placental MVM and BM from AGA and IUGR pregnancies. Activity is defined as potassium elicited and ouabain inhibitable release of phosphate in nanomoles per milligram protein and minute. Error bars represent SEM. *, P Ͻ 0.02, t test. Mean Ϯ SEM. Alkaline phosphatase (ALP) activity (in nanomoles PO 4 /sec ϫ milligrams) is a marker for the MVM, and adenylate cyclase (AC) activity (in picomoles/minute ϫ milligrams) represents a marker for the BM. Enrichment is given within parentheses. Enzyme enrichments were calculated as vesicle enzyme specific activity relative to that for the homogenate (Hom) for ALP or P2 for AC. gestation by linear regression demonstrated a significant reduction in activity with increasing gestation in MVM from AGA and BM from IUGR placentas.

Real time PCR Na ϩ /K ϩ ATPase mRNA quantification
The amplification efficiencies for target gene and the endogenous control over a 1:1-1:64 dilution series of cDNA are shown in Fig. 4, along with the regression analysis data. The efficiencies were similar, indicating that they could be used to compare RNA levels. The steady state mRNA levels of Na ϩ /K ϩ ATPase are shown in Fig. 5. No significant differences in mRNA expression levels of Na ϩ /K ϩ ATPase could be detected between AGA (n ϭ 8) and IUGR (n ϭ 6) groups.

Discussion
Many transport processes across MVM, the maternalfacing plasma membrane of the syncytiotrophoblast, are coupled to the transport of Na ϩ and energized by the Na ϩ gradient across MVM. Examples of such transport processes are the uptake of neutral amino acids mediated by system A, uptake of taurine by the taurine transporter, and the extrusion of protons mediated by Na ϩ /H ϩ exchanger. Because Na ϩ /K ϩ ATPase is the primary transporter maintaining the electrochemical gradient for Na ϩ across syncytiotrophoblast plasma membranes, the decreased Na ϩ /K ϩ ATPase activity in MVM in association with IUGR may impair placental nutrient uptake and proton elimination. As a consequence, these changes could indirectly affect nutrient transport to the fetus and placental and fetal pH regulation. A direct link between Na ϩ /K ϩ ATPase activity and nutrient uptake in transporting epithelia is supported by numerous experimental studies. For example, inhibition of Na ϩ /K ϩ ATPase resulted in a reduced nutrient uptake in the transporting ep-ithelia of the kidney and intestine (27,28). Similarly, ouabain markedly inhibited amino acid uptake in different in vitro preparations of human placental tissue (29 -31).
Previous studies on placental Na ϩ /K ϩ ATPase in IUGR are scarce. Fuchi et al. (32) showed in a rat model of IUGR that the activity of placental Na ϩ /K ϩ ATPase is depressed in IUGR in the stroke prone simultaneously hypertensive rats strain. The severity of IUGR was directly proportional to the decrease in activity of Na ϩ /K ϩ ATPase. In terms of implications for human IUGR, these results should be interpreted with some caution. The SHRSP rat strain is used as a model for pregnancy-induced hypertension and IUGR, but the degree of similarity to the human condition remains speculative. Furthermore, the rat placenta is hemotrichorial and structurally different from its human hemomonochorial counterpart. The activity was assayed on placental homogenates, and because the Na ϩ /K ϩ ATPase is ubiquitously expressed in all mammalian cells, it is difficult to ascribe a reduction in Na ϩ /K ϩ ATPase activity in whole homogenate to a specific placental cell type. Maxwell et al. (33) showed a reduction of ␣ 1 -Na ϩ /K ϩ ATPase mRNA in placentas from pregnancies complicated by preeclampsia and IUGR. The decrease was, however, not statistically significant and was not supported by a corresponding reduction in protein expression levels.
IUGR is associated with reduced activity of several Na ϩcoupled transporters in MVM, such as the system A amino acid transporter (7), the taurine transporter (8), and Na ϩ /H ϩ exchanger (34), as well as Na ϩ -independent transport systems such as system L (6). We have suggested previously that the apical localization of Na ϩ /K ϩ ATPase in the syncytiotrophoblast might constitute a route of rapid elimination of sodium that has entered the syncytium via cotransport systems (12). Therefore, it may be argued that the decreased MVM Na ϩ /K ϩ ATPase activity in IUGR may be secondary to a reduction of the activity of Na ϩ -coupled transporters in MVM that would tend to decrease the influx of sodium into the syncytium. This issue cannot be resolved easily and might, for example, require direct measurement of intrasyncytial Na ϩ concentrations in primary villous fragments obtained from IUGR pregnancies. If the reduced MVM Na ϩ K ϩ FIG. 4. The amplification efficiency for Na ϩ /K ϩ -ATPase and 18S. The amplification efficiencies were similar for target gene and endogenous control throughout a series of 2-fold dilutions from 1:1 to 1:64. f, 18 S; E, Na ϩ /K ϩ -ATPase.
FIG. 5. Na ϩ /K ϩ -ATPase mRNA expressional levels in RNA isolated from AGA and IUGR placentas. Control MVM is arbitrarily assigned the value of 1, and the IUGR results have been normalized to that; n ϭ 8 for AGA, and n ϭ 6 for IUGR. Error bars represent SEM.
ATPase activity in IUGR is secondary to decreased activity of Na ϩ -coupled transporters, one might expect an up-regulation of Na ϩ K ϩ ATPase in conditions that are characterized by an increased activity of Na ϩ -dependent transport system. However, in diabetes, MVM system A is markedly upregulated (35), whereas syncytiotrophoblast Na ϩ K ϩ ATPase activity is unaltered (36).
Our data suggest that the reduced activity of Na ϩ /K ϩ ATPase in MVM from IUGR placentas cannot be explained by a decreased transcription of ␣ 1 -Na ϩ /K ϩ ATPase mRNA. It should be emphasized, however, that mRNA was isolated from placental homogenates and may not fully represent changes in mRNA levels in the syncytiotrophoblast. More importantly, the density of the ␣ 1 -Na ϩ /K ϩ ATPase protein in MVM obtained from IUGR placentas was reduced only to a limited extent (Ϫ10%) compared with the quite pronounced reduction in Na ϩ /K ϩ ATPase activity (Ϫ35%). These data indicate that the observed change in Na ϩ /K ϩ ATPase activity is due to modulation of the transporter protein, e.g. by alterations in phosphorylation. In order for Na ϩ /K ϩ ATPase to attain full activity, association of ␣and ␤-subunits is required, and it is possible that the decreased MVM Na ϩ /K ϩ ATPase in IUGR may be due to a reduction in the protein expression of Na ϩ /K ϩ ATPase ␤-subunits.
The number of samples studied was limited, making subgroup analysis for the effect of gestational age uncertain. For example, Na ϩ /K ϩ ATPase activity was not significantly different between preterm and term AGA and IUGR groups. However, when using linear regression analysis, Na ϩ /K ϩ ATPase activity declined significantly with increasing gestational age (wk 28 -40) in MVM from AGA placentas. The MVM from IUGR placentas showed a lower level of activity throughout this period, and the activity did not change significantly with advancing gestation. These findings resemble recently published data concerning the effects of IUGR on Na ϩ H ϩ exchanger (NHE) activity (34). That study showed a significant decrease in NHE activity and expression in MVM from preterm IUGR compared with preterm AGA, but not in term MVM samples. It was speculated that the IUGR pregnancies that were delivered prematurely may represent a subgroup of patients with severe growth restriction and/or reduced capacity to adapt, perhaps due to alterations in placental NHE function that lead to fetal acidosis. It is possible that MVM Na ϩ /K ϩ ATPase activity is also primarily altered in IUGR fetuses delivered prematurely; however, a larger sample size is needed to clarify the relationship between gestational age and Na ϩ /K ϩ ATPase activity in MVM and BM isolated from normally grown and IUGR placentas.
In both the Western blot and activity measurements of Na ϩ /K ϩ ATPase, the data are expressed per unit membrane protein, which is assumed to represent membrane area. The comparison with IUGR could be confounded if there is a marked difference in the amount of protein per membrane area in this pregnancy complication. However, this is not the case because the phospholipid/protein ratio in IUGR MVM and BM are similar to that in AGA vesicles (37). The total transport capacity is dependent on transporter activity per membrane area as well as the total membrane surface area available for exchange. Total trophoblastic surface area has been reported to be reduced roughly in proportion to the reduced fetal and placental weight in idiopathic IUGR (38). Therefore, the finding of a decrease in Na ϩ K ϩ ATPase activity per MVM membrane area in IUGR may result in a total transport capacity that is more reduced than fetal and placental size.
Na ϩ /K ϩ ATPase in different tissues is regulated by a wide variety of hormonal and nonhormonal factors (39). The Na ϩ /K ϩ ATPase expression and activity have been shown to be both up-regulated and down-regulated during disease states. In skeletal muscle, insulin-treated diabetes mellitus and hyperthyroidism lead to an up-regulation of Na ϩ /K ϩ ATPase (40,41). The opposite is seen during hypothyroidism and McArdle's disease (42). It is known that IUGR is associated not only with impaired growth but also with an altered hormonal profile of both the mother and the fetus (43)(44)(45). It is therefore conceivable that the perturbed Na ϩ /K ϩ ATPase activity might be a result of a changed hormonal status associated with IUGR.