1 Department of Obstetrics and Gynecology, Hokkaido University School of Medicine, Sapporo 060-8638, 2 Osaka Research Laboratories, Wako Pure Chemical Industries, Ltd., Amagasaki, 661-0963, Japan
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words:
-fetoprotein/AFP-L3/Down's syndrome/Lens culinaris agglutinin/prenatal screening
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
An association between trisomy 21 and low levels of -fetoprotein (AFP) in maternal serum in mid-trimester was reported in the 1980s (Cuckle et al., 1984
; Merkatz et al., 1984
; Nicolini et al., 1988
; Suzumori et al., 1997
). In the early 1990s, biochemical markers such as human chorionic gonadotrophin (HCG) (Spencer et al., 1997
), unconjugated oestriol (uE3) (Cheng et al., 1993
; David, 1996
), inhibin A (Van Lith et al., 1992
), and pregnancy-associated plasma protein (Brambati et al., 1993
) were studied to determine their potential for DS screening. Recently, Cole et al have focused on the presence of variant N-linked oligosaccharides on HCG-related molecules in DS pregnancies, and those may be an alternative to HCG (Cole et al., 1997
; Cole, 1999
; Bahado-Singh, 2000
). There continues to be a need for more reliable screening marker for DS.
It is well known that AFP is a glycoprotein produced by the fetal liver, hepatocellular carcinoma and yolk sac tumours. The carbohydrate structure of human AFP has been analysed using preparations purified from ascites fluid of a patient with hepatocellular carcinoma (Yoshima et al., 1980) and a yolk sac tumour (Yamashita et al., 1983
). The sugar chain microheterogeneities of AFP, especially Lens culinaris agglutinin-reactive AFP (AFP-L3), have been studied using lectin affinity electrophoresis in relation to hepatocellular carcinoma (Breborowicz et al., 1981
; Taketa et al., 1993
; Yamashita et al., 1996
). The relationship between AFP glycoforms in maternal serum at the second trimester and DS-affected pregnancies has only been determined using Concanavalin A (Gembruch, 1987; Kim, 1990
; Los et al., 1995
). In this study, we determined the percentage of AFP variants that react with lectins as a way to find out whether the analysis of the carbohydrate chain microheterogeneities of maternal serum AFP could be useful for prenatal DS screening.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Measurement of total AFP, HCG, and uE3
Total AFP, HCG and uE3 concentrations were determined routinely using commercially available kits (AFP; Abbott Laboratories, IL, USA; HCG; Wallak Oy., Turku, Finland; uE3; Diagnostic Products Corp., CA, USA). The multiple of the median (MoM) of the AFP, HCG and uE3 in this clinical study was calculated based on 4256 Japanese unaffected pregnancies at 1420 weeks gestation (Suzumori et al., 1997).
Lectin-affinity electrophoresis analysis of AFP
The percentage of AFP reactive with Lens culinaris agglutinin (LCA) was obtained by lectin-affinity electrophoresis coupled with antibody-affinity blotting (AFP Differentiation Kit L; Wako Pure Chemical Industries, Ltd., Amagasaki, Japan) (Shimizu et al., 1993). The percentage of other lectin-reactive AFPs was measured similarly using Concanavalin A (Con A), erythroagglutinating phytohemagglutinin E4 (EPHA) and Ricinus communis agglutinin-120 (RCA) in agarose gels (Shimizu et al., 1996
). Band intensities separated by a lectin-affinity electrophoresis are expressed as percentages of the total band intensity. Faint bands that were not detectable by densitometer were expressed as <0.5% and the minimum detection level of AFP in one band was 2 ng/ml in a sample (8 pg/band) (Shimizu et al., 1993
). The AFP band 1 had no affinity for lectins because the mobility was identical to that on gels without lectin, and bands numbered 2 and higher had correspondingly higher affinities for lectins. The correspondence between separated AFP bands and the estimated structure of the carbohydrate chain has been described elsewhere (Shimizu et al., 1996
). Both AFP-L2 and AFP-L3 have an
1
6 fucose residue, and the sum of them was evaluated, although AFP-L2 appears generally in amniotic fluids.
Extraction of AFP from liver tissue
Liver tissue samples ~3x3 mm were washed with ice-cold 10 mmol/l phosphate buffer including 0.15 mol/l NaCl, pH 8.0, to remove blood. The tissue weight was measured after removal of the buffer, and then it was homogenized with 0.5 ml of buffer using a glass homogenizer in an ice-bath. AFP in the fetal liver was obtained from the supernatant by centrifugation (1200 g, 10 min), and it was analysed by lectin affinity electrophoresis.
Sialidase digestion of AFP obtained from fetal liver
A volume of 5 µl of 150 IU/ml sialidase (Streptococcus sp.) in 0.1 mol/l 2-(N-morpholinoethanesulphonic acid/NaOH buffer, pH 6.0, was added to 45 µl of the supernatant of fetal liver tissue and incubated at 37°C for 24 h, and the resulting solution was analysed by lectin affinity electrophoresis.
Statistical analysis
Statistical significance was determined by Student's t test, Welch t-test and the MannWhitney U-test. To determine the clinical accuracy of the percentages of lectin-reactive AFP bands and other markers, the area under the receiver operating characteristics (ROC) curve was compared by Z score testing.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
We first focused on any changes in the AFP carbohydrate chain structure during the processes ranging from the AFP production in the fetal liver to its presence in maternal serum. AFP is produced in the developing fetus by both the yolk sac and the fetal liver (Gembruch et al., 1987). At around 12 weeks gestation, the yolk sac degenerates and the fetal liver becomes the main site of synthesis. It was expected that AFP in very early amniotic fluid would mainly be of the yolk sac type sugar chain (Los et al., 1995
). After 14 weeks gestation, AFP synthesis decreases with advancing gestation; therefore, maximum fetal plasma concentration of AFP is reached at 13 weeks gestation and declines exponentially thereafter (Blair et al., 1987
). No significant difference was reported in the steady-state level of fetal liver AFP mRNA levels in trisomy 21 and 18 groups, and the decrease in maternal serum AFP concentration is unlikely to be the consequence of impaired transcription of the AFP gene by the fetal liver (Brizot et al., 1996
). In maternal blood, AFP concentrations rise throughout the first and second trimesters and decline only after 32 gestational weeks. This increase is thought to be due to increased placental permeability to fetal plasma proteins with advancing gestation (Gitlin, 1975
; van Lith et al., 1991
). In the present study, we determined that the carbohydrate structure of AFP in the trisomy 21 fetal liver was simple and uniform in three cases, namely, almost all have biantennary structure with ~30% containing
1
6 fucose residue, 70% without fucose residue, and the reducing end is uneven, with ~44% of agalactosyl, ~6% mono-galactosyl and ~50% of the di-galactosyl biantennary chain.
We examined specimens obtained from fetal liver, amniotic fluid and maternal serum from three pregnant women with trisomy 21 fetuses since the presence in these different tissues may be affected by membrane-permeability, diffusion and active transport. The lectin affinity electrophoresis in these cases revealed relatively identical results. Lectin-reactive AFP bands AFP-L2, AFP-C1, AFP-P4 and AFP-P5 in the amniotic fluid, which have the additional branching sugar to a biantennary structure, indicated relatively greater values than those in both fetal liver and maternal serum. That is possibly because these AFPs having multi-antennary and/or branching sugars do not easily permeate through the membrane of the placenta, although the production in fetal liver is limited in quantity. We failed to detect AFP-L2, which has both 1
6 fucose residue and an additional branching sugar to a biantennary structure in the maternal serum, but we were able to detect it in the amniotic fluid. Even though AFP-L2 has extremely low permeability, the percentage of AFP-L3 in amniotic fluid and maternal serum was similar in the DS pregnancies, thereby suggesting that the presence of
1
6 fucose residue on N-acetylglucosamine at the reducing end does not generally affect the permeability; however the additional branching sugar to the biantennary structure does affect the permeability. AFP-C1 has an additional branching sugar or a multi-antennary carbohydrate chain, while AFP-C2+C3 shows a biantennary chain. Compared with the amniotic fluid and maternal serum, the biantennary chain ratio in the maternal serum tends to increase, indicating that permeability and/or transport at a biantennary chain in the placenta or fetal membrane is relatively high or that of a multi-antennary chain is relatively low.
Because of the special nature of the extracted fluid from the trisomy 21 fetal liver tissues, the difference between sugar-chain bands at different stages in the process by which AFP reaches the maternal blood is great, with the exception of AFP-L3. Looking at the barriers between the amniotic fluid and maternal serum, it is necessary to keep in mind the difference in fluid quantity of the two. The total AFP concentration in the amniotic fluid is 13 722 ± 6586 ng/ml, while that in the maternal serum is 42 ± 7 ng/ml, suggesting that the total AFP hardly permeates the membrane. No correlation was obtained between the AFP concentration in amniotic fluids and that in maternal serum in either pregnancies with DS or unaffected pregnancies, as shown elsewhere (Barford et al., 1985). However, the percentage of AFP-L3 in the cases of the pregnancies with a trisomy 21 fetus seems to be relatively unaffected by such factors as the permeability and/or transport of fetal membrane and placenta, or metabolism in the amniotic fluid and maternal tissues. These findings make AFP-L3 an appropriate choice for a trisomy 21 biochemical marker.
Our second focus was to determine the clinical utility of examining microheterogeneities of the AFP carbohydrate chain in maternal serum for prenatal screening for DS. All AFP bands in maternal serum and amniotic fluid were compared for discriminating efficiency between 17 DS affected pregnancies and 20 unaffected pregnancies, with only AFP-L3 showing a significant difference as a maternal serum biochemical marker. With ROC analysis using 22 DS-affected pregnancies and 227 unaffected pregnancies, the percentage of AFP-L3 in maternal serum was revealed to have a greater area under the curve than those of MoM AFP and MoM uE3, and it was revealed to have a tendency towards a larger area than that of MoM HCG.
This is a first report consisting of findings based on the study of fewer than 250 pregnancies at around 16 weeks gestation. Considering that AFP concentrations in the fetal liver peak at 13 weeks gestation (Brizot et al., 1996), a prospective study of the measurement of AFP-L3 in maternal serum at the end of first trimester to early in the second trimester of pregnancy is advisable. Due to the technical complexity of measurement, it is difficult to measure large numbers of samples in a short time. The DS screening utility of this potentially valuable marker with the use of an automatic analyser for measuring AFP-L3 (Katoh et al., 1998
) will be tested further.
![]() |
Notes |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Barford, D.A., Dickerman, L.H. and Johnson, W.E. (1985) -Fetoprotein: Relationship between maternal serum and amniotic fluid levels. Am. J. Obstet. Gynecol., 151, 10381041.[ISI][Medline]
Blair, J.I., Carachi, R., Gupta, R. et al. (1987) Plasma -fetoprotein reference ranges in infancy: effect of prematurity. Arch. Dis. Child., 62, 362369.[Abstract]
Brambati, B., Macintosh, M.C.M., Teisner, B. et al. (1993) Low maternal serum levels of pregnancy associated plasma protein A (PAPP-A) in the first trimester in association with abnormal fetal karyotype. Br. J. Obstet. Gynecol., 100, 324326.[ISI][Medline]
Breborowicz, J., Mackiewicz, A. and Breborowicz, D. (1981) Microheterogeneity of -fetoprotein in patient serum as demonstrated by lectin affino-electrophoresis. Scand. J. Immunol., 14, 1520.[ISI][Medline]
Brizot, M.L., Mckie, A.T., Kaisenbeg, C.S. et al. (1996) Fetal hepatic alpha-fetoprotein mRNA expression in fetuses with trisomy 21 and 18 at 1215 weeks. Early Hum. Dev., 44, 155159.[ISI][Medline]
Cheng, E.Y., Luthy, D.A., Zebelman, A.M. et al. (1993) A prospective evaluation of a second trimester screening test for fetal Down syndrome using maternal serum alpha-fetoprotein, hCG, and unconjugated estriol. Obstet. Gynecol., 81, 7277.[Abstract]
Cole, L.A., Cermik, D. and Bahado-Singh, R.O. (1997) Oligosaccharide variants of hCG-related molecules: potential screening markers for Down syndrome. Prenat. Diagn., 17, 11881190.[ISI][Medline]
Cole, L.A. (1999) Hyperglycosylated human chorionic gonadotropin (invasive trophoblast antigen) immunoassay: A new basis for gestational Down syndrome screening. Clin. Chem., 45, 21092119.
Cuckle, H.S., Wald, N.J. and Lindenbaum, R.H. (1984) Maternal serum alpha-fetoprotein measurement: a screening test. Lancet, i, 926929.
David, M. (1996) Unconjugated estriol as maternal serum marker for the detection of Down syndrome pregnancies. Fetal Diagn. Ther., 11, 99105.[ISI][Medline]
Gembruch, U., Bellmann, O. and Hansmann, M. (1987) Concanavalin A reactivity of alpha-fetoprotein in the amniotic fluid in autosomal trisomies of the fetus. Geburtshilfe Frauenheilkd, 47, 601603.[ISI][Medline]
Gitlin, D. (1975) Normal biology of alpha-fetoprotein. Ann. New Yolk Acad. Sci., 259, 715.
Jacobs, P.A. and Morton, N.E. (1977) Origin of human trisomics and polyploid. Hum. Hered., 27, 5972.[ISI][Medline]
Jou, H.J. (2000) Maternal serum screening for down syndrome by using alpha-fetoprotein and human chorionic gonadotropin in an asian population. A prospective study. Fetal Diagn., 15, 108111.
Katoh, H., Nakamura, K., Tanaka, T. et al. (1998) Automatic and simultaneous analysis of Lens culinaris agglutinin-reactive -fetoprotein ratio and total
-fetoprotein concentration. Anal. Chem., 70, 21102114.[ISI][Medline]
Kim, T.L. (1990) Carbohydrate heterogeneity of human alpha-fetoprotein in pregnancy. Methodological and clinical studies. Danish Medical Bulletin, 37, 4157.[ISI][Medline]
Lam, Y.H. (1998) Second-trimester maternal serum alpha-fetoprotein and human chorionic gonadotropin screening for Down's syndrome in Hong Kong. Prenat. Diagn., 18, 585589.[ISI][Medline]
Los, F.J., Janse, H.C., Brandenburg, H. et al. (1995) Concanavalin A variants of alpha-fetoprotein in first trimester fetuses with trisomy 21 and with normal karyotypes. Gynecol. Obstet. Invest., 39, 149152.[ISI][Medline]
Merkatz, I.R., Nitowsky, H.M., Macri, J.N. and Johnson, W.E. (1984) An association between low maternal serum alpha-fetoprotein and fetal chromosomal abnormalities. Am. J. Obstet. Gynecol., 14, 886894.
Nicolini, U., Hubinont, C., Santolaya, J. et al. (1988) Fetal serum alpha-fetoprotein in fetuses with chromosomal abnormalities. Lancet, ii, 13161317.
Shimizu, K., Taniichi, T., Satomura, S. et al. (1993) Establishment of assay kits for the determination of microheterogeneities of alpha-fetoprotein using lectin-affinity electrophoresis. Clin. Chim. Acta,., 214, 312.[ISI][Medline]
Shimizu, K., Katoh, H., Yamashita, F. et al. (1996) Comparison of carbohydrate structures of serum -fetoprotein by sequential glycosidase digestion and lectin affinity electrophoresis. Clin. Chim. Acta,., 254, 2340.[ISI][Medline]
Spencer, K. (1999) Second trimester prenatal screening for Down's syndrome using alpha-fetoprotein and free beta hCG: A seven year review. Br. J. Obstet. Gynecol., 106, 12871293[ISI][Medline]
Spencer, K., Muller, F. and Aitken, D.A. (1997) Biochemical markers of trisomy 21 in amniotic fluid. Prenat. Diagn., 17, 3137[ISI][Medline]
Suzumori, K., Tanemura, M., Murakami, I. et al. (1997) A retrospective evaluation of maternal serum screening for the detection of fetal aneuploidy. Prenat. Diagn., 17, 861866[ISI][Medline]
Taketa, K. (1995) Oncofetal alterations of AFP sugar chains. Jpn. J. Electroph., 39, 133136.
Taketa, K., Endo, Y., Sekiya, C. et al. (1993) A collaborative study for the evaluation of lectin-reactive -fetoprotein in early detection of hepatocellular carcinoma. Cancer Res., 53, 54195423[Abstract]
van Lith, J.M.M., Beekhuis, J.R., van Loon, A.J. et al. (1991) Alpha-fetoprotein in fetal serum, amniotic fluid, and maternal serum. Prenat. Diagn., 11, 625628.[ISI][Medline]
Van Lith, J.M.M., Pratt, J.J., Beekhuis, J.R. and Mantingh, A. (1992) Second-trimester maternal serum immunoreactive inhibin as a marker for Down's syndrome. Prenat. Diagn., 12, 801806[ISI][Medline]
Yamashita, K., Hitoi, A., Tsuchida, Y. et al. (1983) Sugar chain of -fetoprotein produced in human yolk sac tumor. Cancer Res., 43, 46914695[Abstract]
Yamashita, K., Taketa, K., Nishi, S. et al. (1993) Sugar chains of human cord serum -fetoprotein: Characteristics of N-linked sugar chains of glycoproteins produced in human liver and hepatocellular carcinomas. Cancer Res., 53, 29702975[Abstract]
Yamashita, F., Tanaka, M., Satomura, S. and Tanikawa, K. (1996) Prognostic significance of Lens culinaris agglutinin A-reactive -fetoprotein in small Hepatocellular carcinoma. Gastroenterology, 111, 9961001[ISI][Medline]
Yoshima, H., Mizuochi, T., Ishii, M. and Kobata, A. (1980) Structure of the asparagine linked sugar chains of -fetoprotein purified from human ascites fluid. Cancer Res., 40, 42764281.[Abstract]
Submitted on March 19, 2001; accepted on July 25, 2001.