Share this article on:

Alpha-Fetoprotein and Hematopoietic Growth Factors in Amniotic Fluid


Original Research

Objective To determine whether a relationship exists between alpha-fetoprotein (AFP) and hematopoietic growth factors in amniotic fluid.

Methods Forty-one women at 15 weeks' gestation were included in the study. Gestational age was assessed by obtaining a reliable menstrual history and scanning. Amniocentesis was performed, and each woman subsequently delivered anatomically and chromosomally normal infants. The level of AFP was determined using a standard automated procedure. The concentrations of stem cell factor, interleukin 3, interleukin 6, erythropoietin, and granulocyte colony-stimulating factor (G-CSF) were measured using a commercially available immunoassay. The relationships between AFP and the studied cytokines were evaluated using the Pearson linear correlation test. Significant correlations were studied further by linear and nonlinear regression to obtain the best predictive model.

Results There was a significant correlation between AFP and stem cell factor (r = .47, P = .002). No significant correlations between AFP and the rest of the studied cytokines were found (r = −.07, r = .02, r = −.02, and r = −.11 for erythropoietin, G-CSF, interleukin 3, and interleukin 6, respectively).

Conclusion Alpha-fetoprotein is significantly correlated with stem cell factor in early pregnancy and might play a role in fetal hematopoiesis.

Alpha-fetoprotein has correlates with stem cell factor and might play a role in fetal hematopoiesis.

Departments of Obstetrics and Gynecology and Clinical Chemistry, University Hospital of Puerto Real, Puerto Real, Cádiz, Spain.

Address reprint requests to: Jose Luis Bartha, MD, Department of Obstetrics and Gynecology, University Hospital of Puerto Real, Carretera Nacional IV, Km 665, Puerto Real, Cádiz 11510, Spain; E-mail:

This work was supported by a grant from the Asociación para la Promoción de la Investigación en Ginecología y Obstetricia.

Received December 27, 1999. Received in revised form March 22, 2000. Accepted April 7, 2000.

The biologic role of alpha-fetoprotein (AFP) in normal development is unclear. In the last decade, the growth-regulatory properties of AFP have aroused interest. This molecule is capable of regulating growth in ovarian, placental, uterine, phagocytic, bone marrow, and lymphatic cells.1

Several facts may point to a hypothetical relationship between AFP and fetal hematopoiesis. First, AFP is synthesized mainly by the fetal liver and the yolk sac. Both are hematopoietic organs in the fetus, and both tumors of the yolk sac and hepatocellular carcinomas can produce AFP.2,3 Second, an association has been reported between higher serum levels of AFP and a paraneoplastic syndrome characterized by erythrocytosis in patients with hepatocellular carcinoma.3 Finally, AFP is higher in women carrying fetuses with acquired anemia due to different causes.4,5

In a previous study, we demonstrated a significant correlation between fetal red blood cell (RBC) count, hemoglobin, hematocrit, erythropoietin, and AFP.6 Gestational age seemed to affect the magnitude of the correlations, which were higher at earlier developmental stages. Therefore, we suspected that the possible role of AFP in fetal hematopoiesis could be more important in early pregnancy than later. Because of the difficulties involved in obtaining fetal blood in early pregnancy, we decided to perform a study on amniotic fluid (AF). The synergistic action of AFP with various growth factors, with AFP enhancing mitogenic activity, has been reported.7 We then sought to evaluate the relationships between AFP and hematopoietic growth factors, to clarify the mechanism through which AFP might act.

Back to Top | Article Outline

Materials and Methods

Between October 1998 and February 1999, 41 pregnant women were enrolled in the study. Demographics are shown in Table 1. All were consecutive patients undergoing genetic amniocentesis in our antenatal clinics at 15 weeks' gestation for reasons of advanced maternal age (more than 35 years) (38 cases) or previous delivery of an infant with chromosomal abnormalities (three cases). Gestational age was assessed by obtaining a reliable menstrual history and scanning. However, all women had regular menses, and there were no discordances between gestational age estimated using the date of the last menstrual period and that determined by ultrasound. At the time of amniocentesis, and after written informed consent was obtained, an extra amount of AF was taken (to a maximum of 5 mL), and this AF was centrifuged, frozen, and stored in six aliquots at −80C. All women delivered anatomically and chromosomally normal infants.

Table 1

Table 1

The level of AFP was determined using a standard automatic procedure (AFP immunoassay, Elecsys 1010/2010 Systems; Hitachi, Tokyo, Japan) at appropriate dilutions. Levels of stem cell factor, interleukin 3, interleukin 6, granulocyte colony-stimulating factor (G-CSF), and erythropoietin were measured using a commercially available immunoassay (Research and Diagnostics Systems Inc., Minneapolis, MN). Duplicates were assayed in every experiment. The main methodology for these immunoassays was the same for all studied cytokines. A monoclonal antibody specific for the cytokine was coated onto a microtiter plate provided in the kit. Standards with known concentrations of cytokines and samples were pipetted into wells, and the studied cytokine was bound by an immobilized antibody during a first incubation at room temperature. After any unbound proteins were washed away, an enzyme-linked polyclonal antibody specific for the studied cytokines was added to the wells, and the samples were incubated again at room temperature. After a wash to remove unbound antibody reagent, a substrate solution was added to the wells, and a color developed in proportion to the amount of cytokine bound in the initial step. Color development was stopped after 20 minutes, and the intensity of the color was measured using a microtiter plate reader. The minimum detectable dose was 7.4 pg/mL for interleukin 3, 0.70 pg/mL for interleukin 6, 7 pg/mL for G-CSF, 9 pg/mL for stem cell factor, and 2 mU/mL for erythropoietin. Because of various problems before or during procedures (limited amount of AF or accidentally broken container), a few samples had to be excluded from some assays. A total of 41 samples were studied for AFP, stem cell factor, erythropoietin, and G-CSF; 39 were studied for interleukin 6; and 30 were studied for interleukin 3.

Variables were analyzed using SPSS statistical software (Statistical Package for Social Sciences; SPSS Inc., Chicago, IL). Distributions were checked using histograms and the Kolmogorov-Smirnov test. When a variable was not distributed in a normal manner, log-transformed values were used to perform statistical analyses. The relationships between AFP and the studied cytokines were evaluated using the Pearson linear correlation test. Significant correlations were studied further by linear and nonlinear regression to obtain the best predictive model. Significance had been set previously at 95% (P < .05).

Back to Top | Article Outline


Table 2 shows the mean, standard deviation, median, interquartile range, and range of each studied variable. Some variables (interleukin 3, interleukin 6, and G-CSF) had a nonnormal distribution, and log-transformed values were used. The level of AFP in AF had a significant correlation (Table 2) with stem cell factor (r = .47, P = .002) (Figure 1). No significant correlations between AFP and the rest of the studied cytokines were found (r = −.07, r = .02, r = −.02, and r = −.11 for erythropoietin, G-CSF, interleukin 3, and interleukin 6, respectively).

Figure 1

Figure 1

Table 2

Table 2

To examine in more detail the relationship between AFP and stem cell factor, regression models that estimated stem cell factor over AFP concentrations were fitted. Even though linear regression showed significant links between both variables (R 2 = 0.22, P = .002) (Figure 1), the equation that best fit the data was a power model—a nonlinear model whose equation is Y = b0 × (t b1)t or ln(b0) + (b1 × ln(t), with, in our case, b0 = 50.7, b1 = 0.27, Y = stem cell factor, and t = AFP (R 2 = 0.54, P < .001) (Figure 2).

Figure 2

Figure 2

Back to Top | Article Outline


It is suspected that the physiologic role of AFP is to modulate various growth-regulatory pathways during fetal development.8 However, the exact biologic role of this abundant protein in fetal life is still unclear.

In a previous study, we found a significant correlation between maternal serum alpha-fetoprotein (MSAFP) and fetal hemoglobin obtained both by prenatal cordocentesis and during elective cesarean delivery before labor.6 In a second part of that study, we also found a significant negative correlation between fetal serum AFP and fetal RBC count, hemoglobin, hematocrit, and erythropoietin. Therefore, it was reasonable to suggest that AFP could have a role in fetal erythropoiesis.

In the late 1970s, several investigators suspected a relationship between AFP and fetal hematopoiesis.9,10 Congote et al11 studied AFP as well as the synthesis of heme associated with hemoglobin in short-term cultures of human fetal liver cells, to correlate the relationship between AFP and erythroid cell function. They concluded that there is no cause-and-effect relationship between AFP production and erythroid cell function in human fetal liver cells and that the two processes occur independently in different cell types.

In vivo hematopoiesis is a complex system regulated by various mechanisms that appear to be under the control of many hemopoietic growth factors. In this context, the synthesis of heme associated with hemoglobin is only one of the last steps in a long chain that includes activation of stem cells, division of progenitors, proliferation, and differentiation.12

We have found a significant correlation between AFP and stem cell factor. Stem cell factor is the ligand for the tyrosine kinase receptor c-kit, which is expressed on both primitive and mature hematopoietic progenitor cells. This cytokine, synergizing with other growth factors, preserves the viability of hematopoietic stem and progenitor cells, preventing their apoptosis, and influences their entry into the cell cycle while it facilitates their proliferation and differentiation.13 In vivo, stem cell factor, in combination with G-CSF, also enhances mobilization of peripheral blood progenitor cells.14 Heidari et al15 found that AF promoted proliferation of both fetal liver cells and adult bone marrow cells, and the proliferation-promoting activity was partially neutralized by the anti–stem cell factor antibody. In another study, the same investigators found that the proliferation-promoting activity of murine AF was present only in mice younger than 2 weeks.16 These results indicate that AF potentially contains multiple growth factors, including stem cell factor, that preferentially affect the early stage of hematopoiesis. Therefore, if AFP has a connection with stem cell factor, it might have a role in fetal hematopoieis, at least at this early stage.

Some relationships between AFP and stem cell factor are well documented. In animal models, bile duct ligation can be used to induce proliferation of a hepatic stem cell compartment that participates in the renewal process of epithelial cell populations in the liver. Stem cell factor is one of the growth factor or receptor systems associated with early activation of this compartment. When this procedure is performed, expression of AFP, used as an indicator of activation of the stem cell compartment, is followed by a consecutive increase in expression of stem cell factor.17 In addition, oval cells, which produce AFP, express the c-kit receptor tyrosine kinase and its ligand, stem cell factor. The stem cell factor/c-kit system might be involved in the early activation of the hepatic stem cells as well as in the expansion and differentiation of oval cells.18 Moreover, both AFP and stem cell factor are genes known to be present in fetal liver.19

Nevertheless, several other factors might influence stem cell factor concentration and action significantly. These factors include cytokines such as tumor necrosis factor α,20 interleukin 1,21 and tumor growth factor β 1,22 and hormones such as FSH23 and glucocorticoids.24 We did not find correlations between AFP and other hematopoietic growth factors. However, there are significant doubts about whether erythropoietin and other classic hematopoietic cytokines such as interleukin 3 are the main promoting factors in fetal liver hematopoiesis. While studying a fetal hepatocyte cell line capable of supporting hematopoiesis, Hata et al25 found that the hematopoietic activities were attributed to granulocyte-macrophage colony-stimulating factor and macrophage colony-stimulating factor, whereas no influences for interleukin 3, interleukin 6, G-CSF, or erythropoietin were identified. Fetuses of preeclamptic women have shown severely reduced hepatic erythropoiesis in spite of an increase in erythropoietin and interleukin 3 levels.26 These results suggested a discrepancy between increased numbers of erythrocytes in peripheral blood and depression of hematopoiesis at the main production site, the fetal liver. Another possible explanation could be that these hematopoietic factors might affect later stages of fetal hematopoiesis.

Ultrastructural differences are described in hepatocytes according to whether the cells produce AFP. These observations led to the hypothesis that AFP might correspond to a specific functional state of the hepatocyte in the human fetal liver.27 Expression of AFP is a constant feature of extramedullary hematopoiesis, occurring not only in physiologic situations but also in pathologic ones. Expression of AFP occurs not only when hepatic tissue is present, such as in cases of hepatoblastoma or hepatic adenoma of the placenta,28 but also in some kinds of leukemia, such as juvenile chronic myeloid leukemia or erythroleukemia, in which there is a major reversion to fetal erythropoiesis related to the phenomenon of fetal protein synthesis.29

We can speculate about three possible mechanisms by which AFP can modulate fetal hematopoiesis. First, it has been reported that AFP might serve as an antioxidant on the basis of amino acid domain and motif structure.1 Given that the main antioxidant enzyme, myeloperoxidase, acts only during the late stages of myeloid differentiation,30 AFP might be useful as an antioxidant in early stages, when cells are too immature to transcribe myeloperoxidase. Second, AFP might enhance the proliferative activity of several growth factors,7 including stem cell factor. The mechanism of the growth-regulatory properties remains unclear, but such modulation might be attributed to transcriptional enhancement and/or suppression activities.31 Third, AFP could act to avoid apoptosis1 in cells participating in early stages of fetal hematopoiesis. Thus, AFP could enhance the aforementioned action of stem cell factor to prevent apoptosis of hematopoietic stem and progenitor cells.

From an obstetric perspective, determining the relationships between AFP and fetal hematopoiesis could be help explain poorly understood clinical phenomena such as increased levels of MSAFP in conditions associated with fetal anemia or decreased levels in clinical situations associated with fetal hemoconcentration (eg, in women at risk of carrying fetuses with chromosomal abnormalities or in pregnant women with diabetes).32 Moreover, unexplained increases in MSAFP levels have been associated with an increased risk of poor perinatal outcome, but the explanation for this phenomenon is still unclear,33,34 and fetal anemia leading to greater hepatic hematopoietic activity could be involved in some of these cases.

Back to Top | Article Outline


1. Mizejewski GJ. α-Fetoprotein as a biologic response modifier: Relevance to domain and subdomain structure. Proc Soc Exp Biol Med 1997;215:333–62.
2. Sakisaka S, Watanabe M, Tateishi H, Harada M, Shakado S, Mimura Y, et al. Erythropoietin production in hepatocellular carcinoma cells associated with polycythemia: Immunohistochemical evidence. Hepatology 1993;18:1357–62.
3. Kew MC, Fisher JW. Serum erythropoietin concentrations in patients with hepatocellular carcinoma. Cancer 1986;58:2485–8.
4. Carrington D, Gilmore DH, Whittle MJ, Aitken D, Gibson AA, Patrick WJ. Maternal serum alpha-fetoprotein: A marker of fetal aplastic crisis during intrauterine human parvovirus infection. Lancet 1982;1(8530):433–4.
5. Seppälä M, Ruoslahti E. Alpha-fetoprotein in Rh-immunized pregnancies. Obstet Gynecol 1973;42:701–5.
6. Bartha JL, Comino-Delgado R, Arce F, Alba P, Broullon JR, Barahona M. Relationship between alpha-fetoprotein and fetal erythropoiesis. J Reprod Med 1999;44:689–97.
7. Keel BA, Eddy KB, Cho S, May JV. Synergistic action of purified α-fetoprotein and growth factors on the proliferation of porcine granulosa cells in monolayer culture. Endocrinology 1991;129:217–25.
8. Mizejewski GJ. Alpha-fetoprotein binding proteins: Implications for transmembrane passage and subcellular localization. Life Sci 1994;56:1–9.
9. Tourne CE, Brettes JP, Guez G, Ritter J, Gandar R. Alpha fetoprotein and neonatal jaundice: Contribution to the study of a physiopathologic mechanism [in French]. J Gynecol Obstet Biol Reprod (Paris) 1977;6:311–7.
10. Ikonen RS, Lindgren J, Niemi E, Sorto AE, Seppälä M, Ruoslahti E. Alpha fetoprotein levels in neonatal hyperbilirubinaemia. Acta Paediatr Scand 1980;69:59–63.
11. Congote LF, Bruno F, Solomon S. Alpha-fetoprotein production and erythropoiesis in cell cultures of human fetal liver. Can J Biochem 1977;55:571–5.
12. De Haan G, Dontje B, Nijhof W. Concepts of hemopoietic cell amplification: Synergy, redundancy and pleiotropy of cytokines affecting the regulation of erythropoiesis. Leuk Lymphoma 1996;22:385–94.
13. Hassan HT, Zander A. Stem cell factor as a survival and growth factor in human normal and malignant hematopoiesis. Acta Haematol 1996;95:257–62.
14. McNiece IK, Briddell RA. Stem cell factor. J Leukoc Biol 1995;58:14–22.
15. Heidari Z, Isobe K, Kiuchi K, Goto S, Nakashima I, Tomoda Y. Promotion of proliferation of murine hematopoietic stem cells by growth factors in murine amniotic fluid. J Reprod Immunol 1996;31:51–64.
16. Heidari Z, Isobe K, Goto S, Nakashima I, Kiuchi K, Tomoda Y. Characterization of the growth factor activity of amniotic fluid on cells from hematopoietic and lymphoid organs of different life stages. Microbiol Immunol 1996;40:583–9.
17. Omori M, Evarts RP, Omori N, Hu Z, Marsden ER, Thorgeirsson SS. Expression of alpha-fetoprotein and stem cell factor/c-kit system in bile duct ligated young rats. Hepatology 1997;25:1115–22.
18. Fujio K, Evarts RP, Hu Z, Marsden ER, Thorgeirsson SS. Expression of stem cell factor and its receptor, c-kit, during liver regeneration from putative stem cells in adult rat. Lab Invest 1994;70:511–6.
19. Malhotra K, Luehrsen KR, Costello LL, Raich TJ, Sim K, Foltz L, et al. Identification of differentially expressed mRNAs in human fetal liver across gestation. Nucleic Acids Res 1999;27:839–47.
20. Jacobsen FW, Dubois CM, Rusten LS, Veiby OP, Jacobsen SE. Inhibition of stem cell factor-induced proliferation of primitive murine hematopoietic progenitor cells signaled through the 75-kilodalton tumor necrosis factor receptor: Regulation of c-kit and p53 expression. J Immunol 1995;154:3732–41.
21. Koenig A, Yakisan E, Reuter M, Huang M, Sykora KW, Corbacioglu S, et al. Differential regulation of stem cell factor mRNA expression in human endothelial cells by bacterial pathogens: An in vitro model of inflammation. Blood 1994;83:2836–43.
22. de Vos S, Brach MA, Asano Y, Ludwig WD, Bettelheim P, Gruss HJ, et al. Transforming growth factor-beta 1 interferes with the proliferation-inducing activity of stem cell factor in myelogenous leukemia blasts through functional down-regulation of the c-kit proto-oncogene product. Cancer Res 1993;53:3638–42.
23. Yan W, Linderborg J, Suominen J, Toppari J. Stage-specific regulation of stem cell factor gene expression in the rat seminiferous epithelium. Endocrinology 1999;140:1499–504.
24. Kassel O, Schmidlin F, Duvernelle C, de Blay F, Frossard N. Up- and down-regulation by glucocorticoids of the constitutive expression of the mast cell growth factor stem cell factor by human lung fibroblasts in culture. Mol Pharmacol 1998;54:1073–9.
25. Hata M, Nanno M, Doi H, Satomi S, Sakata T, Suzuki R, et al. Establishment of a hepatocytic epithelial cell line from the murine fetal liver capable of promoting hemopoietic cell proliferation. J Cell Physiol 1993;154:381–92.
26. Stallmach T, Karolyi L, Lichtlen P, Maurer M, Hebisch G, Joller H, et al. Fetuses from preeclamptic mothers show reduced hepatic erythropoiesis. Pediatr Res 1998;43:349–54.
27. Peyrol S, Grimaud JA, Pirson Y, Chayvialle JA, Touillon C, Lambert B. Ultrastructural immunoenzymatic study of alpha-fetoprotein-producing cells in the human fetal liver. J Histochem Cytochem 1977;25:432–8.
28. Khalifa MA, Gersell DJ, Hansen CH, Lage JM. Hepatic (hepatocellular) adenoma of the placenta: A study of four cases. Int J Gynecol Pathol 1998;17:241–4.
29. Forget BG, Glass J, Housman D. Erythroid cell differentiation. Hamatol Bluttransfus 1976;19:109–24.
30. Lubbert M, Miller CW, Koeffler HP. Changes of DNA methylation and chromatin structure in the human myeloperoxidase gene during myeloid differentiation. Blood 1991;78:345–56.
31. Mizejewski GJ. Alpha-fetoprotein signal sequences: A proposed mechanism for subcellular localization and organelle targeting. J Theor Biol 1995;176:103–13.
32. Salvesen DR, Brudenell JM, Snijders RJM, Ireland RM, Nicolaides KH. Fetal plasma erythropoietin in pregnancies complicated by maternal diabetes mellitus. Am J Obstet Gynecol 1993;168:88–94.
33. Katz VL, Chescheir NC, Cefalo RC. Unexplained elevations of maternal serum alpha-fetoprotein. Obstet Gynecol Surv 1990;45:719–26.
34. Boyd PA. Why might maternal serum AFP be high in pregnancies in which the fetus is normally formed? Br J Obstet Gynaecol 1992;99:93–5.
© 2000 The American College of Obstetricians and Gynecologists