Evidence of eosinophil granule major basic protein in human placenta.

A protein immunochemically related to the eosinophil granule major basic protein (gMBP) is found in increased concentration in the plasma of pregnant women and has been localized to placental trophoblasts by immunofluorescence. Pregnancy MBP (pMBP) is indistinguishable from gMBP in its reactivity with polyclonal antisera and a panel of 14 mouse mAbs. We report the purification of pMBP from human placenta by: (a) affinity chromatography over mAb immobilized on Sepharose, (b) gel filtration in 6 M guanidine.HCl buffer, and (c) reversed-phase HPLC. Purified pMBP and gMBP are biochemically indistinguishable in that both: (a) bind to DNA, (b) polymerize and bind to carrier proteins via disulfide linkages, (c) have a molecular weight of 14,000, (d) have isoelectric points greater than 10.6, (e) comigrate in two-dimensional gels, (f) coelute during reversed-phase HPLC on C18 columns, (g) have identical peptide maps after three different digestions, and (h) have partial amino acid sequence identity. This physicochemical identity has important implications as to the role of pMBP in human placentation.

Immunohistological and elution studies of the human placenta.

An immunohistological survey of 28 full-term human placentas has demonstrated deposits of IgG, beta1C, beta1E, and fibrinogen/fibrin in areas of fibrinoid necrosis and on the trophoblast basement membrane in approximately 35% of placental villi. Traces of IgM were detected at similar sites in 18 of 28 full-term placentas. In 11 specimens of immature placentas (10-18 wk gestation) traces of IgG and beta1C and deposits of fibrinogen/fibrin were also present, but IgM was not detected in this material. IgG was recovered in acidic eluates from an homogenized placenta which behaved as an antibody reactive with unidentified material present in fibrinoid deposits and on the thickened trophoblast basement membrane of some villi. It could not be determined whether this IgG was derived from the maternal or fetal circulation.

Plasminogen activator-specific inhibitors produced by human monocytes/macrophages.

Human monocytes/macrophages produce plasminogen activator-specific inhibitors (PAIs) that form covalent complexes with urokinase-type plasminogen activator (uPA). We have characterized two functionally and antigenically related forms of PAIs produced by resting and phorbol myristate acetate (PMA)-treated U 937 cells: an Mr 40,000 form, presumably nonglycosylated, with a pI of 5.2, that is constitutively synthetized by these cells and that remains predominantly intracellular; a PMA-induced form of heterogeneous Mr (50,000-65,000) with a pI of 4.7, that is preferentially secreted; this PAI is glycosylated with terminal sialic acid residue(s). Biosynthetic labeling experiments demonstrated that both PAIs are synthetized by U 937 cells. They are inactivated upon treatment with propanol, heat, and acid; the covalent and equimolar complexes formed between these PAIs and 125I-uPA are dissociated by ammonium hydroxide, suggesting that the PAIs are linked to uPA via an ester bond. Human peripheral blood monocytes/macrophages also produce the two forms of PAI. These PAIs are clearly different from the main plasma protease inhibitors and they are both antigenically related to the PAI-2 characterized in human placenta.

Expression of transforming growth factor-beta 1 in specific cells and tissues of adult and neonatal mice.

We have used immunohistochemical techniques to detect transforming growth factor-beta 1 (TGF-beta 1) in many tissues of adult and neonatal mice. Each of two antibodies raised to the amino-terminal 30 amino acids of TGF-beta 1 selectively stained this molecule in either intracellular or extracellular locations. Strong intracellular staining was found in adrenal cortex, megakaryocytes and other cells of the bone marrow, cardiac myocytes, chondrocytes, renal distal tubules, ovarian glandular cells, and chorionic cells of the placenta. Marked staining of extracellular matrix was found in cartilage, heart, pancreas, placenta, skin, and uterus. Staining was often particularly intense in specialized cells of a given tissue, suggesting unique roles for TGF-beta within that tissue. Levels of expression of mRNA for TGF-beta 1 and its histochemical staining did not necessarily correlate in a given tissue, as in the spleen. The present data lend further support to the concept that TGF-beta has an important role in controlling interactions between epithelia and surrounding mesenchyme.

Adhesive and degradative properties of human placental cytotrophoblast cells in vitro.

Human fetal development depends on the embryo rapidly gaining access to the maternal circulation. The trophoblast cells that form the fetal portion of the human placenta have solved this problem by transiently exhibiting certain tumor-like properties. Thus, during early pregnancy fetal cytotrophoblast cells invade the uterus and its arterial network. This process peaks during the twelfth week of pregnancy and declines rapidly thereafter, suggesting that the highly specialized, invasive behavior of the cytotrophoblast cells is closely regulated. Since little is known about the actual mechanisms involved, we developed an isolation procedure for cytotrophoblasts from placentas of different gestational ages to study their adhesive and invasive properties in vitro. Cytotrophoblasts isolated from first, second, and third trimester human placentas were plated on the basement membrane-like extracellular matrix produced by the PF HR9 teratocarcinoma cell line. Cells from all trimesters expressed the calcium-dependent cell adhesion molecule cell-CAM 120/80 (E-cadherin) which, in the placenta, is specific for cytotrophoblasts. However, only the first trimester cytotrophoblast cells degraded the matrices on which they were cultured, leaving large gaps in the basement membrane substrates and releasing low molecular mass 3H-labeled matrix components into the medium. No similar degradative activity was observed when second or third trimester cytotrophoblast cells, first trimester human placental fibroblasts, or the human choriocarcinoma cell lines BeWo and JAR were cultured on radiolabeled matrices. To begin to understand the biochemical basis of this degradative behavior, the substrate gel technique was used to analyze the cell-associated and secreted proteinase activities expressed by early, mid, and late gestation cytotrophoblasts. Several gelatin-degrading proteinases were uniquely expressed by early gestation, invasive cytotrophoblasts, and all these activities could be abolished by inhibitors of metalloproteinases. By early second trimester, the time when cytotrophoblast invasion rapidly diminishes in vivo, the proteinase pattern of the cytotrophoblasts was identical to that of term, noninvasive cells. These results are the first evidence suggesting that specialized, temporally regulated metalloproteinases are involved in trophoblast invasion of the uterus. Since the cytotrophoblasts from first trimester and later gestation placentas maintain for several days the temporally regulated degradative behavior displayed in vivo, the short-term cytotrophoblast outgrowth culture system described here should be useful in studying some of the early events in human placen

Heparin facilitates the extraction of tissue fibronectin.

Extraction of fibronectin from two human tissues, lung parenchyma and placental villi, was facilitated by the incorporation of heparin into extraction media. The effect of heparin was additive to the effect of urea which is known to extract fibronectin. These experiments provide further evidence that fibronectin and glycosaminoglycans are associated in connective tissues and the use of heparin forms the basis for a simple method for extraction and quantitation of tissue fibronectin.

A novel human gene closely related to the abl proto-oncogene.

A DNA sequence related to the abl proto-oncogene was identified in human placenta. Molecular cloning and nucleotide sequence analysis revealed two putative exons whose predicted amino acid sequence was most homologous to the corresponding sequences of c-abl and v-abl but was related to other tyrosine kinase genes as well. The new sequence was localized by in situ hybridization and somatic cell genetic analysis to human chromosome 1q24-25, which differs from the location of any previously identified tyrosine kinase gene. The detection of a novel 12-kb transcript by this gene in human normal and tumor cells establishes it as a new member of the tyrosine kinase family that is closely related to but distinct from c-abl.

Detection of smoking-related covalent DNA adducts in human placenta.

The presence of covalent DNA chemical addition products (adducts) in human term placentas was investigated by recently developed immunologic and 32P-postlabeling assays. DNA from placental specimens of smokers showed a small but not statistically significant increase in adduct levels when tested by antibodies to DNA modified with a benzo[a]pyrene dihydrodiol epoxide (BPDE-I), the ultimate carcinogenic derivative of benzo[a]pyrene. The postlabeling assay detected several modified nucleotides, one of which (adduct 1) strongly related to maternal smoking during pregnancy. This adduct was present in placental tissue from 16 of 17 smokers, but only 3 of 14 nonsmokers. Among smokers, levels of adduct 1 in general were only weakly related to questionnaire and biochemical measures of the intensity of smoking exposures, which suggests modulation by individual susceptibility factors. The adduct seemed to be derived from an aromatic carcinogen, but it may not result from several of the most intensely studied polycyclic aromatic hydrocarbons or aromatic amines in tobacco smoke. The data show the association of cigarette smoking with covalent damage to human DNA in vivo.

Radioimmunoassay of the insulin receptor: a new probe of receptor structure and function.

A sensitive and specific radioimmunoassay for the insulin receptor has been developed employing receptor autoantibodies from the serum of a patient with insulin-resistant diabetes. The assay detects insulin binding sites at concentrations as low as 0.1 nanomolar; distinguishes between receptors originating from human placental membranes, human lymphoblastoid cells, and mouse liver membranes; and measures the receptor independently of its binding function. Down-regulation, or loss of binding after exposure to insulin, is associated with loss of immunoreactive receptor.

Human retinoblastoma susceptibility gene: cloning, identification, and sequence.

Recent evidence indicates the existence of a genetic locus in chromosome region 13q14 that confers susceptibility to retinoblastoma, a cancer of the eye in children. A gene encoding a messenger RNA (mRNA) of 4.6 kilobases (kb), located in the proximity of esterase D, was identified as the retinoblastoma susceptibility (RB) gene on the basis of chromosomal location, homozygous deletion, and tumor-specific alterations in expression. Transcription of this gene was abnormal in six of six retinoblastomas examined: in two tumors, RB mRNA was not detectable, while four others expressed variable quantities of RB mRNA with decreased molecular size of about 4.0 kb. In contrast, full-length RB mRNA was present in human fetal retina and placenta, and in other tumors such as neuroblastoma and medulloblastoma. DNA from retinoblastoma cells had a homozygous gene deletion in one case and hemizygous deletion in another case, while the remainder were not grossly different from normal human control DNA. The gene contains at least 12 exons distributed in a region of over 100 kb. Sequence analysis of complementary DNA clones yielded a single long open reading frame that could encode a hypothetical protein of 816 amino acids. A computer-assisted search of a protein sequence database revealed no closely related proteins. Features of the predicted amino acid sequence include potential metal-binding domains similar to those found in nucleic acid-binding proteins. These results provide a framework for further study of recessive genetic mechanisms in human cancers.

Molecular cloning of two types of GAP complementary DNA from human placenta.

The ras p21 GTPase-activating protein (GAP) was purified from human placental tissue. Internal amino acid sequence was obtained from this 120,000-dalton protein and, by means of this sequence, two types of complementary DNA clones were isolated and characterized. One type encoded GAP with a predicted molecular mass of 116,000 daltons and 96% identity with bovine GAP. The messenger RNA of this GAP was detected in human lung, brain, liver, leukocytes, and placenta. The second type appeared to be generated by a differential splicing mechanism and encoded a novel form of GAP with a predicted molecular mass of 100,400 daltons. This protein lacks the hydrophobic amino terminus characteristic of the larger species, but retains GAP activity. The messenger RNA of this type was abundantly expressed in placenta and in several human cell lines, but not in adult tissues.

A single receptor binds both insulin-like growth factor II and mannose-6-phosphate.

Amino acid sequences deduced from rat complementary DNA clones encoding the insulin-like growth factor II (IGF-II) receptor closely resemble those of the bovine cation-independent mannose-6-phosphate receptor (Man-6-P receptorCI), suggesting they are identical structures. It is also shown that IGF-II receptors are adsorbed by immobilized pentamannosyl-6-phosphate and are specifically eluted with Man-6-P. Furthermore, Man-6-P specifically increases by about two times the apparent affinity of the purified rat placental receptor for 125I-labeled IGF-II. These results indicate that the type II IGF receptor contains cooperative, high-affinity binding sites for both IGF-II and Man-6-P-containing proteins.

Effect of perturbation of specific folate receptors during in vitro erythropoiesis.

Although antisera to specific placental folate receptors inhibits the uptake of 5-methyltetrahydrofolate into cultured malignant human cells, little is known of the functional significance of folate receptors in normal human cells. Human bone marrow cells were therefore assayed for erythropoietic burst-forming units in the presence of an antihuman placental folate receptor serum and preimmune serum to determine the role of such a receptor in erythroid differentiation. When marrow cells were assayed in the presence of anti-receptor antiserum, there was (i) a threefold increase in erythropoietic burst formation and a twofold increase in the number of cells per erythroid burst; (ii) morphological evidence for nuclear/cytoplasmic dissociation of orthochromatic normoblasts composing erythroid bursts (megaloblastic erythropoiesis); (iii) intracellular folate deficiency with a 70% reduction of intracellular folate in antiserum treated cells as compared with control cells; and (iv) complete reversal of antiserum-induced changes on preincubation of antiserum with purified human placental folate receptor. These data support the conclusion that folate receptors on marrow cells provide an important function in the cellular uptake of folates during in vitro erythropoiesis. This process of folate uptake also appears to play a pivotal role in the differentiation and proliferation of erythroid progenitor cells.

Extraction and characterization of a thyrotropic material from the human placenta.

A material with thyrotropic activity was extracted from fresh human placentas. After chromatography of the extract on carboxymethyl cellulose, thyroid-stimulating activity ranged from 0.12 to 1.06 mU of thyroid-stimulating hormone (TSH) per mg of protein in bioassays of eight preparations. The amount of TSH per placenta varied from 113 to 2200 mU and approximated the content of the pituitary gland. Additional purification by gel filtration on Sephadex G-100 gave a maximum activity of 8 mU/mg. The most active portion was eluted in the same position as (125)I-labeled bovine or human TSH, a fact suggesting that the molecular size of this thyrotropic substance was similar to that of pituitary TSH. Another placental fraction with weaker activity was eluted earlier indicating that the placental material was heterogeneous. In the McKenzie mouse bioassy, the response of the placental thyrotropin paralleled that of the beef TSH standard. There was no long-acting thyroid stimulator effect. Antibodies to both human and bovine pituitary TSH neutralized the biologic activity of the placental TSH. Placental thyrotropin cross-reacted very weakly in a sensitive radioimmunoassay for human pituitary TSH; it cross-reacted completely in a radioimmunoassay for bovine pituitary TSH, and this assay was used for following the purification. The role of this thyrotropic material as a possible cause of thyroid hypertrophy and hyperfunction in pregnancy and in patients with trophoblastic tumors remains to be investigated.

Measurement, characterization, and source of somatostatin-like immunoreactivity in human amniotic fluid.

Somatostatin-like immunoreactivity (SLI) is widely distributed in tissues and biological fluids. To determine whether SLI is also present in amniotic fluid, samples obtained by amniocentesis from 30 normal and 27 abnormal pregnancies were studied by radioimmunoassay. Direct incubation of [(125)I-Tyr(1)]tetradecapeptide somatostatin (SRIF) with amniotic fluid resulted in 89% tracer degradation. Damage was reduced to <5% when samples were acidified and boiled before the assay. With this technique, SLI was detectable in all normal amniotic fluid samples; the mean level at 15-20 wk of gestation (320+/-55 pg/ml, n = 15) being 4.5 times higher than the mean at 32-43 wk (70+/-12 pg/ml, n = 15) (P < 0.001). In cases of preeclampsia (n = 6), gestational diabetes (n = 5), anencephaly (n = 1), and meningomyelocele (n = 1), SLI values were in the normal range, but in one juvenile diabetic and one patient with chronic renal failure, SLI was undetectable (<10 pg/ml). In a pair of monochorionic diamniotic twins, SLI levels were very different (33 and 197 pg/ml), which suggests that fetal factors are more important than materno-placental ones in determining amniotic fluid SLI. Serial dilutions of amniotic fluid showed parallelism with standard SRIF. When concentrates of pooled amniotic fluid were chromatographed on Sephadex G-25 columns, all SLI eluted in the void volume ahead of SRIF even after treatment with 8 M urea and dithiothreitol. This "big" SLI incubated in amniotic fluid showed 100% stability over 24 h at 37 degrees C, whereas SRIF was rapidly inactivated (t((1/2)) congruent with 7 min). Extracts of placenta and fetal membranes contained no SLI, but small amounts (6-20% of total amniotic fluid SLI) were found in cells from fresh fluid. Radioimmunoassay of SLI in extracts of seven paired cord arterial and venous plasma samples showed no arteriovenous gradient consistent with fetal origin of cord blood SLI. It is concluded that (a) amniotic fluid contains SLI which is of fetal origin and (b) normal levels vary with gestational age. The SLI has a higher molecular weight (>/=5,000) and is more stable in amniotic fluid than SRIF.

Characterization and histologic localization of human growth hormone-variant gene expression in the placenta.

The human growth hormone-variant (hGH-V) gene is one of five highly similar growth hormone-related genes clustered on the short arm of chromosome 17. Although the pattern of expression of the adjacent normal growth hormone (hGH-N) and chorionic somatomammotropin (hCS) genes in this cluster are well characterized, the expression of the hGH-V gene remains to be defined. In previous studies, we have demonstrated that the hGH-V gene is transcribed in the term placenta and expressed as two alternatively spliced mRNAs: one is predicted to encode a 22-kD hormone (hGH-V), the other retains intron 4 in its sequence resulting in the predicted synthesis of a novel 26-kD hGH-V-related protein (hGH-V2). In the present report, we document the expression of both of these hGH-V mRNA species in the villi of the term placenta, demonstrate an increase in their concentrations during gestation, and directly sublocalize hGH-V gene expression to the syncytiotrophoblastic epithelium of the term placenta by in situ cDNA-mRNA histohybridization. The demonstrated similarity in the developmental and tissue-specific expression of the hGH-V gene with that of the related hCS gene suggests that these two genes may share common regulatory elements.

Detection of O6-methyldeoxyguanosine in human placental DNA.

A monoclonal antibody specific for O6-methyldeoxyguanosine (O6-MedGuo) was developed. When used in a competitive enzyme-linked immunosorbent assay, 50% inhibition of binding was achieved with 0.51 pmol O6-MedGuo. When the competitive enzyme-linked immunosorbent assay was coupled with high-performance liquid chromatography, 2 mg of DNA could be analyzed giving a lower limit of detection of 0.5 mumol O6-MedGuo/mol deoxyguanosine. This assay was used to test for O6-MedGuo in DNA from placentas of smoking and nonsmoking women. Two of 10 DNA samples from smoking women and three of 10 from nonsmoking women had detectable concentrations of O6-MedGuo. Concentrations ranged from 0.6 to 1.6 mumol O6-MedGuo/mol deoxyguanosine. Activity of O6-alkylguanine DNA alkyltransferase was also measured. There was no apparent relationship between O6-alkylguanine DNA alkyltransferase activity and O6-MedGuo concentrations in the 20 subjects, nor did mean O6-alkylguanine DNA alkyltransferase activity differ between the two groups. Although no apparent relationship between smoking history and O6-MedGuo concentration was found in this preliminary study, this is the first report of a structurally identified DNA adduct in human placenta.

Comparative 32P-analysis of cigarette smoke-induced DNA damage in human tissues and mouse skin.

Previous studies using a highly sensitive 32P-postlabeling assay for the analysis of carcinogen/mutagen-induced DNA damage have shown the presence of tobacco smoking-related DNA adducts in human placenta (Everson, R.B., Randerath, E., Santella, R.M., Cefalo, R.C., Avitts, T. A., and Randerath, K., Science (Wash. DC), 231: 54-57, 1986). The occurrence of such adducts in smokers' bronchus and larynx is reported here. Since the chemical nature of these adducts could not be characterized by direct methods due to the extremely low levels of individual adducts (less than 0.03 fmol per microgram DNA), we have sought an experimental animal model for studying the formation of tobacco-related DNA adducts. Because cigarette smoke condensate is known to initiate tumors in mouse skin, ICR mice were treated topically with cigarette tar equivalent to 1.5, 3, 6, and 9 cigarettes for 0.4, 3, 5, and 7 days, respectively, and skin DNA was isolated 1 day after the last treatment. When DNA from exposed mice was analyzed by the 32P-postlabeling assay, 12 distinct 32P-labeled DNA adduct spots, as well as a diagonal radioactive zone, which presumably reflected the presence of incompletely resolved adducts, were noted on polyethyleneimine-cellulose TLC fingerprints. One derivative in particular (adduct 1) was seen to increase rapidly during the early treatment phase and also to persist to 8 days after treatment. The prominent adduct 1 was observed in the same location on the fingerprints of DNA samples from smokers. Cochromatography experiments suggested identity of human and mouse DNA adduct 1. Similarly, several other human and mouse adducts (adducts 3, 5, 6, and 9) appeared identical, and the diagonal radioactive zone was also present on DNA adduct maps from smokers. While absolute levels of individual human adducts were too low to be accurately quantitated, semiquantitative estimation of total tobacco-related aromatic DNA adducts in the human specimens gave values of 1 adduct in (1.7-2.9) X 10(7) nucleotides (0.10-0.18 fmol per micrograms DNA), with adduct 1 constituting 8.5-14% of the total. On the basis of these results, it appears now feasible to determine the chemical origin of smoking-induced DNA adducts in human tissues by preparation of authentic 32P-labeled reference adducts from animals treated with characterized subfractions of cigarette tar, 32P-postlabeling, and cochromatography of 32P-labeled human and animal adducts.

Characterization of sequences related to the mouse mammary tumor virus that are specific to MCF-7 breast cancer cells.

Mouse mammary tumour virus (MuMTV) DNA hybridized more strongly to two human EcoRI fragments (6.6 and 9.5 kilobases) in DNA from the MCF-7 human breast cancer cell line than in DNA from normal human placenta. Seven recombinants (NMWV 1E1-7) containing these MuMTV-related sequences were identified. Hybridization of NMWV 1E probes to Southern transfers of human DNA suggested that these probes hybridize to multiple sequences in the human genome. The pattern obtained was very similar to that obtained using MuMTV gag-pol DNA. Analysis of the cloned DNA showed that the NMWV 1E MuMTV-related sequences are arranged as tandem repeats and are contained in EcoRI fragments of 6.6 or 9.5 kilobases. Only two NMWV 1E EcoRI fragments (9.5 and 15 kilobases) were detected in 17 DNA samples prepared from human placenta and blood. In contrast the 6.6-kilobase EcoRI fragment was detected in two (MCF-7 and EFM-19) of seven breast cancer cell lines and the MDA MB-231 cell line contained NMWV 1E sequences in an EcoRI fragment of 9.8 kilobases.

Ah receptor in human placenta: stabilization by molybdate and characterization of binding of 2,3,7,8-tetrachlorodibenzo-p-dioxin, 3-methylcholanthrene, and benzo(a)pyrene.

Aryl hydrocarbon hydroxylase (AHH, cytochrome P1-450) is highly inducible in several human cells and tissues exposed to specific halogenated and nonhalogenated aromatic chemicals of the "3-methylcholanthrene-type." In laboratory animals AHH induction is known to be regulated by binding of inducers to the Ah receptor, a soluble intracellular protein. However, the induction mechanism in the human species is incompletely understood largely because the Ah receptor, which seems to be essential to the induction process, has not previously been detectable in certain human cells and tissues (including placenta) that are highly responsive to AHH induction. We found that human placenta contains high concentrations of Ah receptor (comparable to the receptor concentrations in rat and mouse liver) but that special modifications were necessary in the assay techniques in order to detect and accurately quantitate receptor binding. Receptor was detected at concentrations approximately equal to 100 fmol/mg cytosol protein using [3H]2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) as the radioligand. This high concentration of specific binding sites was present only if the placental tissue was initially homogenized in a buffer containing sodium molybdate (10 or 20 mM). Without molybdate in the homogenizing buffer, specific [3H]TCDD binding was only about 35 fmol/mg. Specific Ah receptor binding also was detectable with [3H]-3-methylcholanthrene and, to a lesser extent, with [3H]-benzo(alpha)pyrene. The receptor sedimented near 9S on sucrose gradients whether molybdate was present or not. About 80% of specific binding was lost if excessive charcoal was used to adsorb "nonspecifically bound" ligand from cytosol prior to gradient analyses. The apparent affinity with which [3H]TCDD bound to Ah receptor in human placental cytosol was relatively low (apparent Kd approximately equal to 5 to 8 nM) when compared with the affinity of [3H]TCDD binding in rat or mouse hepatic cytosols (Kd approximately equal to 1 to 3 nM). These data suggest that while molybdate has very little effect on the quantity or molecular size of the rodent Ah receptor assay, it is very important in stabilizing the human Ah receptor. Our experiments demonstrate that human placenta contains a high concentration of Ah receptor and suggest that AHH induction in placenta is mediated through a receptor mechanism analogous to that previously established in tissues and cells from laboratory animals.