The role of fat metabolism during human pregnancy and in placental growth and function is poorly understood. Mitochondrial fatty acid oxidation disorders in an affected fetus are associated with maternal diseases of pregnancy, including preeclampsia, acute fatty liver of pregnancy, and the hemolysis, elevated liver enzymes, and low platelets syndrome called HELLP. We have investigated the developmental expression and activity of six fatty acid β-oxidation enzymes at various gestational-age human placentas. Placental specimens exhibited abundant expression of all six enzymes, as assessed by immunohistochemical and immunoblot analyses, with greater staining in syncytiotrophoblasts compared with other placental cell types. β-Oxidation enzyme activities in placental tissues were higher early in gestation and lower near term. Trophoblast cells in culture oxidized tritium-labeled palmitate and myristate in substantial amounts, indicating that the human placenta utilizes fatty acids as a significant metabolic fuel. Thus human placenta derives energy from fatty acid oxidation, providing a potential explanation for the association of fetal fatty acid oxidation disorders with maternal liver diseases in pregnancy.
- acute fatty liver of pregnancy
- hemolysis, elevated liver enzymes, and low platelets syndrome
because the placenta provides the fetus with nutrients needed for growth and serves as an excretory organ to eliminate wastes from fetal metabolism, placental pathology profoundly affects the developing fetus. The placenta grows exponentially during gestation, from an average of 6 g at 3 wk of gestation to ∼470 g at term. Moreover, the villous surface increases from 830 cm2 at 3 wk of gestation to ∼125,000 cm2 at term, and the maternal-fetal diffusion distance decreases from 55 to 4.8 mm (16). The placenta requires a constant and abundant source of energy to supply the needs for its own rapid growth and maturation and to transport the nutrients, ions, vitamins, waste, and other molecules required for fetal growth and homeostasis from the maternal to the fetal circulation and vice versa.
A common belief among fetal physiologists (14, 16, 18, 20,26) is that glucose transported to and across the placenta from the maternal circulation provides all placental and fetal energy needs via glycolysis and the citric acid cycle. Because this supply of glucose is constant, consistent, and reliable, it has been suggested that the placenta and fetus do not need to regulate energy-producing metabolic pathways. The focus of most research has been on transplacental passage of nutrients, including both amino acids and fatty acids, but the metabolic fuel required by the human placenta has not been determined conclusively. The presence of multiple glucose transporters and enzymes of glycolysis and the citric acid cycle in the placenta is consistent with glucose being a major energy source (14). It has been postulated that adequate glucose supply, conversion of acetyl-CoA to malonyl-CoA by acetyl-CoA carboxylase, and the resultant inhibition of carnitine palmitoyltransferase I (CPT I) (26) inhibit fatty acid uptake and oxidation by placental mitochondria in utero (20). Although fatty acids are actively transported across the placenta to the fetus, there are scant data to assess the role of lipids as a metabolic fuel for placental growth and development (18).
Fatty acid oxidation (FAO) defects are autosomal recessive and potentially fatal disorders that are now diagnosed with increased frequency in the perinatal and infantile periods. Uniquely among inherited metabolic defects, FAO enzyme disorders in the affected fetus may cause significant maternal morbidity and mortality (11, 13,15, 23, 27, 32, 33, 39). We (11, 13, 23) and others (15, 27, 32, 33, 39) have recently shown that maternal acute fatty liver of pregnancy (AFLP), the HELLP syndrome (hemolysis, elevated liver enzymes, and low platelets), placental floor infarction, and preeclampsia are associated with defects in FAO in the fetus. The majority of neonates born after such pregnancies are premature, exhibit growth restriction, and present in the newborn period or early infancy with fasting-induced hypoketotic hypoglycemia and hepatic encephalopathy, which may progress to coma and death (11, 33).
Because FAO disorders in the fetus are associated with maternal complications and because the human placenta is mostly of fetal origin, we hypothesized that energy supplied from fatty acid β-oxidation in the placenta could be an important metabolic energy source for survival, growth, and function in both the placenta and the fetus. As a corollary, if FAO were active in the placenta and because late-gestation placenta is of fetal origin, fetal defects in this pathway would generate long-chain fatty acids that could enter the maternal circulation in levels toxic to the mother. We report here the developmental expression and activity of six different FAO enzymes of the mitochondrial β-oxidation spiral in human placenta (Fig.1): medium-chain acyl-CoA dehydrogenase (MCAD), long-chain acyl-CoA dehydrogenase (LCAD), very-long-chain acyl-CoA dehydrogenase (VLCAD), short-chainl-3-hydroxyacyl-CoA dehydrogenase (SCHAD), long-chainl-3-hydroxyacyl-CoA dehydrogenase (LCHAD), and long-chain 3-ketoacyl-CoA thiolase (LKAT). We used immunohistochemistry, Western blot analysis, enzyme activity measurements, and metabolic flux studies to demonstrate substantial FAO expression and function of these enzymes in the human placenta, consistent with our hypotheses.
Placental tissues and cells.
This study was approved by the Human Studies Committee of Washington University School of Medicine. Placental specimens were collected at gestational ages ranging from 12 to 43 wk. Tissue was placed in chilled, buffered saline to remove maternal blood and then snap-frozen in liquid nitrogen and stored at −80°C or fixed in 10% neutral buffered formalin solution at 4°C for 24 h before being processed for paraffin embedding and immunohistochemical studies. Twenty-eight placental specimens were used for enzyme activity and Western blot studies. Specimens were divided for analysis into the following gestational age groups: 12–19 (n = 3), 20–28 (n = 3), 29–34 (n = 5), 35–37 (n = 4), 38 wk (n = 9), and >40 wk (n = 4). All samples were utilized to measure enzyme activity, and nine of these samples from 12 to 43 wk of gestation were used to run Western blots for the six FAO enzymes.
For enzyme activity and metabolic flux studies, primary human cytotrophoblasts were isolated from normal-term human placentas (n = 4) by use of the trypsin-DNAse-dispase/Percoll method, as described (25). Cultures were plated at a density of 350,000 cells/cm2 and maintained in Earle's medium 199 (M199) containing fetal bovine serum, 20 mM HEPES, pH 7.4, 0.5 mM l-glutamine, penicillin (10 U/ml), streptomycin (10 mg/ml), and fungizone (0.25 mg/ml). All cultures were maintained at 37°C in a 5% CO2 atmosphere, and the medium was changed every 24 h. Our trophoblast cell isolates were 95–97% cytokeratin positive, which indicates that they were trophoblast cells (25).
Term placentas from uncomplicated pregnancies were collected, and 5-μm-thick sections of paraffin-embedded tissue were cut, applied to glass slides, deparaffinized in xylene, and rehydrated in an ethanol gradient. Endogenous peroxidase activity was quenched by incubating the specimens in 3% H2O2 in methanol for 30 min. After equilibration for 5 min in distilled water, the samples were subjected to heat antigen retrieval using citrate buffer (pH 6.0). The samples were heated at maximum power in a microwave for 5 min, cooled for 5 min, reheated for 5 min, and allowed to stand at room temperature for 20 min.
The slides were then washed and blocked using an avidin-biotin blocking kit (Vector Labs, Burlingame, CA) for 30 min followed by a blocking buffer (NEN-Life Sciences, Boston, MA) for 30 min. The blocking buffer was removed, and the sections were exposed to primary rabbit polyclonal antisera raised against one of the following enzymes at the indicated dilution: MCAD (1:200), LCAD (1:400), VLCAD (1:200), SCHAD (1:200), LCHAD (1:400), LKAT (1:400), or human β-chorionic gonadotropin (β-HCG; Dako, Carpinteria, CA). The primary antibody was applied with 0.3% Triton X-100 in phosphate-buffered saline (PBS) overnight at 4°C. After two washes with PBS the next day, secondary goat anti-rabbit biotinylated antibody (NEN-Life Sciences) was applied at a concentration of 1:800 for 1 h at room temperature. The tertiary reagent was streptavidin-horseradish peroxidase (Dako) at a concentration of 1:1,000 for 1 h at room temperature followed by application of 3,3-diaminobenzidine substrate for 1–5 min. The slides were rinsed, counterstained with Mayer's hematoxylin, dehydrated in ethanol, cleared with xylene, and mounted with glass coverslips using Histomount (Zymed Laboratories). Two to five sets of placental tissue from term pregnancies were stained for all six FAO enzymes.
Western blot analyses.
Placental tissue freed of maternal blood (100–250 mg) was lysed in a buffer containing 0.1 M sodium phosphate, 0.5 mM EDTA, and 0.5% Triton with protease inhibitors by means of a polytron. The placental lysate was sonicated three times for 10 s each on ice. The lysates were subjected to centrifugation at 3,000 g for 5 min, and the protein concentration of the supernatant was measured by the Bradford method. Fifty micrograms of protein were analyzed by immunoblotting with rabbit polyclonal antisera raised against one of the six different FAO spiral enzymes at the following dilutions: MCAD (1:1,000), LCAD (1:5,000), VLCAD (1:500), SCHAD (1:5,000), LCHAD (1:3,000), and LKAT (1:2,000). Incubation with secondary antibody (goat anti-rabbit, 1:1,000 dilution) and visualization with diaminobenzidine reagent were done until the protein bands were visible. Two to five blots were prepared for each enzyme, and a representative immunoblot was analyzed with an AlphaImager 3400 (Alpha Innotech, San Leandro, CA) using its AlphaEase image analysis software for densitometry. Densitometry data were subjected to statistical analysis to determine any relationship of gestational age to antigen expression.
Enzyme kinetics and metabolic flux studies.
The activities of SCHAD, LCHAD, and LKAT in placental homogenates were measured as described (3, 35, 38). Trophoblast cells were harvested from four uncomplicated term pregnancies, and three 75-cm2 flasks from each placenta were used for experiments. Metabolic flux studies were performed using tritiated water released from [9,10-3H]palmitate and [9,10-3H]myristate in 24-well microplates (22,24). The cells were grown for 20 h while being incubated with 22 μM [9,10-3H]palmitate or 110 μM [9,10-3H]myristate, respectively. Each experiment was run in triplicate, and in all assays, palmitic or myristic acids were complexed with defatted bovine serum albumin (0.45 mg/ml). By use of these tritium-labeled substrates, where 3H is distributed equally between two adjacent carbon atoms, 75–100% of the label is converted to 3H2O during complete β-oxidation cycle in intact cytotrophoblast cells (22, 24,36).
We made three enzyme activity measurements from each sample and compared a total of 28 samples with one another after categorizing them into gestational age groups; we used multivariate analysis of variance (MANOVA) with calculation of Wilk's λ and P values. Data were later subjected to Student-Newman-Keuls post hoc analysis for multiple comparisons using statistical software SPSS for PC, version 11.01.
We carried out univariate and multivariate regression analyses on densitometric data for the six FAO enzymes to assess the significance of regression, i.e., the relationship between gestational age and densitometric readings. The slope of regression was compared with zero; an R 2 value, F statistic, and significance (P) value were calculated for each slope. Data are presented as means ± SD unless stated otherwise; statistical significance was set at P < 0.05.
Figure 1 shows the metabolic pathway of entry of fatty acids into the cell and their breakdown through the mitochondrial β-oxidation spiral. Among the six enzymes of the spiral studied here, the highly homologous enzymes MCAD, LCAD, and VLCAD catalyze the first step by using substrates of differing chain lengths. SCHAD and LCHAD, which are also highly homologous, catalyze the second and third reaction of the spiral, and LKAT performs the final cleavage step in this pathway.
We examined expression of these six enzymes by immunohistochemistry in 2–5 term human placentas (Fig. 2). Minimal nonspecific background staining was observed in control sections processed without primary antibody (Fig. 2 A). After incubation with β-HCG antibody as a positive control (Fig.2 B), intense staining of the syncytiotrophoblast layer was observed and, as expected, there was no reaction in cells from the rest of chorionic villi. Figure 2, C-H, shows immunoreactivity for MCAD, SCHAD, LKAT, VLCAD, LCAD and LCHAD, respectively. The intensity of staining for all six enzymes was highest in syncytiotrophoblast and similar to that of β-HCG, suggesting abundant expression of all FAO enzymes. No specific immunoreactivity for any FAO enzyme was detected in the chorionic villous vessels or connective tissue. Although the highest levels of FAO enzyme immunoreactivity were in the syncytiotrophoblast layer of the chorionic villi, there was also detectable expression of all FAO enzymes in the villous cytotrophoblasts. These results show abundant and cell type-specific expression of the FAO enzymes in both syncytiotrophoblast and cytotrophoblast but not in core cells of the villi.
Expression and activity of FAO enzymes during human placental development.
Figure 3 is a composite immunoblot to analyze expression of the six FAO enzyme antigens in placental villi by use of specimens from between 12 and 43 wk of gestation. Densitometric analysis of immunoblots indicated that expression of LCHAD, VLCAD, and SCHAD was 2- to 2.5-fold higher at the lower gestational ages of 12 and 17 wk compared with term placenta. LKAT expression was four- and threefold higher at 12 and 17 wk of gestation, respectively, compared with term placenta. Our multivariate regression analysis found an inverse correlation between expression of LCHAD, SCHAD, VLCAD, and LKAT with gestational age (R 2 = 0.65, 0.75, 0.51, and 0.73 for LCHAD, SCHAD, VLCAD, and LKAT, respectively). The slope of regression was significantly different from zero for LCHAD, SCHAD, VLCAD, and LKAT, with P values of 0.008, 0.002, 0.02, and 0.002, respectively. There was no measurable difference among various gestational ages for MCAD and LCAD expression. Figure4 shows LCHAD, SCHAD, and LKAT enzyme activities measured from extracts of 28 placental samples. The specific activities (nmol · min−1 · mg tissue−1) were significantly higher for all three enzymes at lesser gestational ages (12–28 wk) compared with term placentas with MANOVA, with Wilk's λ of 0.290 and P value of 0.02. Groupwise comparisons using the Student-Newman-Keuls test showed a significantly higher activity of LCHAD and LKAT at 12–28 wk of gestation compared with term and postterm placentas and a significantly higher activity of SCHAD at 12–19 wk compared with term placentas. These results show that enzymatic activities are regulated during the course of placental development.
Tissue-specific activities of FAO enzymes.
We compared the enzyme activities of LCHAD, SCHAD, LKAT, and CPT II in fresh, crude, placental tissue extracts with our previously published data of activities in fresh human liver and skeletal muscle extracts and from cultured fibroblasts from normal individuals (2, 3,38). The enzyme activities in placenta were two- to fivefold less than in liver throughout gestation. Compared with skeletal muscle, both LCHAD and LKAT, components of trifunctional protein that utilize long-chain substrates, were similar or higher in placental extracts from 12- to 19-wk-gestation pregnancies. FAO enzyme activities in crude placental extract were comparable to those in cultured human fibroblasts, but freshly isolated cytotrophoblasts from term pregnancies had substantially greater (2- to 8-fold) LCHAD, SCHAD, and LKAT activities than cultured fibroblasts. The enzyme activities in term trophoblast cells in primary culture were two- to threefold greater than those in fresh placental tissue from term pregnancies. These data show that placenta contains levels of FAO enzymes comparable to those present in mature, fatty acid-dependent tissues such as skeletal muscle, especially between 12 and 19 wk of gestation, and that cytotrophoblasts contain long-chain activities (LKAT and LCHAD) comparable to those of liver.
Fatty acid flux in term human cytotrophoblast cells in culture.
Table 1 shows the results of metabolic flux studies in cytotrophoblasts with tritiated palmitic acid or myristic acid as substrates. Primary trophoblast cells cultured for 36–48 h used the two fatty acids as metabolic fuel, with overall oxidation of fatty acids 30–100% greater than fibroblasts in culture (P < 0.05).
Our results show that expression and activities of FAO enzymes as well as overall FAO of palmitate and myristate are substantial in human placental villi during gestation and identify trophoblast components as the primary FAO sites in the term placenta. Expression of FAO enzymes and measured activities at lower gestational ages of 12–28 wk are comparable to that in mature slow skeletal muscle, a tissue that uses fatty acids as a substrate to satisfy high energy requirements. The data provide direct evidence that fatty acids undergo extensive mitochondrial β-oxidation in the placenta. Our major conclusion is that fatty acids are utilized as a significant metabolic fuel and energy source in this organ, consistent with our hypothesis. This overall conclusion is novel and at variance with the current concept that glucose is the sole energy source in the placenta (14,16). On the basis of this conclusion, we speculate that human placental mitochondrial FAO is critical for normal growth and maturation of the placenta and for fueling the energy-consuming functions of ion, nutrient, and waste transplacental transport.
Our study's limitations include a lack of correlation between our in vitro data and in vivo FAO in the fetal-placental unit and some discrepancy between our Western blot and enzyme activity data. We are also limited by the fact that we used term normal placenta to extract trophoblast cells, and our metabolic flux data cannot be applied directly to trophoblast cells from earlier gestations. Thus our in vitro data need to be correlated with in vivo experiments, preferably in human subjects. FAO has been studied in humans by use of13C-labeled stable isotope technology, and a similar study model can be applied to pregnant women to elucidate our hypothesis more conclusively.
Our metabolic flux data for palmitate and myristate in Table 1 have high standard deviations, indicating variability in results. Variability among primary cultures is not uncommon, and trophoblast cells are no exception. Our comparison between fibroblasts, a well-studied approach to quantification of fatty acid oxidation in humans with suspected FAO disorders, and trophoblasts showed no statistical difference; the major conclusion from these studies is that trophoblasts indeed oxidize fatty acids. Because all cell lines were cultured in media containing glucose and not in physiological conditions present in vivo during fasting, we have shown that significant metabolic flux through the FAO pathway occurs even when glucose is the available energy substrate source. Our previous data in fibroblasts show that, when fatty acids are supplied as substrate, expression of FAO enzymes and flux through the β-oxidation pathway increases. Thus the flux measurement presented in Table 1represents minimal values.
Previous studies are consistent with our conclusion that fatty acid uptake and metabolism are prominent in placenta. Lipoprotein lipase is highly expressed on the maternal surface of the syncytiotrophoblast and hydrolyzes maternal plasma triacylglycerol (17, 29). This enzyme activity would make long-chain free fatty acids available for uptake. Maternal triglyceride levels rise two- to threefold in late gestation, thereby increasing availability of fatty acids for uptake and metabolism (10, 17). Fatty acid-binding proteins that are critical for uptake of free long-chain fatty acids are also located on the microvillous membranes of the syncytiotrophoblasts facing the maternal circulation (1, 4, 5). This location favors unidirectional flow of maternal fatty acids into the placenta. Furthermore, the VLDL/apolipoprotein E receptor is positioned on the microvillous surface in human placental trophoblast cells, consistent with a role in placental lipid uptake and transport (40). Perhaps most importantly, carnitine, an essential factor for transfer of long-chain fatty acids from the cytosol to mitochondria for subsequent β-oxidation, is actively transported across the placenta by an organic cation/carnitine, sodium-dependent transporter (OCTN2) that is highly expressed in placental tissue (37). Thus previously reported results, in conjunction with our data, strongly suggest that the human placenta is capable not only of transporting fatty acids to the fetus but also of using them as a metabolic fuel.
Our major conclusion has important implications for understanding the known association of maternal liver diseases of pregnancy, AFLP, HELLP syndrome, placental floor infarction, and preeclampsia, with LCHAD or complete trifunctional protein deficiency in the fetus (11, 13,15, 23, 27, 32, 33, 39). We (11, 13, 23) and others (33, 34) have shown that pregnancies carrying an affected fetus with the missense mutation G1528C in the α-subunit of mitochondrial trifunctional protein, a mutation that occurs in the active site of the LCHAD enzyme, have one of these maternal complications 75% of the time. This mutation is relatively common among individuals of northern European ancestry (11, 33), with a carrier frequency of 1 in 175 in the United States and 1 in 680 in the Netherlands (8). In LCHAD deficiency, accumulation of the long-chain hydroxyacylcarnitines, free plasma hydroxy-long-chain fatty acids, and dicarboxylic acids occurs. These metabolites are cytotoxic because they inhibit mitochondrial FAO enzymes, uncouple oxidative phosphorylation, and impair ATP production (7, 28, 30, 31). Long-chain acylcarnitines are also known to damage isolated canine myocyte sarcolemmal membranes and potentiate free-radical-induced lipid membrane peroxidative injury in ischemia (21).
We previously postulated (11) that long-chain fatty acids and their metabolites cause maternal liver damage in AFLP and other maternal liver diseases. However, we were puzzled, because the source of the postulated toxic metabolites was unclear given that the fetus does not utilize fatty acids for energy production. The data presented here provide a likely explanation of the seemingly conflicting clinical findings, that is, that the LCHAD-deficient fetal-derived placenta may be the source of these harmful metabolites, particularly as the placental mass and energy requirements increase substantially during the third trimester. In addition to direct toxicity, contributing factors to maternal liver disease might be the 50% decrease in maternal hepatic long-chain fatty acid oxidation capacity related to maternal heterozygosity for LCHAD mutations and the increased liberation of fatty acids during the latter half of pregnancy due to increased lipoprotein lipase and increased reliance on fat as an energy source by the mother late in gestation. Thus we suggest that the placental FAO defect causes maternal liver disease in families with LCHAD or trifunctional protein (TFP) mutations.
A second implication of our findings is that placental insufficiency due to lack of energy production in pregnancies with FAO-deficient fetuses may occur. We (11, 13) and others (15, 27,32, 33, 39) have noted that fetal growth restriction and prematurity are common among LCHAD-affected fetuses. In addition, we have shown that ablation of the trifunctional protein α-subunit in mice causes intrauterine growth restriction and perinatal lethality (12). Moreover, in VLCAD- and LCAD-deficient mice, late-gestation prenatal fetal death is common, despite the fact that the fetus does not rely on β-oxidation for energy (Exil VJ, Sims HF, Qin W, Roberts R, Rinaldo P, Zimmerman F, and Strauss AW, unpublished observations; 19). These results are all consistent with the hypothesis we posed: that placental FAO is critical for the health of the fetal-placental-maternal unit.
Other FAO defects may also be associated with maternal liver disease but only rarely. Single case reports of maternal liver diseases occurring during pregnancies with fetuses affected by CPT I and SCAD deficiency exist (15, 28). Three cases of maternal liver disease during pregnancies carrying fetuses with complete trifunctional protein deficiency have also been published (6). However, among families with MCAD deficiency, the commonest FAO disorder, and VLCAD deficiency, maternal liver diseases of pregnancy are extremely rare. This raises the possibility that 3-hydroxy- and other long-chain fatty acids that must accumulate in isolated LCHAD or complete TFP deficiencies are particularly toxic.
A second conclusion from our results is that FAO enzymes are expressed in a cell-specific manner within the placenta and that there is some developmental regulation of expression during gestation. Trophoblast cells from term placenta express key enzymes of the β-oxidation spiral, and expression was higher in the synctiotrophoblast layer than in the cytotrophoblasts (Fig. 2). The syncytiotrophoblast layer of chorionic villi plays an important role in the uptake of lipids, ions, and glucose into the placenta and their transfer to the fetus (17, 29), functions consistent with a large energy requirement. Our measurements revealed modestly higher enzyme activities at lower gestational ages (Fig. 4), emphasizing a key role for FAO early during gestation.
In summary, we have demonstrated the expression and activity of six enzymes involved in the FAO β-oxidation spiral in human placenta, with enzyme expression in a cell-specific manner localized to the syncytiotrophoblast layer, with lesser activity in the cytotrophoblast cells, and with no expression in villous core cells. Fatty acids are used as a major metabolic fuel by human placentas at all gestational ages, and any defect within this energy-producing pathway may hamper the growth, differentiation, and function of the placenta, thereby compromising fetal growth.
Defects of FAO in the fetal-placental unit are associated with accumulation of abnormal metabolic precursors, including hydroxyacylcarnitines and dicarboxylic acids. Such toxic fatty acids are likely transferred to the maternal circulation and may contribute to the pathophysiology of preeclampsia, AFLP, and HELLP syndrome in these families.
We gratefully acknowledge the help of Dr. Beverly Rogers with enzyme kinetics analysis, Karen Hutton with immunohistochemistry, and Prof. James K. Dias from the Medical College of Georgia for statistical data analysis.
This study was supported in part by Grants P30-DK-52574, AM-20407, and DK-02574 from the National Institutes of Health.
Address for reprint requests and other correspondence: P. Shekhawat, Dept. of Pediatrics, Section of Neonatology, BIW 6033E, Medical College of Georgia, Augusta, GA 30912 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published February 11, 2003;10.1152/ajpendo.00481.2002
- Copyright © 2003 the American Physiological Society