Carnitine plays an important physiological role, in particular, in β-oxidation because it facilitates long-chain fatty acid transport across the inner mitochondrial membrane. Moreover, carnitine is involved in intracellular coenzyme A homeostasis and functions as an antioxidant (Bremer, 1983; Arduini et al., 1992; Pons and De Vivo, 1995). Only a few organs like brain, liver, and kidney have the ability to biosynthesize carnitine (Bremer, 1983), whereas other tissues like skeletal and heart muscles, where β-oxidation plays a major role in energy metabolism, are highly dependent on active carnitine uptake from blood to maintain their carnitine steady-state concentration (Siliprandi et al., 1989).

Recent studies describe the organic cation transporter novel type II (OCTN2) as a high affinity uptake system for carnitine. The OCTN2 cDNA codes for 557 amino acids consisting of 12 putative transmembrane domains with a predicted molecular mass of 63 kDa (Wu et al., 1998). The transport of carnitine is sodium-dependent (Tamai et al., 1998), whereas other compounds such as tetraethylammonium are transported by OCTN2 in a sodium-independent way (Ohashi et al., 2001).

Besides its physiological function, OCTN2 is of pharmacological relevance. Drugs like verapamil, pyrilamine, and β-lactam antibiotics have been characterized as substrates of OCTN2 and/or inhibitors of carnitine transport (Ohashi et al., 1999, 2001; Wu et al., 1999; Ganapathy et al., 2000). Using screening approaches, the OCTN2 mRNA was detected in kidney, heart, skeletal muscle, and placenta (Tamai et al., 1998). Although OCTN2 expression in the kidney seems to be mainly involved in carnitine reabsorption, expression in muscle cells is assumed to be responsible for carnitine homeostasis in this tissue. The physiological role is further supported by studies of systemic carnitine deficiency, which was associated with mutations in the OCTN2 gene (Vaz et al., 1999; Lamhonwah et al., 2002).

Interestingly, previous studies indicated a limited capacity of the fetal organism for fatty acid oxidation, which is a consequence of incompletely matured carnitine biosynthesis (Novak et al., 1981). This assumption leads to the conclusion that placental uptake of carnitine is pivotal for the fetal organism (Schmidt-Sommerfeld et al., 1981). In fact, sodium-dependent carnitine uptake has been demonstrated for the human choriocarcinoma-derived JAR cell line and brush-border syncytiotrophoblast vesicles. The protein involved in placental uptake of carnitine, however, has not been identified. (Prasad et al., 1996; Roque et al., 1996).

We therefore examined placental carnitine transport in relation to expression, localization, and function of OCTN2. We assayed OCTN2 mRNA levels in term and preterm placenta, localized the protein by immunofluorescence staining to the apical membrane of syncytiotrophoblasts, and determined sodium-dependent carnitine uptake in basal and apical fractions of syncytiotrophoblast membrane vesicles from term and preterm placentas. Moreover, the influence of cellular differentiation on OCTN2 expression has been assessed using isolated trophoblasts.

Materials and Methods

Human Samples. Chorionic villous tissues were obtained from women undergoing caesarian and normal birth. A total of 31 samples from preterm (17) and term (14) placentas were used in the present study following written informed consent. Placentas were collected after normal vaginal deliveries and caesarian sections. Pathological samples such as gestational diabetes and preeclampsia were excluded. Samples for isolation of trophoblasts were taken from term placentas of normal deliveries. Demographic data of the placentas are summarized in Table 1.

Cytotrophoblast Culture. Cytotrophoblasts were isolated as described previously by Kliman et al. (1986). Placental cotyledons were prepared, minced, and washed with 0.9% saline. The tissue was then digested three times using trypsin and DNase I dissolved in Hanks′ balanced salt solution without Ca²⁺ and Mg²⁺ and 25 mM HEPES (pH 7.4). After incubation in a shaking water bath at 37°C, 140 ml of the cell solution were decanted and filtered through gaze and four layers of mull. Enzymatic activity in the disaggregated supernatant was stopped by centrifugation through 5 ml of 90% FCS for 10 min at 1000g. The pellet was resolved in DMEM/25 mM HEPES/DNase. After a further centrifugation step at 500g for 10 min with 90% Percoll, the pellet was resolved in ice-cold DMEM/25 mM HEPES. Cytotrophoblasts were separated using density gradient centrifugation at 1500g for 45 min on discontinuous Percoll (10-70%). Cells between the 40 and 50% Percoll bands were collected, washed with DMEM, and plated onto 35-mm culture dishes at a density of 5 × 10⁶ cells/dish. Cells were grown in M199 medium, supplemented with 10% FCS, 100 units/ml penicillin/streptomycin, and 5 ng/ml epidermal growth factor at 37°C in a 5% CO₂ humidified atmosphere and later on maintained in culture for 5 days.

RNA Isolation and Analysis. RNA was isolated from 17 term and 14 premature placenta (and four different cytotrophoblasts) isolations using RNeasy Mini extraction kit (QIAGEN GmbH, Hilden, Germany). Villous material of the placenta was separated and stored at -80°C. After mechanical homogenization at 2500 rpm for 2 min, frozen samples were homogenized in guanidinium thiocyanate-containing buffer. The isolation of RNA was performed using the kit according to the manufacturer's instructions. Integrity of RNA was controlled by ethidium bromide staining in a formaldehyde-containing 1% agarose gel.

The isolated RNA was reversely transcribed using random hexamer primers and the TaqMan reverse transcription kit (Applied Biosystems, Darmstadt, Germany). The resulting cDNA was amplified by real-time PCR with intron-spanning primers and probes for human OCTN2, OATP-B, and MRP2. Primer and probe oligonucleotides for OCTN2/PCR were designed based on the cDNA sequence published under GeneBank accession number AB015050. Real-time quantitative PCR for OCTN2 was performed using forward primer 5′-aattttgagatgtttgtcgtgctg-3′, reverse primer 5′-caagaatttctgtccccaggac-3′, and probe 5′-6FAM-tccttgtaggcatgggccagatctcc-3′ and for MRP2, the forward primer 5′-ctgggaacatgattcggaagc-3′, reverse primer 5′-gaggatttcccagagccgac-3′, and probe 5′-6FAM-cagtccgagatgtgaacctggacat-3′ were used. The detection of OATP-B was performed according to St Pierre et al. (2002). For quantification of 18S rRNA, a predeveloped primer and probe mix was purchased from Applied Biosystems. Analysis of OCTN2, OATP-B, and MRP2 was performed using 10-ng reverse-transcribed RNA, although for 18S rRNA, 0.10 ng was used. A TaqMan universal mastermix (Applied Biosystems) was used for real-time PCR reaction. PCR was performed using a real-time PCR cycler ABI Prism 7700 Sequence Detector (Applied Biosystems). For quantification of OCTN2, OATP-B, MRP2, and 18S rRNA, signals were applied to a cloned standard resulting in absolute copy numbers for the respective gene.

OCTN2 Antibody. The anti-OCTN2 antibody was raised in rabbits against the 15 amino acids at the carboxy terminus of the deduced OCTN2 sequence (KDGQERPTILKSTAF). The peptide was synthesized automatically and then coupled to maleimide-activated keyhole limpet hemocyanin (Peptide Specialty Laboratories GmbH, Heidelberg, Germany). New Zealand white rabbits (Charles River, Sülzfeld, Germany) were immunized with this conjugate after the followed protocol. On day 1, 7 mg of immunogen were emulsified with equal volumes of complete Freund′s adjuvants and injected subcutaneously. Boosting with 7 mg of immunogen in incomplete Freund′s adjuvants into the subcutaneous tissue was performed on day 28 and 54, respectively. Serial bleeds were collected from each animal prior the immunization and before new boosting.

Immunofluorescence Microscopy. Localization of OCTN2, OATP-B, and MRP2 was investigated by confocal laser scanning immunofluorescence microscopy using polyclonal antibodies from rabbits raised against the 15 C-terminal amino acids of human OCTN2 and OATP-B. For MRP2, the monoclonal mouse anti-MRP2 antibody M₂III-6 (Alexis Biochemicals, Grünberg, Germany) was used.

After delivery, placenta samples were fixed in formalin, and paraffin sections of 2 μm were generated. The slides were deparaffinized in Xylene substitute for 10 min, and after that, the paraffin sections were hydrated in alcohol at different dilutions for 5 min (100, 98, 75, 50 and 0%). Then the sections were boiled for 10 min in 10 mM citrate buffer (pH 6.0) for OCTN2 and OATP-B staining or in 1 mM EDTA buffer (pH 8.0) for MRP2/OCTN2 double-staining. Sections were washed twice with PBS and blocked for 1 h using 5% FCS in PBS. Incubation with primary antibodies was performed at 4°C overnight. After washing with PBS, the sections were incubated for 2 h with Alexa Fluor 488-labeled anti-rabbit IgG for detection of OCTN2 and OATP-B. For the MRP2/OCTN2 double-staining, the Alexa Fluor 568-labeled anti-rabbit and the Alexa Fluor 488-labeled anti-mouse antibody were used. Staining of nuclei was performed using a 1:2000 dilution of TOTO-3 iodide with DAKO fluorescent mounting medium (DakoCytomation California Inc., Carpinteria, CA). All secondary antibodies, as well as the TOTO dye, were purchased from Molecular Probes (Eugene, OR). For the OCTN2 peptide competition, 50 μl of 1:500 diluted antiserum were preincubated with 100 μg of OCTN2 peptide overnight at 4°C. After that, sections were processed as described above. The fluorescence was detected by confocal laser scanning microscopy.

Preparation of Trophoblastic Vesicles from Human Placental Tissue. A term placenta of normal delivery was used for preparation of basal and apical vesicles according to the methods described previously (Booth et al., 1980; Kelley et al., 1983). In brief, placental cotelydons were divided in two different fractions, the maternal localized part with the higher proportion of extravillous cytotrophoblasts and the fetal localized with multinuclear syncytiotrophoblasts. After several washing steps with cold PBS, 20 to 100 g of the placental tissue were minced and then washed again. Thereafter, the tissue was homogenized in incubation buffer (250 mM sucrose and 10 mM Tris/HCl, pH 7.4) supplemented with protease inhibitors (0.1 mM phenylmethylsulfonyl fluoride, 0.3 μM aprotinin, and 1 μM pepstatin) using a Potter homogenizer (20 strokes, 1000 rpm). After incubation for 1 h on ice, the homogenate was centrifuged at 9000g for 10 min. Supernatants were centrifuged (100,000g, 35 min) and the pellets were resuspended in incubation buffer and homogenized by 15 strokes with a loose-fitting Dounce B homogenizer. For separation of basal and apical membranes, MgCl₂ was added to a final concentration of 10 mM. After 10 min of incubation on ice, a centrifugation step (2200g, 12 min) was performed, separating the basal (pellet) and apical (supernatant) membranes. Pellets were resuspended in homogenization buffer and both fractions were homogenized using a tight-fitting Dounce B potter (30 strokes), washed by centrifugation (100,000g, 35 min), and homogenized again using the tight-fitting Dounce B potter (30 strokes). After centrifugation (100,000g, 35 min), the pellets of both membrane fractions were resuspended in 1 to 3 ml of homogenization buffer, and the membrane suspensions were passed through a 27-gauge needle 20 times for vesicle formation. Membrane vesicles were frozen and stored in liquid nitrogen.

To assess purity of the membrane fractions, the activity of the alkaline phosphatase as marker for the apical membrane was assayed according to Pekarthy et al. (1972). To determine the orientation and integrity of the vesicles, the activity of the ectoenzyme nucleotide pyrophosphatase was measured in the presence and absence of detergent as described earlier (Bohme et al., 1994).

Immunoblot Analysis. Crude membrane vesicle fractions were loaded onto a 10% sodium dodecyl sulfate-polyacrylamide gel after incubation in sample buffer at 95°C for 10 min. Immunoblotting was performed using a tank-blotting system (Bio-Rad, Hercules, CA) and an enhanced chemiluminescence detection system (Amersham Biosciences Inc., Freiburg, Germany). Primary antibodies were diluted in Tris-buffered saline containing 0.05% Tween 20 and 5% bovine serum albumin to the following final concentrations: OCTN2 antiserum 1:1000, MRP2 monoclonal antibody 1:500 (Alexis Biochemicals), and OATP-B affinity-purified antiserum 1:2000. Secondary horseradish peroxidase-conjugated goat anti-rabbit or goat anti-mouse IgG antibodies (Bio-Rad) were used at a 1:2000 dilution.

Vesicle Transport Studies. Vesicles prepared from basal and apical membrane fractions were used for measuring the transport of [³H]carnitine (specific activity: 2.96 TBq/mmol; Hartmann Analytic, Braunschweig, Germany). Transport studies were performed using a rapid filtration through nitrocellulose filters as described previously (Jedlitschky et al., 2000). Total vesicular protein (1 μg/μl) was incubated in the presence of 100 nM [³H]carnitine using an incubation buffer containing 180 mM sucrose, 100 mM NaCl, 10 mM MgCl₂, 10 mM Tris/HCl, and 0.2 mM CaCl₂ in a total volume of 100 μl. In control experiments, NaCl was replaced by KCl. Aliquots of 20 μl were taken after 0.5, 1, 2.5, 5, and 10 min for time dependence studies, and reaction was stopped by adding 1 ml of ice-cold KCl containing incubation buffer. For kinetic studies, 20 μg of vesicular probes were incubated with different concentrations of unlabeled carnitine (1 to 100 μM) and 250 nM [³H]carnitine. For pH studies, the incubation buffer was adjusted by using NaOH or HCl. Samples were filtered immediately through nitrocellulose filters (0.2-μm pore size, presoaked in incubation buffer) and rinsed three times with 3-ml ice-cold KCl containing incubation buffer. For detection of radioactivity, filters were dissolved in 10 ml of scintillation cocktail (ROTISZINT; Roth, Karlsruhe, Germany) and measured in a scintillation β-counter (type 1409; Amersham Biosciences, Uppsala, Sweden). Rates of carnitine transport were given in femtomoles or picomoles of [³H]carnitine × mg protein^-1 × min^-1 or as ratio of control.

Results

OCTN2 mRNA Expression. The expression of OCTN2 was analyzed by quantitative real-time PCR in 36 samples from preterm (gestational age under 37 weeks) and term placentas. OCTN2 values were normalized to expression of 18S rRNA.

OCTN2 mRNA levels were detectable in all 36 samples with copy numbers between 1000 and 10,000. The mRNA expression normalized to 18S rRNA in preterm placentas was around 15% higher compared with term placentas (term, 1.64 × 10^-5; preterm, 1.94 × 10^-5); however, this difference was not significant. Furthermore, mRNA expression for the membrane transport proteins OATP-B and MRP2 were determined and normalized to 18S rRNA. Compared with the mRNA expression of OCTN2, expression of MRP2 was about 60% of the OCTN2. In contrast, mRNA levels of OATP-B were more than 6-fold higher compared with OCTN2 (data not shown).

Immunolocalization of OCTN2. From immunofluorescence the OCTN2 protein appears to be expressed at the syncytiotrophoblast membrane. To further characterize the exact localization of OCTN2 in syncytiotrophoblasts, placenta sections were stained with antibodies against OCTN2 (Fig. 1D), OATP-B as a basal marker (Fig. 1E), and MRP2 as an apical marker (Fig. 1F). Although staining of OATP-B could be assigned to the basal membrane, the double-staining of OCTN2 and MRP2 revealed a colocalization in the apical membrane (yellow fluorescence). Specificity of the OCTN2 antibody was demonstrated by incubating the sections with preimmune serum and antiserum preincubated with OCTN2 peptide (Fig. 1, A-C).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Immunofluorescence microscopy of OCTN2 in human placenta sections. A to C, 20× objective. A, OCTN2 antiserum (dilution: 1:500); B, preimmune serum (dilution 1:500); C, OCTN2 antiserum (dilution 1:500) preincubated with OCTN2 peptide (100 μg). D to F, 40× objective. D, staining of OCTN2 (dilution of antiserum 1:200); E, staining of OATP-B (dilution of affinity-purified antiserum 1:100); F, double-staining of OCTN2 (red fluorescence) and MRP2 (green fluorescence). A to F, staining of cell nuclei with TOTO-3 iodide (blue fluorescence; dilution 1:1000).

Transport of l-Carnitine into Trophoblastic Membrane Vesicles. To demonstrate functional activity of OCTN2 in human placenta, transport of tritium-labeled l-carnitine was measured for apical and basal vesicle preparation (Fig. 2). Membrane vesicles enriched in basal or apical membranes were prepared from term placentas and characterized for markers as described under Materials and Methods. Immunoblots with 25 μg of protein were performed and the separation of membranes further evaluated by detection of OATP-B and MRP2 protein. As expected from previous reports (St Pierre et al., 2000, 2002), the OATP-B protein could be detected in basal and apical preparation with a significantly stronger signal for the basal fraction, whereas the MRP2 protein was detected mainly in the apical fraction of membrane vesicles (Fig. 2A). Furthermore, the activity of the alkaline phosphatase as a marker enzyme for the apical membrane was determined. The results show an enrichment factor of 27.5 ± 11.5 for the apical fraction and 6.9 ± 0.8 for the basal one (compared with the homogenate in both cases). To assess the orientation and integrity of the vesicles, the activity of the ectoenzyme nucleotide pyrophosphatase was measured. The results show, for the basal fraction, 45 ± 2.5% “right side out” vesicles and 57 ± 5.5% for the apical one (data represent the mean value and S.D. for three preparations with triplicate determination). The Western blot analysis of OCTN2 using immunoblots with 25 μg of each vesicular protein fraction showed a significantly stronger signal for the apical fraction compared with the basal one (Fig. 2A).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

A, Western blot analysis for OCTN2, OATP-B, and MRP2. Vesicles (25 μg) were blotted on a nitrocellulose membrane and detected with rabbit OCTN2 and OATP-B polyclonal antisera (dilutions 1:1000 and 1:2000) and a monoclonal mouse antibody against MRP2 (dilution 1:500). B, time-dependent uptake of carnitine into apical (○) and basal (▪) syncytiotrophoblast membrane vesicle. Vesicle fractions from a term placenta were incubated with 100 nM tritium-labeled carnitine at 37°C. At indicated times, uptake was stopped by adding ice-cold stopping buffer, vesicles were isolated using the rapid filtration method, and radioactivity was determined (n = 3, data shown as mean ± S.D.).

The corresponding carnitine transport studies into the apical fraction show a time-dependent uptake with a maximum uptake after 5 min, whereas the basal fraction showed only little uptake of carnitine (Fig. 2B). In comparison, the maximum uptake into the apical vesicles was about eight times higher compared with basal vesicles.

Moreover, we investigated sodium dependence of carnitine uptake into both fractions. Although carnitine uptake into vesicles of the apical fraction was significantly decreased in the absence of sodium, transport of basal vesicle was affected to a minor extent (Fig. 3A). Carnitine uptake to apical vesicles was inhibited at pH 6.4, whereas higher values up to pH 8.4 had only a minor influence compared with physiological pH (Fig. 3B). Verapamil, a known inhibitor of OCTN2, also showed an inhibitory effect on carnitine uptake of about 70% for 50 μM verapamil. To determine the kinetic parameters of carnitine uptake, an apical vesicle preparation was incubated with different concentrations of unlabeled carnitine and 250 nM tritium-labeled carnitine. Total transport of carnitine was calculated for each concentration, and kinetic parameters were calculated from double reciprocal plot according to Lineweaver-Burk (Fig. 4). Calculated values were 21 μM for K_m and 112 pmol × min^-1 × mg^-1 protein for V_max.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Sodium and pH dependence of carnitine uptake. A, sodium-dependent uptake of carnitine into apical and basal syncytiotrophoblast membrane vesicle. Vesicle fractions from a term placenta were incubated with 100 nM tritium-labeled carnitine at 37°C in the presence of 100 mM NaCl (closed bars) or KCl (open bars). Uptake was stopped by adding ice-cold stopping buffer after 1.5 min, vesicles were isolated using the rapid filtration method, and radioactivity was determined (mean ± S.D. for n = 3; ^*, significant difference with p < 0.05, Student's t test; ns, not significant). B, pH-dependence of carnitine uptake into apical placental membrane vesicles. Uptake of 100 nM tritium-labeled l-carnitine was measured at different extravesicular pH. Data given as mean + S.D. of percent of transport at physiological pH (7.4) (n = 3; ^*, significant difference with p < 0.05, Student's t test).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Kinetic carnitine uptake into apical membrane vesicles. Vesicles from a term placenta were incubated for 3 min with different carnitine concentrations and 250 nM tritium-labeled carnitine. The kinetic constants were determined from direct curve-fitting to Michaelis-Menten (A) and from double-reciprocal plot according to Lineweaver-Burk (B) (data shown as mean ± S.D. for n = 3).

Expression of OCTN2 in Differentiating Cytotrophoblasts. Cytotrophoblasts were isolated and cultured in epidermal growth factor-containing medium. After 2, 4, and 6 days in culture, RNA from differentiated cells was isolated, and OCTN2 mRNA levels were determined. The expression of OCTN2 is increasing until day 4 and then regressing at day 6 (151 ± 17, 206 ± 82, 124 ± 39% of the 0 h value) (Fig. 5). As a biochemical marker for differentiation, the hCG production of the cultured cytotrophoblasts was assessed in the supernatant showing maximal concentration (above 14,000 U/l) at the 4th day of culture (Fig. 5).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Expression of OCTN2 mRNA (bars) and concentration of secreted hCG (curve) in supernatant in cultured trophoblasts. Data were given as ratio of OCTN2/18SrRNA expression + S.D. (for 3 independent trophoblast isolations).

Discussion

In this study, we examine the expression and function of OCTN2 in human term and preterm placentas. As shown previously, the expression of a carnitine-transporting protein in human placenta may play an important role in the carnitine supply of the growing fetus. For this reason, a directed carnitine transport from the maternal to the fetal circulation is assumed to be pivotal for the fetus especially in early gestation. Indications for such a sodium-dependent carnitine transport across the apical syncytiotrophoblast membrane have been described (Prasad et al., 1996; Roque et al., 1996). Our results suggest that OCTN2 is a candidate protein mediating this transport.

We have detected significant levels of OCTN2 mRNA in placenta; the OCTN2/18S RNA ratios are in the same order of magnitude as that for MRP2. There was, however, no significant difference between term and preterm placentas indicating a role in fetal supply with carnitine during the entire pregnancy.

Although the presence of OCTN2 mRNA in placenta has already been described (Wu et al., 1998), little was known about its protein expression, localization, and function. To examine expression and localization of OCTN2 protein, an anti-rabbit antibody against the 15 C-terminal amino acids of the human OCTN2 has been generated. The antibody displayed a well defined signal in immunofluorescence. The specificity of this signal could be demonstrated by the comparison of placenta sections incubated with the antiserum and the preimmune serum. Furthermore, specificity of the antibody has been verified by peptide competition experiments.

In the placenta, the immunofluorescence signal is localized in the syncytiotrophoblast membrane and seems to be restricted to the apical membrane. This observation has been verified by double-staining with antibodies against OATP-B and MRP2, which are known to be localized in the basal and apical membrane, respectively (St Pierre et al., 2000, 2002) and by Western blot analysis (Fig. 2A).

In addition, we investigated the function of placental OCTN2. OCTN2 transports carnitine as well as TEA. Although the carnitine transport is sodium-dependent, TEA is transported in a sodium-independent manner. Ohashi et al. (2001) reported that TEA uptake is enhanced by extracellular carnitine and that intracellular TEA is enforcing carnitine uptake. The authors suggest that after each carnitine transport the OCTN2 protein has to be reactivated for further uptake by transport of one molecule of TEA (Ohashi et al., 2001). We separated syncytiotrophoblast membranes in apical and basal fractions. The quality of membrane purification was tested by Western blot analysis for OATP-B and MRP2 resulting in the expected bands in the respective fractions. We then measured carnitine uptake in vesicles generated from these fractions. The apical vesicles showed a significant uptake of carnitine, which was highly sodium-dependent. When replacing sodium by potassium, the carnitine uptake of the apical fraction was reduced to the level of the basal fraction. Further experiments showed that the observed carnitine uptake into the apical fraction exhibited kinetic parameters similar to recombinant OCTN2 and was inhibited by verapamil and unlabeled carnitine, an effect that has also been described for OCTN2-mediated carnitine uptake (Tamai et al., 1998; Wu et al., 1999; Ohashi et al., 2001). Furthermore, the apical fraction showed a pH-dependent decrease in carnitine uptake which is already known for organic cation transporter (Kekuda et al., 1998; Wu et al., 1998).

Although there are other transport proteins like OCTN1 and OCTN3 that transport carnitine with lower affinity than OCTN2 and, in the case of OCTN3, in a sodium-independent manner (Tein, 2003), our data concerning the protein expression and transport characteristics lead to the conclusion that OCTN2 is the major player in carnitine transport from maternal to fetal circulation. The localization of OCTN2 in the apical membrane of the syncytiotrophoblast suggests a dominant role for carnitine uptake from maternal circulation. It remains to be elucidated how carnitine crosses the basal membrane of syncytiotrophoblast.

Carnitine is pivotal for the fetal organism because it seems to be an important stimulator of the mitochondrial respiratory chain (Huertas et al., 1992) and production of surfactant (Lohninger et al., 1990). Another issue is its potential role in energy metabolism. After delivery, the energy supply changes from oxidative metabolism of carbohydrates to β-oxidation of fatty acids as the main energy source (Arenas et al., 1998). For this process, carnitine is an essential cofactor because it facilitates the transport of long-chain fatty acids across the inner mitochondrial membrane (Bremer, 1983). For this reason and because of the limited capacity especially of the early fetus for carnitine biosynthesis, transplacental carnitine transport is essential for the fetal organism and after that for the first days after delivery (Nakano et al., 1989). Because of the described function and localization of OCTN2, this transporter can play an important role in transplacental carnitine transport. Therefore, malfunction caused by modified expression of OCTN2 or inhibiting pharmacological argents may result in carnitine deficiency of the fetus and the newborn child similar to the systemic carnitine deficiency syndrome described for the skeletal and heart muscles (Vaz et al., 1999; Amat Di San et al., 2003).

We did not observe a significant difference between the OCTN2 mRNA level in term and preterm placentas, which indicates functional relevance of OCTN2 already for the early fetus. We could, however, show an increased expression of OCTN2 mRNA in cultured trophoblast cells, which form a multinuclear syncytium after several days in culture (Richards et al., 1994). This increase was paralleled with the hCG secretion in supernatant of cells as a biochemical marker for differentiation (Richards et al., 1994; Malek et al., 2001) and indicates a functional expression of placental OCTN2 regulated by the cytotrophoblast. These cytotrophoblast cells are progenitor cells of the syncytiotrophoblast, which is assumed to play a major role in the control of placental transfer of various substances (Young et al., 2003).

Taken together, we show an apical expression of OCTN2 in syncytiotrophoblast membrane. Carnitine is transported across this membrane with characteristics similar to those described for OCTN2. Very recently, a study on the same topic was published by Lahjouji et al. (2004). The results of that study concerning carnitine transport and OCTN2 expression in placenta were in line to ours. Therefore, we conclude that OCTN2 can play an important role in placental carnitine transport.