Autoantibodies against Folate Receptors in Women with a Pregnancy Complicated by a Neural-Tube Defect
http://www.100md.com
《新英格兰医药杂志》
Background In the absence of clinical folate deficiency, periconceptional supplementation with folic acid reduces a woman's risk of having an infant with a neural-tube defect. Since antiserum to folate receptors induces embryo resorption and malformations in rats, we hypothesized that autoantibodies against folate receptors in women may be associated with pregnancy complicated by a neural-tube defect.
Methods Serum from 12 women who were or had been pregnant with a fetus with a neural-tube defect and from 24 control women (20 with current or prior normal pregnancies and 4 who were nulligravid) was analyzed for autoantibodies by incubation with human placental folate receptors radiolabeled with folic acid. The properties of these autoantibodies were characterized by incubating serum and the autoantibodies isolated from serum with placental membranes, ED27 cells, and KB cells, which express the folate receptors.
Results Serum from 9 of 12 women with a current or previous affected pregnancy (index subjects) and 2 of 20 control subjects contained autoantibodies against folate receptors (P<0.001). The autoantibodies blocked the binding of folic acid to folate receptors on placental membranes and on ED27 and KB cells incubated at 4°C and blocked the uptake of folic acid by KB cells when incubated at 37°C.
Conclusions Serum from women with a pregnancy complicated by a neural-tube defect contains autoantibodies that bind to folate receptors and can block the cellular uptake of folate. Further study is warranted to assess whether the observed association between maternal autoantibodies against folate receptors and neural-tube defects reflects a causal relation.
Neural-tube defects, which include spina bifida, anencephaly, craniorachischisis, and encephalocele, occur in approximately 1 per 1000 births in the United States, and women who have one fetus with this complication are at increased risk in subsequent pregnancies.1 Although periconceptional folic acid supplementation reduces the occurrence and recurrence of neural-tube defects by approximately 70 percent,2,3 most women who are pregnant with a fetus with this complication do not have clinical folate deficiency.4 Though some polymorphisms for folate-pathway enzymes have been identified,5 they cannot account for the 70 percent decrease in the incidence of this birth defect with folate supplementation.
Studies in animals have suggested the importance of folate receptors in embryogenesis.6 Inactivation of both alleles encoding the mouse homologue of the human folate receptor gene was uniformly fatal in embryos with neural-tube defects.7,8 Folinic acid given to the pregnant dams resulted in normal development in 80 percent of the embryos that lacked the folate receptor gene in both alleles.8 Despite considerable research, no specific polymorphisms or mutations of the human folate receptor gene have been identified that might explain the reduction in the incidence of neural-tube defects with folic acid supplementation.9
Our observation that the administration of an antiserum to folate receptors10 to pregnant rats resulted in the resorption of or multiple developmental abnormalities in embryos11 prompted us to speculate that an autoantibody against folate receptors in women could induce similar embryonic and fetal abnormalities. Therefore, we developed an assay to measure these autoantibodies in the serum of women with a current or prior pregnancy complicated by a neural-tube defect and that of women without such a history.
Methods
Subjects
Between April 2002 and August 2003, all 12 women at our institution with a current or prior pregnancy complicated by a neural-tube defect who consented to participate in this study were enrolled and had their serum analyzed for autoantibodies against the folate receptor. The women were attending the obstetrical clinic or were the mothers of children with neural-tube defects who were being followed in the pediatric neurology clinic. Eligible women were between the ages of 18 and 40 years at the time of the affected pregnancy.
Initially, we enrolled 12 women with a current or previous affected pregnancy (index subjects) and 18 control subjects who had never had a pregnancy involving a neural-tube defect, 4 of whom were nulligravid. An initial analysis comparing the frequency of the autoantibodies in the group of index subjects and these 18 controls showed a significant difference (P<0.001).
Subsequently, we decided that women who were pregnant at the time of testing with a fetus that did not have a neural-tube defect would be more appropriate additional controls. Accordingly, the first six women who met this criterion and provided informed consent were also enrolled in the study. Thus, there were 24 controls; 8 were pregnant with an unaffected fetus, 12 had given birth to one or more unaffected infants, and 4 were nulligravid. Because pregnancy was a criterion for study entry, the four nulligravid control women who were initially tested for the autoantibody were excluded from the statistical analysis. Although the number of subjects in each group was not prespecified, every woman who consented to participate was included in the study. The institutional review board approved the study, and written informed consent was obtained from all subjects.
Preparation of Folic Acid–Labeled Purified Folate Receptors
Folate receptors from human placental membranes were solubilized at 37°C for two hours in 0.01 M sodium phosphate buffer, pH 7.4, containing 1 percent Triton X-100, and purified with Sepharose 4B to which folic acid was covalently coupled.12 The matrix was washed with the phosphate buffer, and the receptors were eluted from the matrix with 0.1 M glycine–hydrochloric acid, pH 3.5, containing 0.1 percent Triton X-100. The pH of the eluate was raised to 7.4 by the addition of 0.2 M sodium barbiturate, and the receptors were quantified on the basis of the level of binding of folic acid (specific activity, 31 Ci per millimole; Moravek), which binds at a 1:1 molar ratio.
Radiolabeled receptors were prepared by incubating the purified apo-folate receptors with a 20 percent excess of folic acid for 30 minutes at 4°C. At a pH between 7 and 9, there is virtually no dissociation of folic acid from the receptors incubated overnight at 4°C. The addition of a 10-fold greater concentration of the folate receptors, saturated with nonradioactive folic acid, confirmed the specificity of the folate receptor–autoantibody complex.
Assay of Serum for the Autoantibody against the Folate Receptors
For the assay for the autoantibody against folate receptors, 40 μl of serum was added to 200 μl of 0.1 M glycine buffer, pH 3.0, incubated for 10 minutes at 25°C, and then added to a pellet of dextran-coated charcoal to adsorb free folate and folate that dissociated from soluble receptors. At this pH, soluble receptors will also dissociate from autoantibodies in the serum. The charcoal was pelleted, and the supernatant fraction was added to 200 μl of 0.2 M sodium barbiturate, pH 8.9, containing phenylmethylsulfonyl fluoride, EDTA, 0.1 percent Triton X-100, and the folic acid–folate receptor preparation. Another aliquot of serum was added to a mixture of the folic acid–folate receptors containing a 10-fold excess of the receptors saturated with nonradioactive folic acid.
The volume of each reaction was increased to 600 μl by the addition of glycine–sodium barbiturate buffer (final pH, 8.6), incubated overnight at 4°C, and 600 μl of a 10 percent suspension of staphylococcal protein A membranes (Pansorbin, Calbiochem) was added. The incubation continued for 15 minutes; the membranes were then pelleted and washed, and the level of radioactivity was determined. The level of background radioactivity measured in duplicate reactions that lacked serum was subtracted from the test serum samples.
Autoantibodies and the Binding of Folic Acid to Folate Receptors
Membranes were prepared from human placenta homogenized in 0.01 M sodium phosphate buffer, then pelleted, and washed. The membranes were suspended in 0.1 M acetic acid to dissociate endogenous folate from the receptors, washed twice in this solution, and resuspended in 0.01 M sodium phosphate buffer, pH 7.4. The binding of folic acid was used to determine the quantity of the apo-folate receptors.
The effect of the autoantibodies on the binding of folic acid to the receptors was determined by incubating 200 μl of a 1 percent suspension of the membranes with charcoal-treated serum overnight at 4°C. The membranes were washed with cold phosphate-buffered saline, incubated for 30 minutes at 4°C with folic acid (125 pg) in 1 ml of phosphate-buffered saline, washed with cold buffer, and solubilized with 1 N sodium hydroxide, and the radioactivity was determined.
ED27 cells, a human-placenta cell line,13 and KB cells, a human epidermoid carcinoma-cell line,14 were each plated in triplicate at a density of 20,000 cells in 1.83-cm2 wells containing folate-deficient Dulbecco's minimum essential medium with 10 percent fetal-calf serum and the test serum (pretreated with charcoal to remove free endogenous folate) and incubated overnight at 37°C. The cells were washed with cold Hanks' balanced salt solution (HBSS) and incubated at 4°C for 30 minutes with folic acid (125 pg) in 1 ml of HBSS. The samples were washed with cold HBSS, and the cell-associated radioactivity was determined.
To obtain the apparent association constant (Ka) for the binding of the autoantibodies to the folate receptors, the charcoal-treated serum was incubated overnight at 4°C with placental membranes containing the apo-folate receptors, followed by incubation with folic acid for 30 minutes. The free ligand was removed by pelleting the membranes, and the radioactivity of the membrane-bound folic acid was determined. This value represented the apo-folate receptors that were not blocked by the autoantibodies. Scatchard analysis15 was used to compute the apparent Ka.
Autoantibodies and the Cellular Uptake of Folic Acid
Ten milliliters of serum from Index Subject 8 and from Control Subject 14 (see Supplementary Appendix 1 and Supplementary Appendix 2, available with the full text of this article at www.nejm.org) was incubated overnight at 4°C with 1 ml of packed placental membranes to bind the autoantibodies to the folate receptors. The samples were washed with cold HBSS to remove unbound serum proteins, and autoantibodies bound to the folate receptors were eluted with 0.1 M acetic acid containing 0.1 percent bovine serum albumin. The eluate was dialyzed against HBSS at 4°C overnight and concentrated by vacuum dialysis.
KB cells (20,000) plated in duplicate in 3.5-cm2 culture dishes were incubated overnight at 37°C with the isolated autoantibodies, the similarly treated control serum, or medium alone. The cells were washed with HBSS, medium containing folic acid (125 pg) was added to the wells, and the incubation was continued for 15, 30, and 60 minutes at 37°C, with duplicate sets incubated at 4°C. The cells were washed again with HBSS at 4°C and lysed with 1 N sodium hydroxide, and the radioactivity was determined. The difference in radioactivity between the cells incubated at 4°C and those incubated at 37°C represents the cellular uptake of folic acid.
The analysis of serum from index and control subjects for other autoantibodies was performed by Universal Diagnostic Laboratory with use of enzyme-linked immunosorbent assays (anti-DNA antibodies from Wampole Laboratories; anticardiolipin antibodies from DiaSorin; antithyroid antibodies from Diagnostic Products).
Statistical Analysis
The percentage of women with autoantibodies to the folate receptor in the two groups was compared with the use of Fisher's exact test. Mean binding levels of the folic acid–folate receptor complex in serum were compared between groups with the use of Student's t-test; Satterwaite's correction of the degrees of freedom was applied because of the unequal variances. Clinical variables in the two groups were compared with the use of the Wilcoxon two-sample test and Fisher's exact test. Two-tailed tests were used in all cases; P values of 0.05 or less were considered to indicate statistical significance. SAS software (SAS Institute) was used.
Results
Table 1 shows the demographic and clinical characteristics of the 12 index subjects and the 20 control subjects who had been or were pregnant. There were no significant differences between the groups. Other clinical characteristics of the index and control subjects (including the four nulligravid controls who were not included in the statistical analyses) are provided in Supplementary Appendix 1 and Supplementary Appendix 2 (available with the full text of this article at www.nejm.org).
Table 1. Clinical Characteristics of the Subjects.
Autoantibodies against the folate receptors were identified in the serum of 9 of 12 index subjects (75 percent) — 2 with a current pregnancy complicated by a neural-tube defect (Subjects 4 and 5) and 7 (Subjects 1, 2, 3, 6, 7, 8, and 9) with a history of a pregnancy complicated by a neural-tube defect (Figure 1A) — as compared with 2 of the 20 control subjects (10 percent, P<0.001) (Figure 1B). The mean value for the folic acid–labeled folate receptors bound by the autoantibodies per milliliter of serum from the index subjects (27,705±5842 dpm) was significantly greater than that for the 20 controls (3279±1285 dpm, P=0.002) or that for the 8 control subjects who were pregnant when blood was obtained (5779±2982 dpm, P<0.001).
Figure 1. Autoantibodies against Folate Receptors in Serum from Women with a Current or Previous Pregnancy Complicated by a Neural-Tube Defect (Index Subjects, Panel A) and Control Subjects (Panel B).
Serum from index and control subjects was incubated with folic acid–labeled folate receptors and with a 10-fold excess (10x) of the folate receptors saturated with nonradioactive folic acid, to confirm the specificity of the folate receptor–autoantibody complex. The inset depicts the mean (±SE) values obtained for the index and control subjects. Control Subjects 1, 2, 3, and 4 were nulligravid and were excluded from the statistical analysis because they were not pregnant. The P value was determined with the use of Student's t-test. Asterisks in Panel A indicate that insufficient serum was available for this determination. Index Subjects 4, 5, and 12 were pregnant at the time of blood sampling. Index Subjects 10, 11, and 12 had a pregnancy complicated by a neural-tube defect and did not have the autoantibody against the folate receptors. In Panel B, Control Subjects 5 through 16 had previous pregnancies that were not complicated by a neural-tube defect; Control Subjects 17 through 24 were pregnant at the time of blood sampling. Serum from Control Subjects 16 and 24 contained autoantibodies against the folate receptors.
Figure 2 shows the effect of the autoantibodies on the binding of folic acid to folate receptors. Serum from Index Subjects 7 and 8 was selected on the basis of the availability of serum required for this analysis and contained autoantibodies that blocked the binding of folic acid to placental membrane receptors, ED27 cells, and KB cells at 4°C, whereas serum from five control subjects (Subjects 5, 6, 10, 11, and 14), which tested negative for autoantibodies, did not. Autoantibodies isolated from the serum of Index Subject 8 blocked 95 percent of the cellular uptake of folic acid by KB cells at 37°C (Figure 3), indicating that the epitope on the receptor is either at or close to the ligand-binding site.
Figure 2. Blocking of the Binding of Folic Acid to Folate Receptors on Placental Membranes, KB Cells, and ED27 Cells by Serum Containing Autoantibodies against Folate Receptors.
The number in each bar indicates the percentage of folic acid that is blocked from binding to the apo-folate receptors on the respective cell membranes by the autoantibodies at 4°C. Values are the means (±SE) of three experiments. P values are for the comparison with culture medium alone and were calculated with the use of Student's t-test. Asterisks indicate P<0.002, and daggers P<0.02.
Figure 3. Effect of Isolated Autoantibodies against Folate Receptors on the Cellular Uptake of Folic Acid by KB Cells.
Autoantibodies against folate receptors were isolated from serum from Index Subject 8 as described in the Methods section. KB cells were incubated overnight at 37°C with this isolated fraction of serum from either an index subject or a control subject. The uptake of folic acid by the KB cells was then determined. The results of a control incubation lacking the isolated fraction of serum are also shown. The I bars represent the SE.
Figure 4 shows the binding of apo-folate receptors to the autoantibodies in the serum from three index and two control subjects; the insets show the Scatchard analysis.15 Index Subjects 8 and 9 appeared to have a single order of autoantibody-binding sites on the receptor with similar Ka values (2.1x1010 and 1.8x1010 liters per mole, respectively). Index Subject 7 appeared to have two orders of autoantibody-binding sites on the receptor (Ka of 2.6x1010 and 2.8x109 liters per mole). The autoantibodies in the serum from the two control subjects (Subjects 16 and 24) also appeared to have two orders of binding sites, both of which (in Subject 16) were lower than the Ka for the index subjects.
Figure 4. Determination of the Binding Affinity of the Autoantibodies for the Folate Receptors in Index Subjects 7, 8, and 9 (Panel A) and Control Subjects 16 and 24 (Panel B).Placental membranes with apo-folate receptors were prepared as described in the Methods section and incubated overnight at 4°C with the serum containing autoantibodies. folic acid was then added, and the fraction bound to the folate receptors was subtracted from the total folate-binding capacity of the folate receptors to derive the level (in picomoles) of receptors blocked (B) by the autoantibodies. The ratio of the autoantibody-blocked receptor to the free apo-folate receptor (B/F) was used for the Scatchard analysis15 to compute the apparent association constant (Ka), which is shown in the inset in each panel.
The serum from the index subjects did not contain anticardiolipin, anti–single-stranded DNA, or antithyroid autoantibodies. One control subject (Subject 24) had an IgM anticardiolipin autoantibody, and another control subject (Subject 8) had an antithyroid autoantibody.
Discussion
We have identified autoantibodies against folate receptors in serum from women who have had a pregnancy complicated by a neural-tube defect. This autoantibody blocked the binding of folic acid to folate receptors on placental membranes and on ED27 cells and KB cells incubated at 4°C and inhibited the uptake of folate by KB cells when incubated at 37°C. The mean apparent Ka for the binding of the autoantibodies from the index subjects to the folate receptor was 2.2x1010 liters per mole. This high affinity could explain the blocking of the binding and cellular uptake of folate.
There are multiple causes of neural-tube defects, including drugs (especially antifolate16 and antiepileptic17 agents), chromosomal abnormalities,18 and environmental19 and genetic20 factors. Studies showing a reduction in the incidence of neural-tube defects of approximately 70 percent with periconceptional folic acid supplementation2,3 provide evidence that supplementary folate circumvents either an impaired intracellular folate-dependent enzyme pathway or an inhibitor of the cellular uptake of folate. However, the genetic variants of folate-pathway enzymes or of folate receptors identified in women who have pregnancies complicated by a neural-tube defect do not account for the 70 percent reduction in neural-tube defects associated with folate supplementation.21
The occurrence of the autoantibody might explain the observed benefit of periconceptional folic acid supplementation. The autoantibody-mediated blocking of cellular folate uptake by folate receptors could be bypassed by folic acid, because it is reduced and methylated in vivo and is transported into cells by the reduced folate carrier.22 Another potential mechanism that would be consistent with our observations is that folic acid, which has a high affinity for the receptor, may displace an autoantibody with a lower affinity for the receptor. The identification of the autoantibody against folate receptors in 75 percent of our index subjects — a rate similar to the decrease in the occurrence and recurrence of neural-tube defects of approximately 70 percent with periconceptional folic acid supplementation2,3 — is also consistent with the possibility that folate-responsive neural-tube defects may be due to impaired folate uptake caused by autoantibodies against the folate receptors. However, the results of our cross-sectional study cannot be used to establish a cause-and-effect relation between autoantibodies against the folate receptor and neural-tube defects.
The mechanism by which folate receptors might become self antigens is not known. Since the risk of neural-tube defects may increase after abortions23,24 or miscarriages,25 autoimmunity may be induced by epitopes of the folate receptors exposed as a result of injury and proteolysis of the reproductive tissues, which together with host genetic factors, may trigger the generation of autoantibodies. In our study, 7 of the 12 index subjects had previously had a first pregnancy complicated by a neural-tube defect, suggesting that the autoantibodies against the folate receptors were generated before pregnancy. Their appearance could be the consequence of an unrecognized earlier spontaneous abortion, which occurs with 30 percent of implantations in women with normal fertility.26
Some limitations of our study should be noted. Because it was designed as a pilot study, the sample size is small, and index subjects and controls were not recruited in a systematic manner. However, selection bias is unlikely. Index and control subjects were recruited without knowledge of whether they had autoantibodies against folate receptors or other autoantibodies, and they had similar demographic and clinical characteristics. Every woman who consented to participate in this pilot study was included. We do not have longitudinal serum samples, and thus we cannot assess whether autoantibodies were present before the pregnancy involving a neural-tube defect.
In summary, we have identified an autoantibody against the folate-receptor membrane protein in women who have had a pregnancy complicated by a neural-tube defect. Additional studies are needed to establish the frequency of folate-receptor autoantibodies in women whose pregnancies are complicated by this birth defect and to determine whether these autoantibodies are pathogenic.
Supported by the Raymond and Frances Church Biomedical Research Fund.
We are indebted to Marvin Numeroff, president of Universal Diagnostic Laboratory, for arranging for the analyses of other autoantibodies in the subjects' serum; and to Drs. C. Peterkin, N. Gabbur, and R. Patel for referring subjects from the Department of Obstetrics and Gynecology, SUNY Downstate Medical Center, Brooklyn, N.Y.
Source Information
From the Departments of Medicine (S.P.R., M.P.C., J.M.S., E.V.Q.) and Obstetrics and Gynecology (J.L.R.), the Division of Pediatric Neurology (J.C.), and the Scientific Computing Center (J.W.), State University of New York Downstate Medical Center, Brooklyn.
Address reprint requests to Dr. Rothenberg at SUNY Downstate Medical Center, 450 Clarkson Ave., Box 20, Brooklyn, NY 11203, or at srothenberg@downstate.edu.
References
Cragan JD, Roberts HE, Edmonds LD, et al. Surveillance for anencephaly and spina bifida and the impact of prenatal diagnosis -- United States, 1985-1994. Mor Mortal Wkly Rep CDC Surveill Summ 1995;44:1-13.
Czeizel AE, Dudás I. Prevention of the first occurrence of neural-tube defects by periconceptional vitamin supplementation. N Engl J Med 1992;327:1832-1835.
MRC Vitamin Study Research Group. Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. Lancet 1991;338:131-137.
Giles C. An account of 335 cases of megaloblastic anaemia of pregnancy and the puerperium. J Clin Pathol 1966;19:1-11.
Christensen B, Arbour L, Tran P, et al. Genetic polymorphisms in methylenetetrahydrofolate reductase and methionine synthase, folate levels in red blood cells, and risk of neural tube defects. Am J Med Genet 1999;84:151-157.
Antony AC. Folate receptors. Annu Rev Nutr 1996;16:501-521.
Piedrahita JA, Oetama B, Bennett GD, et al. Mice lacking the folic acid-binding protein Folbp1 are defective in early embryonic development. Nat Genet 1999;23:228-232.
Finnell RH, Wlodarczyk B, Speigelstein O, Triplett A, Gelineau-vanWaes J. Folate transport abnormalities and congenital defects. In: Milstien S, Kapatos G, Levine RA, Shane B, eds. Chemistry and biology of pteridines and folates. Boston: Kluwer Academic, 2002:637-42.
Barber RC, Shaw GM, Lammer EJ, et al. Lack of association between mutations in the folate receptor-alpha gene and spina bifida. Am J Med Genet 1998;76:310-317.
da Costa M, Rothenberg SP. Purification and characterization of folate binding proteins from rat placenta. Biochim Biophys Acta 1996;1292:23-30.
da Costa M, Sequeira JM, Rothenberg SP, Weedon J. Antibodies to folate receptors impair embryogenesis and fetal development in the rat. Birth Defects Res A Clin Mol Teratol 2003;67:837-47.
Sadasivan E, da Costa M, Rothenberg SP, Brink L. Purification, properties, and immunological characterization of folate-binding proteins from human leukemia cells. Biochim Biophys Acta 1987;925:36-47.
Kniss DA, Xie Y, Lin Y, et al. ED(27) trophoblast-like cells isolated from first-trimester chorionic villi are genetically identical to HeLa cells yet exhibit a distinct phenotype. Placenta 2002;23:32-43.
McHugh M, Cheng YC. Demonstration of a high affinity folate binder in human cell membranes and its characterization in cultured human KB cells. J Biol Chem 1979;254:11312-11318.
Scatchard G. The attractions of proteins for small molecules and ions. Ann N Y Acad Sci 1949;51:660-672.
Hernández-Diaz S, Werler MM, Walker AM, Mitchell AA. Folic acid antagonists during pregnancy and the risk of birth defects. N Engl J Med 2000;343:1608-1614.
Dansky LV, Rosenblatt DS, Andermann E, Mechanisms of teratogenesis: folic acid and antiepileptic therapy. Neurology 1992;42:Suppl 5:32-42.
Seller MJ. Neural tube defects, chromosome abnormalities and multiple closure sites for the human neural tube. Clin Dysmorphol 1995;4:202-207.
Finnell RH, Gelineau-van Waes J, Bennett GD, et al. Genetic basis of susceptibility to environmentally induced neural tube defects. Ann N Y Acad Sci 2000;919:261-277.
De Marco P, Moroni A, Merello E, et al. Folate pathway gene alterations in patients with neural tube defects. Am J Med Genet 2000;95:216-223.
Antony AC, Hansen DK. Hypothesis: folate-responsive neural tube defects and neurocristopathies. Teratology 2000;62:42-50.
Henderson GB. Folate-binding proteins. Annu Rev Nutr 1990;10:319-335.
Evans DR. Neural-tube defects: importance of a history of abortion in aetiology. Br Med J 1979;1:975-976.
Carmi R, Gohar J, Meizner I, Katz M. Spontaneous abortion -- high risk factor for neural tube defects in subsequent pregnancy. Am J Med Genet 1994;51:93-97.
Cuckle HS. Recurrence risk of neural tube defects following a miscarriage. Prenat Diagn 1983;3:287-289.
Wilcox AJ, Weinberg CR, O'Connor JF, et al. Incidence of early loss of pregnancy. N Engl J Med 1988;319:189-194.(Sheldon P. Rothenberg, M.)
Methods Serum from 12 women who were or had been pregnant with a fetus with a neural-tube defect and from 24 control women (20 with current or prior normal pregnancies and 4 who were nulligravid) was analyzed for autoantibodies by incubation with human placental folate receptors radiolabeled with folic acid. The properties of these autoantibodies were characterized by incubating serum and the autoantibodies isolated from serum with placental membranes, ED27 cells, and KB cells, which express the folate receptors.
Results Serum from 9 of 12 women with a current or previous affected pregnancy (index subjects) and 2 of 20 control subjects contained autoantibodies against folate receptors (P<0.001). The autoantibodies blocked the binding of folic acid to folate receptors on placental membranes and on ED27 and KB cells incubated at 4°C and blocked the uptake of folic acid by KB cells when incubated at 37°C.
Conclusions Serum from women with a pregnancy complicated by a neural-tube defect contains autoantibodies that bind to folate receptors and can block the cellular uptake of folate. Further study is warranted to assess whether the observed association between maternal autoantibodies against folate receptors and neural-tube defects reflects a causal relation.
Neural-tube defects, which include spina bifida, anencephaly, craniorachischisis, and encephalocele, occur in approximately 1 per 1000 births in the United States, and women who have one fetus with this complication are at increased risk in subsequent pregnancies.1 Although periconceptional folic acid supplementation reduces the occurrence and recurrence of neural-tube defects by approximately 70 percent,2,3 most women who are pregnant with a fetus with this complication do not have clinical folate deficiency.4 Though some polymorphisms for folate-pathway enzymes have been identified,5 they cannot account for the 70 percent decrease in the incidence of this birth defect with folate supplementation.
Studies in animals have suggested the importance of folate receptors in embryogenesis.6 Inactivation of both alleles encoding the mouse homologue of the human folate receptor gene was uniformly fatal in embryos with neural-tube defects.7,8 Folinic acid given to the pregnant dams resulted in normal development in 80 percent of the embryos that lacked the folate receptor gene in both alleles.8 Despite considerable research, no specific polymorphisms or mutations of the human folate receptor gene have been identified that might explain the reduction in the incidence of neural-tube defects with folic acid supplementation.9
Our observation that the administration of an antiserum to folate receptors10 to pregnant rats resulted in the resorption of or multiple developmental abnormalities in embryos11 prompted us to speculate that an autoantibody against folate receptors in women could induce similar embryonic and fetal abnormalities. Therefore, we developed an assay to measure these autoantibodies in the serum of women with a current or prior pregnancy complicated by a neural-tube defect and that of women without such a history.
Methods
Subjects
Between April 2002 and August 2003, all 12 women at our institution with a current or prior pregnancy complicated by a neural-tube defect who consented to participate in this study were enrolled and had their serum analyzed for autoantibodies against the folate receptor. The women were attending the obstetrical clinic or were the mothers of children with neural-tube defects who were being followed in the pediatric neurology clinic. Eligible women were between the ages of 18 and 40 years at the time of the affected pregnancy.
Initially, we enrolled 12 women with a current or previous affected pregnancy (index subjects) and 18 control subjects who had never had a pregnancy involving a neural-tube defect, 4 of whom were nulligravid. An initial analysis comparing the frequency of the autoantibodies in the group of index subjects and these 18 controls showed a significant difference (P<0.001).
Subsequently, we decided that women who were pregnant at the time of testing with a fetus that did not have a neural-tube defect would be more appropriate additional controls. Accordingly, the first six women who met this criterion and provided informed consent were also enrolled in the study. Thus, there were 24 controls; 8 were pregnant with an unaffected fetus, 12 had given birth to one or more unaffected infants, and 4 were nulligravid. Because pregnancy was a criterion for study entry, the four nulligravid control women who were initially tested for the autoantibody were excluded from the statistical analysis. Although the number of subjects in each group was not prespecified, every woman who consented to participate was included in the study. The institutional review board approved the study, and written informed consent was obtained from all subjects.
Preparation of Folic Acid–Labeled Purified Folate Receptors
Folate receptors from human placental membranes were solubilized at 37°C for two hours in 0.01 M sodium phosphate buffer, pH 7.4, containing 1 percent Triton X-100, and purified with Sepharose 4B to which folic acid was covalently coupled.12 The matrix was washed with the phosphate buffer, and the receptors were eluted from the matrix with 0.1 M glycine–hydrochloric acid, pH 3.5, containing 0.1 percent Triton X-100. The pH of the eluate was raised to 7.4 by the addition of 0.2 M sodium barbiturate, and the receptors were quantified on the basis of the level of binding of folic acid (specific activity, 31 Ci per millimole; Moravek), which binds at a 1:1 molar ratio.
Radiolabeled receptors were prepared by incubating the purified apo-folate receptors with a 20 percent excess of folic acid for 30 minutes at 4°C. At a pH between 7 and 9, there is virtually no dissociation of folic acid from the receptors incubated overnight at 4°C. The addition of a 10-fold greater concentration of the folate receptors, saturated with nonradioactive folic acid, confirmed the specificity of the folate receptor–autoantibody complex.
Assay of Serum for the Autoantibody against the Folate Receptors
For the assay for the autoantibody against folate receptors, 40 μl of serum was added to 200 μl of 0.1 M glycine buffer, pH 3.0, incubated for 10 minutes at 25°C, and then added to a pellet of dextran-coated charcoal to adsorb free folate and folate that dissociated from soluble receptors. At this pH, soluble receptors will also dissociate from autoantibodies in the serum. The charcoal was pelleted, and the supernatant fraction was added to 200 μl of 0.2 M sodium barbiturate, pH 8.9, containing phenylmethylsulfonyl fluoride, EDTA, 0.1 percent Triton X-100, and the folic acid–folate receptor preparation. Another aliquot of serum was added to a mixture of the folic acid–folate receptors containing a 10-fold excess of the receptors saturated with nonradioactive folic acid.
The volume of each reaction was increased to 600 μl by the addition of glycine–sodium barbiturate buffer (final pH, 8.6), incubated overnight at 4°C, and 600 μl of a 10 percent suspension of staphylococcal protein A membranes (Pansorbin, Calbiochem) was added. The incubation continued for 15 minutes; the membranes were then pelleted and washed, and the level of radioactivity was determined. The level of background radioactivity measured in duplicate reactions that lacked serum was subtracted from the test serum samples.
Autoantibodies and the Binding of Folic Acid to Folate Receptors
Membranes were prepared from human placenta homogenized in 0.01 M sodium phosphate buffer, then pelleted, and washed. The membranes were suspended in 0.1 M acetic acid to dissociate endogenous folate from the receptors, washed twice in this solution, and resuspended in 0.01 M sodium phosphate buffer, pH 7.4. The binding of folic acid was used to determine the quantity of the apo-folate receptors.
The effect of the autoantibodies on the binding of folic acid to the receptors was determined by incubating 200 μl of a 1 percent suspension of the membranes with charcoal-treated serum overnight at 4°C. The membranes were washed with cold phosphate-buffered saline, incubated for 30 minutes at 4°C with folic acid (125 pg) in 1 ml of phosphate-buffered saline, washed with cold buffer, and solubilized with 1 N sodium hydroxide, and the radioactivity was determined.
ED27 cells, a human-placenta cell line,13 and KB cells, a human epidermoid carcinoma-cell line,14 were each plated in triplicate at a density of 20,000 cells in 1.83-cm2 wells containing folate-deficient Dulbecco's minimum essential medium with 10 percent fetal-calf serum and the test serum (pretreated with charcoal to remove free endogenous folate) and incubated overnight at 37°C. The cells were washed with cold Hanks' balanced salt solution (HBSS) and incubated at 4°C for 30 minutes with folic acid (125 pg) in 1 ml of HBSS. The samples were washed with cold HBSS, and the cell-associated radioactivity was determined.
To obtain the apparent association constant (Ka) for the binding of the autoantibodies to the folate receptors, the charcoal-treated serum was incubated overnight at 4°C with placental membranes containing the apo-folate receptors, followed by incubation with folic acid for 30 minutes. The free ligand was removed by pelleting the membranes, and the radioactivity of the membrane-bound folic acid was determined. This value represented the apo-folate receptors that were not blocked by the autoantibodies. Scatchard analysis15 was used to compute the apparent Ka.
Autoantibodies and the Cellular Uptake of Folic Acid
Ten milliliters of serum from Index Subject 8 and from Control Subject 14 (see Supplementary Appendix 1 and Supplementary Appendix 2, available with the full text of this article at www.nejm.org) was incubated overnight at 4°C with 1 ml of packed placental membranes to bind the autoantibodies to the folate receptors. The samples were washed with cold HBSS to remove unbound serum proteins, and autoantibodies bound to the folate receptors were eluted with 0.1 M acetic acid containing 0.1 percent bovine serum albumin. The eluate was dialyzed against HBSS at 4°C overnight and concentrated by vacuum dialysis.
KB cells (20,000) plated in duplicate in 3.5-cm2 culture dishes were incubated overnight at 37°C with the isolated autoantibodies, the similarly treated control serum, or medium alone. The cells were washed with HBSS, medium containing folic acid (125 pg) was added to the wells, and the incubation was continued for 15, 30, and 60 minutes at 37°C, with duplicate sets incubated at 4°C. The cells were washed again with HBSS at 4°C and lysed with 1 N sodium hydroxide, and the radioactivity was determined. The difference in radioactivity between the cells incubated at 4°C and those incubated at 37°C represents the cellular uptake of folic acid.
The analysis of serum from index and control subjects for other autoantibodies was performed by Universal Diagnostic Laboratory with use of enzyme-linked immunosorbent assays (anti-DNA antibodies from Wampole Laboratories; anticardiolipin antibodies from DiaSorin; antithyroid antibodies from Diagnostic Products).
Statistical Analysis
The percentage of women with autoantibodies to the folate receptor in the two groups was compared with the use of Fisher's exact test. Mean binding levels of the folic acid–folate receptor complex in serum were compared between groups with the use of Student's t-test; Satterwaite's correction of the degrees of freedom was applied because of the unequal variances. Clinical variables in the two groups were compared with the use of the Wilcoxon two-sample test and Fisher's exact test. Two-tailed tests were used in all cases; P values of 0.05 or less were considered to indicate statistical significance. SAS software (SAS Institute) was used.
Results
Table 1 shows the demographic and clinical characteristics of the 12 index subjects and the 20 control subjects who had been or were pregnant. There were no significant differences between the groups. Other clinical characteristics of the index and control subjects (including the four nulligravid controls who were not included in the statistical analyses) are provided in Supplementary Appendix 1 and Supplementary Appendix 2 (available with the full text of this article at www.nejm.org).
Table 1. Clinical Characteristics of the Subjects.
Autoantibodies against the folate receptors were identified in the serum of 9 of 12 index subjects (75 percent) — 2 with a current pregnancy complicated by a neural-tube defect (Subjects 4 and 5) and 7 (Subjects 1, 2, 3, 6, 7, 8, and 9) with a history of a pregnancy complicated by a neural-tube defect (Figure 1A) — as compared with 2 of the 20 control subjects (10 percent, P<0.001) (Figure 1B). The mean value for the folic acid–labeled folate receptors bound by the autoantibodies per milliliter of serum from the index subjects (27,705±5842 dpm) was significantly greater than that for the 20 controls (3279±1285 dpm, P=0.002) or that for the 8 control subjects who were pregnant when blood was obtained (5779±2982 dpm, P<0.001).
Figure 1. Autoantibodies against Folate Receptors in Serum from Women with a Current or Previous Pregnancy Complicated by a Neural-Tube Defect (Index Subjects, Panel A) and Control Subjects (Panel B).
Serum from index and control subjects was incubated with folic acid–labeled folate receptors and with a 10-fold excess (10x) of the folate receptors saturated with nonradioactive folic acid, to confirm the specificity of the folate receptor–autoantibody complex. The inset depicts the mean (±SE) values obtained for the index and control subjects. Control Subjects 1, 2, 3, and 4 were nulligravid and were excluded from the statistical analysis because they were not pregnant. The P value was determined with the use of Student's t-test. Asterisks in Panel A indicate that insufficient serum was available for this determination. Index Subjects 4, 5, and 12 were pregnant at the time of blood sampling. Index Subjects 10, 11, and 12 had a pregnancy complicated by a neural-tube defect and did not have the autoantibody against the folate receptors. In Panel B, Control Subjects 5 through 16 had previous pregnancies that were not complicated by a neural-tube defect; Control Subjects 17 through 24 were pregnant at the time of blood sampling. Serum from Control Subjects 16 and 24 contained autoantibodies against the folate receptors.
Figure 2 shows the effect of the autoantibodies on the binding of folic acid to folate receptors. Serum from Index Subjects 7 and 8 was selected on the basis of the availability of serum required for this analysis and contained autoantibodies that blocked the binding of folic acid to placental membrane receptors, ED27 cells, and KB cells at 4°C, whereas serum from five control subjects (Subjects 5, 6, 10, 11, and 14), which tested negative for autoantibodies, did not. Autoantibodies isolated from the serum of Index Subject 8 blocked 95 percent of the cellular uptake of folic acid by KB cells at 37°C (Figure 3), indicating that the epitope on the receptor is either at or close to the ligand-binding site.
Figure 2. Blocking of the Binding of Folic Acid to Folate Receptors on Placental Membranes, KB Cells, and ED27 Cells by Serum Containing Autoantibodies against Folate Receptors.
The number in each bar indicates the percentage of folic acid that is blocked from binding to the apo-folate receptors on the respective cell membranes by the autoantibodies at 4°C. Values are the means (±SE) of three experiments. P values are for the comparison with culture medium alone and were calculated with the use of Student's t-test. Asterisks indicate P<0.002, and daggers P<0.02.
Figure 3. Effect of Isolated Autoantibodies against Folate Receptors on the Cellular Uptake of Folic Acid by KB Cells.
Autoantibodies against folate receptors were isolated from serum from Index Subject 8 as described in the Methods section. KB cells were incubated overnight at 37°C with this isolated fraction of serum from either an index subject or a control subject. The uptake of folic acid by the KB cells was then determined. The results of a control incubation lacking the isolated fraction of serum are also shown. The I bars represent the SE.
Figure 4 shows the binding of apo-folate receptors to the autoantibodies in the serum from three index and two control subjects; the insets show the Scatchard analysis.15 Index Subjects 8 and 9 appeared to have a single order of autoantibody-binding sites on the receptor with similar Ka values (2.1x1010 and 1.8x1010 liters per mole, respectively). Index Subject 7 appeared to have two orders of autoantibody-binding sites on the receptor (Ka of 2.6x1010 and 2.8x109 liters per mole). The autoantibodies in the serum from the two control subjects (Subjects 16 and 24) also appeared to have two orders of binding sites, both of which (in Subject 16) were lower than the Ka for the index subjects.
Figure 4. Determination of the Binding Affinity of the Autoantibodies for the Folate Receptors in Index Subjects 7, 8, and 9 (Panel A) and Control Subjects 16 and 24 (Panel B).Placental membranes with apo-folate receptors were prepared as described in the Methods section and incubated overnight at 4°C with the serum containing autoantibodies. folic acid was then added, and the fraction bound to the folate receptors was subtracted from the total folate-binding capacity of the folate receptors to derive the level (in picomoles) of receptors blocked (B) by the autoantibodies. The ratio of the autoantibody-blocked receptor to the free apo-folate receptor (B/F) was used for the Scatchard analysis15 to compute the apparent association constant (Ka), which is shown in the inset in each panel.
The serum from the index subjects did not contain anticardiolipin, anti–single-stranded DNA, or antithyroid autoantibodies. One control subject (Subject 24) had an IgM anticardiolipin autoantibody, and another control subject (Subject 8) had an antithyroid autoantibody.
Discussion
We have identified autoantibodies against folate receptors in serum from women who have had a pregnancy complicated by a neural-tube defect. This autoantibody blocked the binding of folic acid to folate receptors on placental membranes and on ED27 cells and KB cells incubated at 4°C and inhibited the uptake of folate by KB cells when incubated at 37°C. The mean apparent Ka for the binding of the autoantibodies from the index subjects to the folate receptor was 2.2x1010 liters per mole. This high affinity could explain the blocking of the binding and cellular uptake of folate.
There are multiple causes of neural-tube defects, including drugs (especially antifolate16 and antiepileptic17 agents), chromosomal abnormalities,18 and environmental19 and genetic20 factors. Studies showing a reduction in the incidence of neural-tube defects of approximately 70 percent with periconceptional folic acid supplementation2,3 provide evidence that supplementary folate circumvents either an impaired intracellular folate-dependent enzyme pathway or an inhibitor of the cellular uptake of folate. However, the genetic variants of folate-pathway enzymes or of folate receptors identified in women who have pregnancies complicated by a neural-tube defect do not account for the 70 percent reduction in neural-tube defects associated with folate supplementation.21
The occurrence of the autoantibody might explain the observed benefit of periconceptional folic acid supplementation. The autoantibody-mediated blocking of cellular folate uptake by folate receptors could be bypassed by folic acid, because it is reduced and methylated in vivo and is transported into cells by the reduced folate carrier.22 Another potential mechanism that would be consistent with our observations is that folic acid, which has a high affinity for the receptor, may displace an autoantibody with a lower affinity for the receptor. The identification of the autoantibody against folate receptors in 75 percent of our index subjects — a rate similar to the decrease in the occurrence and recurrence of neural-tube defects of approximately 70 percent with periconceptional folic acid supplementation2,3 — is also consistent with the possibility that folate-responsive neural-tube defects may be due to impaired folate uptake caused by autoantibodies against the folate receptors. However, the results of our cross-sectional study cannot be used to establish a cause-and-effect relation between autoantibodies against the folate receptor and neural-tube defects.
The mechanism by which folate receptors might become self antigens is not known. Since the risk of neural-tube defects may increase after abortions23,24 or miscarriages,25 autoimmunity may be induced by epitopes of the folate receptors exposed as a result of injury and proteolysis of the reproductive tissues, which together with host genetic factors, may trigger the generation of autoantibodies. In our study, 7 of the 12 index subjects had previously had a first pregnancy complicated by a neural-tube defect, suggesting that the autoantibodies against the folate receptors were generated before pregnancy. Their appearance could be the consequence of an unrecognized earlier spontaneous abortion, which occurs with 30 percent of implantations in women with normal fertility.26
Some limitations of our study should be noted. Because it was designed as a pilot study, the sample size is small, and index subjects and controls were not recruited in a systematic manner. However, selection bias is unlikely. Index and control subjects were recruited without knowledge of whether they had autoantibodies against folate receptors or other autoantibodies, and they had similar demographic and clinical characteristics. Every woman who consented to participate in this pilot study was included. We do not have longitudinal serum samples, and thus we cannot assess whether autoantibodies were present before the pregnancy involving a neural-tube defect.
In summary, we have identified an autoantibody against the folate-receptor membrane protein in women who have had a pregnancy complicated by a neural-tube defect. Additional studies are needed to establish the frequency of folate-receptor autoantibodies in women whose pregnancies are complicated by this birth defect and to determine whether these autoantibodies are pathogenic.
Supported by the Raymond and Frances Church Biomedical Research Fund.
We are indebted to Marvin Numeroff, president of Universal Diagnostic Laboratory, for arranging for the analyses of other autoantibodies in the subjects' serum; and to Drs. C. Peterkin, N. Gabbur, and R. Patel for referring subjects from the Department of Obstetrics and Gynecology, SUNY Downstate Medical Center, Brooklyn, N.Y.
Source Information
From the Departments of Medicine (S.P.R., M.P.C., J.M.S., E.V.Q.) and Obstetrics and Gynecology (J.L.R.), the Division of Pediatric Neurology (J.C.), and the Scientific Computing Center (J.W.), State University of New York Downstate Medical Center, Brooklyn.
Address reprint requests to Dr. Rothenberg at SUNY Downstate Medical Center, 450 Clarkson Ave., Box 20, Brooklyn, NY 11203, or at srothenberg@downstate.edu.
References
Cragan JD, Roberts HE, Edmonds LD, et al. Surveillance for anencephaly and spina bifida and the impact of prenatal diagnosis -- United States, 1985-1994. Mor Mortal Wkly Rep CDC Surveill Summ 1995;44:1-13.
Czeizel AE, Dudás I. Prevention of the first occurrence of neural-tube defects by periconceptional vitamin supplementation. N Engl J Med 1992;327:1832-1835.
MRC Vitamin Study Research Group. Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. Lancet 1991;338:131-137.
Giles C. An account of 335 cases of megaloblastic anaemia of pregnancy and the puerperium. J Clin Pathol 1966;19:1-11.
Christensen B, Arbour L, Tran P, et al. Genetic polymorphisms in methylenetetrahydrofolate reductase and methionine synthase, folate levels in red blood cells, and risk of neural tube defects. Am J Med Genet 1999;84:151-157.
Antony AC. Folate receptors. Annu Rev Nutr 1996;16:501-521.
Piedrahita JA, Oetama B, Bennett GD, et al. Mice lacking the folic acid-binding protein Folbp1 are defective in early embryonic development. Nat Genet 1999;23:228-232.
Finnell RH, Wlodarczyk B, Speigelstein O, Triplett A, Gelineau-vanWaes J. Folate transport abnormalities and congenital defects. In: Milstien S, Kapatos G, Levine RA, Shane B, eds. Chemistry and biology of pteridines and folates. Boston: Kluwer Academic, 2002:637-42.
Barber RC, Shaw GM, Lammer EJ, et al. Lack of association between mutations in the folate receptor-alpha gene and spina bifida. Am J Med Genet 1998;76:310-317.
da Costa M, Rothenberg SP. Purification and characterization of folate binding proteins from rat placenta. Biochim Biophys Acta 1996;1292:23-30.
da Costa M, Sequeira JM, Rothenberg SP, Weedon J. Antibodies to folate receptors impair embryogenesis and fetal development in the rat. Birth Defects Res A Clin Mol Teratol 2003;67:837-47.
Sadasivan E, da Costa M, Rothenberg SP, Brink L. Purification, properties, and immunological characterization of folate-binding proteins from human leukemia cells. Biochim Biophys Acta 1987;925:36-47.
Kniss DA, Xie Y, Lin Y, et al. ED(27) trophoblast-like cells isolated from first-trimester chorionic villi are genetically identical to HeLa cells yet exhibit a distinct phenotype. Placenta 2002;23:32-43.
McHugh M, Cheng YC. Demonstration of a high affinity folate binder in human cell membranes and its characterization in cultured human KB cells. J Biol Chem 1979;254:11312-11318.
Scatchard G. The attractions of proteins for small molecules and ions. Ann N Y Acad Sci 1949;51:660-672.
Hernández-Diaz S, Werler MM, Walker AM, Mitchell AA. Folic acid antagonists during pregnancy and the risk of birth defects. N Engl J Med 2000;343:1608-1614.
Dansky LV, Rosenblatt DS, Andermann E, Mechanisms of teratogenesis: folic acid and antiepileptic therapy. Neurology 1992;42:Suppl 5:32-42.
Seller MJ. Neural tube defects, chromosome abnormalities and multiple closure sites for the human neural tube. Clin Dysmorphol 1995;4:202-207.
Finnell RH, Gelineau-van Waes J, Bennett GD, et al. Genetic basis of susceptibility to environmentally induced neural tube defects. Ann N Y Acad Sci 2000;919:261-277.
De Marco P, Moroni A, Merello E, et al. Folate pathway gene alterations in patients with neural tube defects. Am J Med Genet 2000;95:216-223.
Antony AC, Hansen DK. Hypothesis: folate-responsive neural tube defects and neurocristopathies. Teratology 2000;62:42-50.
Henderson GB. Folate-binding proteins. Annu Rev Nutr 1990;10:319-335.
Evans DR. Neural-tube defects: importance of a history of abortion in aetiology. Br Med J 1979;1:975-976.
Carmi R, Gohar J, Meizner I, Katz M. Spontaneous abortion -- high risk factor for neural tube defects in subsequent pregnancy. Am J Med Genet 1994;51:93-97.
Cuckle HS. Recurrence risk of neural tube defects following a miscarriage. Prenat Diagn 1983;3:287-289.
Wilcox AJ, Weinberg CR, O'Connor JF, et al. Incidence of early loss of pregnancy. N Engl J Med 1988;319:189-194.(Sheldon P. Rothenberg, M.)