- Email: [email protected]
The role of transferrin and ferritin in the fetal-maternalplacental unit
Nicolaides, Warenski, and Rodeck
4. Bowman JM. The management of Rh isoimmunization. Obstet Gynecol 1978;52: l. 5. Phibbs RH , Johnson P, Tooley WH. Cardiorespiratory status of erythroblastotic newborn infants. II . Blood vol ume, hematocrit, and serum albumin concentration in relation to hydrops fetalis. Pediatrics 1974;53: 13. 6. Rodeck CH, Campbell S. Sampling pure fetal blood by fetoscopy in the second trimester of pregnancy. Br Med J 1978;2:728. 7. Rodeck CH, Nicolaides KH. Ultrasound guided invasive procedures in obstetrics. Clin Obstet Gynaecol 1983; 10:529. 8. Rodeck CH, Holman CA, Karnicki J, Kemp JR, Whitmore DN, Austin MA. Direct intravascular fetal blood trans fusion by fetoscopy in severe rhesus isoimmunization. Lancet 1981;1 :625. 9. Rodeck CH, Nicolaides KH, Warsof SL, Fysh WJ, Gamsu HR, Kemp JR. The management of severe rhesus iso immunization by fetoscopic intravascular blood transfu sion. AM J 0BSTET GYNECOL 1984; 150:769. 10. Whitfield CR. A three year assessment of an Action Line method of timing intervention in rhesus isoimmunization. AMJ 0BSTETGYNECOL 1970;108:1239. ll. Deverilll. Kinetic measurements of the immunoprecipitin
June I, 1985 Am J Obstet Gynecol
12. 13. 14.
15. 16.
17. 18.
reaction using the centrifugal analyser. In: Price CP, Spen cer EVJ, eds. Centrifugal analyser in clinical chemistry. New York: Praeger Publishers, 1980: 109. Nicolaides KH, Rodeck CH, Miller D, Mibashan RS. Fetal haematology in rhesus isoimmunization. Br Med J 1985; I :661. Moniz C, Nicolaides KH, Rodeck CH. Biochemical indices in fetal blood in the second and third trimester of preg nancy. J Clin Pathol (in press). Glickman RM, Isselbacher KJ. Abdominal swelling and ascites. In : Petersdorf RG, ed. Harrison's principles of internal medicine. New York: McGraw-Hill Book Co., 1983:209. Baum JD, Harris D. Colloid osmotic pressure in eryth roblastosis fetalis. Br Med J 1972; 1:60 l. Adamsons K,James LS, Lucey JF, Towell ME. The effect of anemia upon cardiovascular performance and acid base state of the fetal rhesus monkey. Ann NY Acad Sci 1969; 162:225. Horger EO, Hutchinson DL. Drug-induced hydrops in the fetal lamb. Obstet Gynecol 1970;35:364. Machin GA. Differential diagnosis of hydrops fetalis. Am J Med Genet 1981;9:341.
The role of transferrin and ferritin in the fetal-maternal placental unit Teruaki Okuyama, M.D., Tetsuo Tawada, M.D., Hiroshi Furuya, M.D., and Claude A. Villee, Ph.D.
Boston, Massachusetts, and Tokyo, japan
The rapidly growing human fetus requires a large supply of iron, which is obtained from the iron stores of the mother. Iron is transported from mother to fetus against a concentration gradient. The placenta plays a key role in regulating the supply of iron in the fetus. In both anemic and nonanemic patients serum ferritin levels decreased and total iron-binding capacity increased as gestation progressed. The total iron-binding capacity is higher in maternal than in umbilical cord blood at delivery; this suggests that little or no ferritin or transferrin is transferred from mother to fetus. Mother and fetus appear to have independent systems controlling iron metabolism. Transferrin was localized on the site facing the intervillous space, on the surface of the microvilli of the syncytiotrophoblasts. The removal of transferrin from the surface of the trophoblast by thiocyanate and its rebinding were demonstrated. Ferritin was shown to be present in all layers of the trophoblast and especially in the syncytiotrophoblast. (AM J OBSTET GYNECOL
1985;152:344-50.)
Key words: Iron transport, placenta, ferritin, transferrin, immunochemistry, iron requirement of fetus The ubiquitous element iron plays key roles in oxy gen transport, in electron transport, and in many enFrom the Department of Biological Chemistry and the Laboratory of Human Reproduction and Reproductive Biology, Harvard Med ical School, and the Department ofObstetrics and Gynecology,jun tendo University School of Medicine. Receivedfor publication August 6, 1984; revised December 5, 1984; accepted january 16, 1985. Reprint requests: Claude A. Villee, Ph.D., LHRRB 501 , Harvard Medical School, 45 Shattuck St., Boston, MA 02115.
344
zymatic reactions. Transferrin, a protein that trans ports iron, is present in the yolk sac of the early embryo in a form that is immunologically similar to that present in the sera of adults.' The developing human fetus has an immense de mand for iron, requiring 250 to 300 mg of iron during gestation! Iron is transported one way from maternal blood to the fetus . The placenta, in addition to trans porting iron from mother to fetus against a concentra tion gradient,' serves as a storage depot for iron.
Volume 152 Number 3
Transferrin and ferritin in fetal-maternal-placental unit 345
In the present studies the authors examined the amounts and dynamics of transferrin and ferritin in the fetal-maternal-placental unit by means of biochem ical and immunohistologic methods.
Tissue
I
Fixed in Periodate-Lysine-Paraformaldhyde( PLP)
Paraffin sectior.
Material and methods Determination of hemoglobin volume, serum fer ritin level, and total iron-binding capacity in the blood of the pregnant woman at each stage of gestation. With 113 normal pregnant women at each stage of gestation as subjects, hemoglobin volume was determined by the cyanmethemoglobin method, serum ferritin level by radioimmunoassay (Spac-Ferritin Kit, Daiichi Radioiso tope) with use of antihuman liver ferritin antibody, and total iron-binding capacity by use of the Resomat-Fe Kit (Daiichi Radioisotope). Determination of serum iron level, serum ferritin level, and total iron-binding capacity in maternal blood and umbilical blood at delivery. From 35 normal parturient women, cubital venous blood was drawn at delivery and umbilical cord venous blood was collected immediately after delivery. The serum iron concen tration of each sample was determined by the baso phenanthroline method, and serum ferritin level and total iron-binding capacity were measured as described above. The localization of transferrin in placental villous tissue. A small cube of villous tissue was cut from the placenta immediately after normal term delivery, rinsed with physiologic saline solution to remove blood, fixed in a periodate-Iysine paraformaldehyde solution for 2 to 4 hours, and embedded in paraffin for sec tioning and observation under light microscopy. For electron microscopy the sample was frozen at - 60° C, and sections were cut. Thereafter transferrin was identified by the enzyme-conjugated antibody method (indirect method) with use of the method of Nakane and Kawaoi! For the first antibody, antihuman liver transferrin rabbit IgG: whole (Dako Corporation) was used. For the second antibody, peroxidase-labeled rabbit IgG and goat IgG: whole (Japan Antibody In stitute) was used . For the control, phosphate-buffered saline solution or nonimmunized rabbit serum was used in place of the first antibody, the nonspecific reaction was checked, and the specificity of the localization of antibody was confirmed. Thus the localization of trans ferrin in placental villous tissue was observed both by light microscopy and by electron microscopy (Fig. I). Small pieces of villous tissue were extracted imme diately from the placenta obtained after normal term delivery, rinsed to remove blood, and incubated for 10 minutes in a solution of 0.5 mol/L of ammonium thio cyanate according to the method of Galbraith et a!." (Fig. 2). Thereafter a part of the sample was fixed in I 0% formalin, and the remainder was rinsed in a buffer solution, incubated again in the sera of the maternal
Deparaffin Wash in PBS Apply rabbit antiserum to human transferrin for 30min . Wash in PBS
Frozen section Wash in PBS
Apply rabbit antiserum to human transferrin for overnight ( 4 'C ) Wash in PBS
Apply peroxodase labeled goat antirabbit lgG . for 2hrs . Wash in PBS
Apply peroxidase labeled goat
antirabbit lgG . for 30min .
Wash in PBS
Incubate in DAB solution for 30min
Incubate in OAB-H 2 0 2 Solution for IOmin. Wash in PBS
Incubate in OAB-H 2 0 2 Solution for 2min . Wash in PBS
Dehydrate
Incubate in 2 % Os-0' 0. I M PBS for 30min . Wash on PBS
Mount
Dehydrate LM observation
Embed
EM observat1on
Fig. 1. Localization of transferrin by the peroxidase-la beled antibody method (indirect enzyme-conjugated antibody method).
blood collected at delivery, rinsed again in a buffer solution, and fixed in formalin. Each sample was treated with the enzyme-conjugated antibody method described above, and the localization of transferrin was observed by light microscopy. Localization of ferritin in placental villous tissue. The enzyme-conjugated antibody method (indirect method) was performed similarly on placental vil lous tissue obtained at normal term delivery. For the first antibody, antihuman placental rabbit IgG: whole (Hoechst Corporation) was used. For the second anti body, peroxidase-labeled antirabbit IgG and goat IgG, F(ab')2: whole (Cappel Corporation) was used. The control test was similarly performed with transferrin, and the. localization of ferritin was observed by light microscopy. Determination of ferritin concentration in placental villous tissue. Thirty samples of placentas at each stage of gestation were used. After the decidua, the fetal mkmbranes, and the umbilical cord were removed, the placenta was cut into small pieces and rinsed several times in physiologic saline solution to remove blood. The fetal vasculature was also removed as completely as possible. To the villous tissue thus obtained, 1.5 times as large an amount of distilled water was added, and the placenta was homogenized. A part of the homog enate was used for protein determination, and the re mainder was centrifuged at 3000 gm for 30 minutes.
346 Okuyama et al.
June I , 1985 Am J Obstet Gynecol
Q
L£ '~"N" LJ ~
placental section
foriOmin . •
0. SM ammonium
- - - - - - - - - • Removal of Transferrin
l
Wash in PBS
Ul
thiocyanate Solution
•
u
incubate for IOmin .
Human Trans f err in Solut ion
~
Peroxifse antibody method
-Replacement of Transferr in
Wash in PBS
Fig. 2. Removal and replacement of transferrin in the trophoblast. Table I. Serum ferritin and related hematologic values in pregnant women Stage of gestation
Nonanemic group (hemoglobin ;;;.JJ gm/dl) First trimester Second trimester Third trimester Anemic group (hemoglobin < llg/dl) First trimester Second trimester Third trimester
N o. of women
Serum ferritin (nglml)
H emoglobin (gmldl)
Total iron-binding capacity (JLgldl)
32 17 17
34.1 ± 19.0 22.9 ± 12.2 19.4 ± 11.7
12.3 ± 0.7 11.8 ± 0.7 11.7±0.8
394 ± 51 434 ± 81 443 ± 53
10 15 22
15.1 ± 7.8 10.2 ± 6.7 5.2 ± 2.2
9.9 ± 1.1 9.9 ± 0.6 9.8 ± 1.1
464 ±II 501 ± 52 509 ± 82
Values are expressed as mean ± SD. Table II. Iron, ferritin, and total iron-binding capacity of maternal and fetal blood at delivery (n = 35) Iron (1-Lg/dl) Ferritin (ng/ml) Total iron-binding capacity (1-Lg/dl)
M atemal blood
Cord blood
59.8 ± 26.2* 9.6 ± 7.2* 495.2 ± 100.3*
161.5 ± 42.1* 160.5 ± 67.2* 177.7 ± 33.9*
*p < 0.05. Values are expressed as mean ± SD. The supernatant fraction was heated at 80° C for 10 minutes, then cooled suddenly, and again centrifuged at 3000 gm for 30 minutes. The supernatant thus ob tained was used for the determination of ferritin by the ferritin test (Research Laboratories, Yamonouchi Phar maceutical) with the use of antihuman placental ferritin antibody. Protein concentration was determined by the method of Lowry et al. 6
Results Hemoglobin volume, serum ferritin level, and total iron-binding capacity in the blood of the pregnant woman at successive stages of gestation. Pregnant
women at each stage of gestation were classified into a nonanemic group if hemoglobin volume was ;;;<11 gm/ dl and into an anemic group if hemoglobin volume was < 11 gm/dl. In both anemic and nonanemic patients the serum ferritin level decreased gradually with the progress of gestation whereas the total iron-binding capacity increased gradually with gestation (Table I). Serum iron level, serum ferritin level, an-i total iron-binding capacity in maternal blood and umbilical blood at delivery. Both the serum iron level and the serum ferritin level in umbilical blood were greater than in maternal blood, whereas the total iron-binding capacity in maternal blood was significantly greater than that in umbilical blood (p < 0.05) (Table II). There was no significant correlation between the con centrations of any of these constituents in maternal and fetal blood. Localization of transferrin in placental villous tis sue. Deposits of reaction products indicating the lo calization of transferrin were detected by light micros copy on the site facing the intervillous space over the surface of the trophoblast and by electron microscopy on the surface of microvilli of syncytiotrophoblasts (Figs. 3 and 4). In the incubation experiments conducted simulta
Transferrin and ferritin in fetal-maternal-placental unit 347
Volume 152 Number 3
I l I .
"'
.i!'
.f•"f "
~
' 1
Fig. 3. Localization of transferrin in placental villous tissue ( x 400). Reaction products of transferrin are found on the surface of the trophoblasts (arrow).
Fig. 4. The localization of transferrin in placental ·villo\}s tissue ( x 7200). Reaction products of transferrin are found on the surface of microvilli of syncytiotrophoblasts (arrow). N, Nucleus; IVS, intervillous space.
neously, the removal of transferrin from the surface of trophoblasts by treatment with thiocyanate and the sub sequent rebinding of transferrin to the surface of tro phoblasts were observed (Fig. 5). Localization of ferritin in placental villous tissue. Deposits of reaction products indicating the localization of ferritin were found ·in all ·layers of the trophoblast and especially in the syn~;;ytiotrophoblast (Fig. 6). Ferritin concentrations in placental villo~s tissue. The concentration of ferritin in placental villous tissue increased as gestation progressed from 0.07 ± .0.03 ~Jog/ mg protei~ in the first trimester to 0.18 ± 0.03 ~J.g/mg protein in the second and 0.25 ± 0.09 ~J.g/mg protein in the third trimester (Table III and Fig. 7). .
Comment Pommerenke' observed that radioactive iron admin istered to a pregnant woman was transferred to the fetus very quickly. In rabbits and in humans iron trans port is one-way from the mother to the offspring, and jn the latter stage of gestation iron is transferred against a concentration gradient. 3 In humans the amount of iron passing through the placenta increases as gestation progresses. 8 In rabbits iron is taken UP, by the placenta even if the umbilical cord is ligated and the traffic be tween the placenta and the fetus is interrupted. 9 In contrast, Woehler 10 found that a high dose of iron administered to pregnant rats did not result in an ex cessive amount of iron in the fetus.
348
Okuyama et al.
June I, 1985 Am J Obstet Gynecol
Fig. 5. A, Removal of transferrin with use of ammonium thiocyanate. Reaction products of transferrin are not visible on the surface of the trophoblasts. The surface seems smooth ( x 200). B, Replacement of transferrin by maternal serum transferrin. Reaction products of transferrin are evident on the surface of the trophoblasts.
Fig. 6. The localization of ferritin in placental villous tissue ( X 200). Reaction products of ferritin are present in all layers of the trophoblasts, especi
Table III. Concentration of ferritin in placental tissue at first, second, and third trimester Stage of gestation
First trimester
Second trimester
Third trimester
No. of cases Ferritin content (J.Lg/mg of protein)
10 0.07 ± 0.03*
5 0.18 ± 0.03*
15 0.25 ± 0.09*
*p < 0.05. Values are expressed as mean ± SD.
In animals with hemochorial placentas such as the human, rabbit, and rat, iron transport from mother to fetus may be one-way. Some specific mechanism may actively take up iron needed by the fetus against a con-
centration gradient and control the iron concentration so that there is no excess of iron in the fetus . Serum ferritin" has been considered to reflect the amount of iron stored in the body. Its concentration
Volume 152 Number 3
characteristically decreases as gestation progresses. The concentration of maternal hemoglobin is also greatly reduced in the latter part of pregnancy even in women who were not anemic. These findings indicate that iron is mobilized from the stored irc:~n pool in the mother as gestation pro gresses in order to meet the increasing demands for iron for hetn;1topoiesis in the fetus and placenta. The amount of stored iron decreases. Latent iron deficiency is probably always present in the mother, even if the hemoglobin level remains within the normal range. At delivery the umbilical cord blood had a signifi cantly higher content of serum iron than the maternal blood. This provides evidence that at least in the ter minal stage of gestation iron is actively transported from mother to child against a concentration gradient. The umbilical cord blood also contains a significantly higher amount of serum ferritin than does the maternal blood. The total iron-binding capacity was significantly higher in the maternal blood than in the umbilical cord blood. This indicates that there is little or no transfer of ferritin and transferrin between mother and child. The mother is in a state of iron deficiency, with low serum ferritin level and high total iron-binding capac ity, and the fetus is in a state of iron sufficiency, with high serum ferritin level and low iron-binding capacity. Thus iron Storage in the mother and ircin metabolism in the fetus are not directly related. !his suggests that mother and fetus have independent systems for the control of iron metabolism. The transport of iron from mother to fetus by the placenta may be under some separate controlling mechanism. Iron is present in maternal serum mainly bound to transferrin and ferritin. When a large volume of fer ritin was loaded on placental villous tissue, ferritin was incorporated into the placenta by pinocytosis in the trophoblasts. 12 However, the amount of serum ferritin in maternal blood is small, and its iron content is also very small. Even though the turnover of iron bound to ferritin is very rapid, the significance of serum ferritin as the source of iron to the fetus may be minimal. The major source of iron to the fetus is the iron bound to transferrin in maternal blood. Our present studies of the localization of transferrin in placental villous tissue by the enzyme-conjugated an tibody method confirmed its localization on the micro villi of the syncytiotrophoblast. The subsequent incu bation experiments showed that this was identical to the transferrin in maternal blood. Receptors for transferrin are present on the surface of placental villous epithelial cells.' This may indicate that transferrin in maternal blood binds specifically to the surface of the microvilli of syncytiotrophoblasts as the first step in the mechanism by which iron is trans ported across the placenta.
Transferrin and ferritin in fetal-maternal-placental unit 349
0.40
c
·a;
"§
Q.
•
0.35
••
0.30
•
•• • •• •
llD
E ....._
llD
3
2c:
c0 4>
•
0. 25
c: :;::; ·c
•• •
•
0.20
0
•
0.15
~
•
•
••
0.10
••• • • •
0.05 (
I
4
I
I
I
I
l
I
I
I
.!
I
8 12 16 20 24 28 32 36 40 42 gestational weeks
Fig. 7. Concentration of ferritin in placental tissues.
The process by which iron is incorporated from the transferrin in maternal blood bound on the surface of microvilli into the cytoplasm of the synEytiotrophoblast remains unclear. Harding et al. 13 demonstrated in rat reticulocytes that the transferrin bound tci transferrin receptors on the cell membrane was incorporated into the cytoplasm by pinocytosis. Thus they demonstrated the existence of receptor-mediated endocytosis, both biochemically and morphologically. We .suggest that a similar mechanism exists on the syncytiotrophoblast of the placenta. It has been maintained 2 that the amount of trans ferrin transported from maternal to fetal blood is very small compared to the amount of iron transported. Therefore the transferrin incorporated into syncytio ti-ophoblasts is considered to transfer iron to other me diators rather than being transferred completely across the placenta to the fetal blood. Oski and Naiman 14 re ported that iron bound to transferrin in maternal blood was incorporated into trophoblasts first and then was bound to ferritin. Our experiments indicate that the ferritin concentration in the placenta increased with the progress of gestation. The amount of iron passing daily through the placenta is increased as gestation pro gresses.9Our experiments show that ferritin is localized in the entire trophoblast. These findings indicate that the amount of ferritin within the trophoblasts rose with the increase in iron transport and that ferritin plays an important role in iron transport across the placenta. Ferritin consists of H and L subunits. Acidic ferritin, in which the H sub unit is dominant, can incorporate iron more efficiently
350
Okuyama et al.
ahd can meet the demand for iron ions more rapidly. Basic ferritin , in which the L subunit is dominant, is more stable metabolically and is associated with the stor age of iron. The ratio of H and L subunits varies with the amount of iron in the tissue. When an adequate amount of iron is present, ihe L subunit becomes dom inant, whereas in iron deficiency the H subunit becomes dominant. Ferritin may be involved not only in the storage of surplus i~on but also more actively in iron metabolism than had been considered previously. The ferritin in human placenta is acidic ferritin with the H subunit dominant. 15 Acidic ferritin characteris tically incorporates iron efficiently and responds quickly to the demands for iron. When an excess amount of iron was administered to rats, their placenta ferritin became basic and the L subunit became dom inant. It can be postulated from these findings that human placental ferritin has the two functions of trans poriing irCil} when it is defiolism, regulating the transport of iron be tween mother and fetus and ensuring a constant supply of iron to the feius. REFERENCES 1. Giblin D, Perricelli A. Synthesis of serum albumin, preal bumin, a-fetoprotein, a 1-antitrypsin and transferrin by the human yolk sac. Nature 1970;228:995. 2. Burman D. Iron in biochemistry and medicine. New York: Academic Press, 1974:543.
. June I , 1985 Am J Obstet Gynecol
3. Furuya H. Studies of anemia ' in pregnant women. Acta Obstet Gynecol 1965; 17:683. 4. Nakane PK, Kawa:oi A. Peroxidase labeled antibody: a new method of conjugation. J Histochem Cytochem 1974;22 : 1084.. . 5. Galbraith GMP, Galbraith RM, Temple A,.faulk WF. Demonstration •pf transferrin receptors on human pla cental ttophoblasts. Blood 1980;55:240. 6. Lowry OH, Rosenbrough NJ , Farf AL, Randall .RJ. PI;O tein measurement with the Folin phenol reagenL J Bioi Chern 1951 ; 193:265 . . . 7. Pommerenke WT. Transmission of radioactive iron to the human fetus .. Amj Physioi 1942;137 :164: . 8. FletcherJ , Suter PEN. The transport of iron by the humari placenta. Clin Sci 1969;36:209. 9. Bothwell TH, Pribilla WF, Mebust W; Finch CA. Iron metabolism in the pregnant rabbi.t: iron transport across the placenta. Am J Physiol 1958;193:615. · 10. Woehler F. Intermediary iron metabolism of the placenta with special consideration of the transport of therapeu tically administered iron through this organ. Curr Ther Res 1964;6:464. 11. Walters GO, Miller TM, Worwood M. Serum ferritin con centration and iron stores in normal subjects. J Clin Pathol 1973;26:770. 12. Yamaguchi R. Studies of placental physiology, concerning fetomaternal transport and metabolism. Acta Obstet Gy necol 1963; 15:743. 13. Harding C, Heuser J, Stahl P. Receptor mediated endo cytosis of transferrin and recycling of the transferrin re ceptor in rat reticulocytes. J Cell Bioi 1983;97:329. 14. Oski FA, Naiman JL. Hematologic problem in the new born. Clin Pediatr 1972;4: l. 15. Brown PJ, Johnson PM, Ogbimi AO, Tappin JA. Char acterization and localization of human placental ferritin . Biochemj 1979; 182:763.
4. Bowman JM. The management of Rh isoimmunization. Obstet Gynecol 1978;52: l. 5. Phibbs RH , Johnson P, Tooley WH. Cardiorespiratory status of erythroblastotic newborn infants. II . Blood vol ume, hematocrit, and serum albumin concentration in relation to hydrops fetalis. Pediatrics 1974;53: 13. 6. Rodeck CH, Campbell S. Sampling pure fetal blood by fetoscopy in the second trimester of pregnancy. Br Med J 1978;2:728. 7. Rodeck CH, Nicolaides KH. Ultrasound guided invasive procedures in obstetrics. Clin Obstet Gynaecol 1983; 10:529. 8. Rodeck CH, Holman CA, Karnicki J, Kemp JR, Whitmore DN, Austin MA. Direct intravascular fetal blood trans fusion by fetoscopy in severe rhesus isoimmunization. Lancet 1981;1 :625. 9. Rodeck CH, Nicolaides KH, Warsof SL, Fysh WJ, Gamsu HR, Kemp JR. The management of severe rhesus iso immunization by fetoscopic intravascular blood transfu sion. AM J 0BSTET GYNECOL 1984; 150:769. 10. Whitfield CR. A three year assessment of an Action Line method of timing intervention in rhesus isoimmunization. AMJ 0BSTETGYNECOL 1970;108:1239. ll. Deverilll. Kinetic measurements of the immunoprecipitin
June I, 1985 Am J Obstet Gynecol
12. 13. 14.
15. 16.
17. 18.
reaction using the centrifugal analyser. In: Price CP, Spen cer EVJ, eds. Centrifugal analyser in clinical chemistry. New York: Praeger Publishers, 1980: 109. Nicolaides KH, Rodeck CH, Miller D, Mibashan RS. Fetal haematology in rhesus isoimmunization. Br Med J 1985; I :661. Moniz C, Nicolaides KH, Rodeck CH. Biochemical indices in fetal blood in the second and third trimester of preg nancy. J Clin Pathol (in press). Glickman RM, Isselbacher KJ. Abdominal swelling and ascites. In : Petersdorf RG, ed. Harrison's principles of internal medicine. New York: McGraw-Hill Book Co., 1983:209. Baum JD, Harris D. Colloid osmotic pressure in eryth roblastosis fetalis. Br Med J 1972; 1:60 l. Adamsons K,James LS, Lucey JF, Towell ME. The effect of anemia upon cardiovascular performance and acid base state of the fetal rhesus monkey. Ann NY Acad Sci 1969; 162:225. Horger EO, Hutchinson DL. Drug-induced hydrops in the fetal lamb. Obstet Gynecol 1970;35:364. Machin GA. Differential diagnosis of hydrops fetalis. Am J Med Genet 1981;9:341.
The role of transferrin and ferritin in the fetal-maternal placental unit Teruaki Okuyama, M.D., Tetsuo Tawada, M.D., Hiroshi Furuya, M.D., and Claude A. Villee, Ph.D.
Boston, Massachusetts, and Tokyo, japan
The rapidly growing human fetus requires a large supply of iron, which is obtained from the iron stores of the mother. Iron is transported from mother to fetus against a concentration gradient. The placenta plays a key role in regulating the supply of iron in the fetus. In both anemic and nonanemic patients serum ferritin levels decreased and total iron-binding capacity increased as gestation progressed. The total iron-binding capacity is higher in maternal than in umbilical cord blood at delivery; this suggests that little or no ferritin or transferrin is transferred from mother to fetus. Mother and fetus appear to have independent systems controlling iron metabolism. Transferrin was localized on the site facing the intervillous space, on the surface of the microvilli of the syncytiotrophoblasts. The removal of transferrin from the surface of the trophoblast by thiocyanate and its rebinding were demonstrated. Ferritin was shown to be present in all layers of the trophoblast and especially in the syncytiotrophoblast. (AM J OBSTET GYNECOL
1985;152:344-50.)
Key words: Iron transport, placenta, ferritin, transferrin, immunochemistry, iron requirement of fetus The ubiquitous element iron plays key roles in oxy gen transport, in electron transport, and in many enFrom the Department of Biological Chemistry and the Laboratory of Human Reproduction and Reproductive Biology, Harvard Med ical School, and the Department ofObstetrics and Gynecology,jun tendo University School of Medicine. Receivedfor publication August 6, 1984; revised December 5, 1984; accepted january 16, 1985. Reprint requests: Claude A. Villee, Ph.D., LHRRB 501 , Harvard Medical School, 45 Shattuck St., Boston, MA 02115.
344
zymatic reactions. Transferrin, a protein that trans ports iron, is present in the yolk sac of the early embryo in a form that is immunologically similar to that present in the sera of adults.' The developing human fetus has an immense de mand for iron, requiring 250 to 300 mg of iron during gestation! Iron is transported one way from maternal blood to the fetus . The placenta, in addition to trans porting iron from mother to fetus against a concentra tion gradient,' serves as a storage depot for iron.
Volume 152 Number 3
Transferrin and ferritin in fetal-maternal-placental unit 345
In the present studies the authors examined the amounts and dynamics of transferrin and ferritin in the fetal-maternal-placental unit by means of biochem ical and immunohistologic methods.
Tissue
I
Fixed in Periodate-Lysine-Paraformaldhyde( PLP)
Paraffin sectior.
Material and methods Determination of hemoglobin volume, serum fer ritin level, and total iron-binding capacity in the blood of the pregnant woman at each stage of gestation. With 113 normal pregnant women at each stage of gestation as subjects, hemoglobin volume was determined by the cyanmethemoglobin method, serum ferritin level by radioimmunoassay (Spac-Ferritin Kit, Daiichi Radioiso tope) with use of antihuman liver ferritin antibody, and total iron-binding capacity by use of the Resomat-Fe Kit (Daiichi Radioisotope). Determination of serum iron level, serum ferritin level, and total iron-binding capacity in maternal blood and umbilical blood at delivery. From 35 normal parturient women, cubital venous blood was drawn at delivery and umbilical cord venous blood was collected immediately after delivery. The serum iron concen tration of each sample was determined by the baso phenanthroline method, and serum ferritin level and total iron-binding capacity were measured as described above. The localization of transferrin in placental villous tissue. A small cube of villous tissue was cut from the placenta immediately after normal term delivery, rinsed with physiologic saline solution to remove blood, fixed in a periodate-Iysine paraformaldehyde solution for 2 to 4 hours, and embedded in paraffin for sec tioning and observation under light microscopy. For electron microscopy the sample was frozen at - 60° C, and sections were cut. Thereafter transferrin was identified by the enzyme-conjugated antibody method (indirect method) with use of the method of Nakane and Kawaoi! For the first antibody, antihuman liver transferrin rabbit IgG: whole (Dako Corporation) was used. For the second antibody, peroxidase-labeled rabbit IgG and goat IgG: whole (Japan Antibody In stitute) was used . For the control, phosphate-buffered saline solution or nonimmunized rabbit serum was used in place of the first antibody, the nonspecific reaction was checked, and the specificity of the localization of antibody was confirmed. Thus the localization of trans ferrin in placental villous tissue was observed both by light microscopy and by electron microscopy (Fig. I). Small pieces of villous tissue were extracted imme diately from the placenta obtained after normal term delivery, rinsed to remove blood, and incubated for 10 minutes in a solution of 0.5 mol/L of ammonium thio cyanate according to the method of Galbraith et a!." (Fig. 2). Thereafter a part of the sample was fixed in I 0% formalin, and the remainder was rinsed in a buffer solution, incubated again in the sera of the maternal
Deparaffin Wash in PBS Apply rabbit antiserum to human transferrin for 30min . Wash in PBS
Frozen section Wash in PBS
Apply rabbit antiserum to human transferrin for overnight ( 4 'C ) Wash in PBS
Apply peroxodase labeled goat antirabbit lgG . for 2hrs . Wash in PBS
Apply peroxidase labeled goat
antirabbit lgG . for 30min .
Wash in PBS
Incubate in DAB solution for 30min
Incubate in OAB-H 2 0 2 Solution for IOmin. Wash in PBS
Incubate in OAB-H 2 0 2 Solution for 2min . Wash in PBS
Dehydrate
Incubate in 2 % Os-0' 0. I M PBS for 30min . Wash on PBS
Mount
Dehydrate LM observation
Embed
EM observat1on
Fig. 1. Localization of transferrin by the peroxidase-la beled antibody method (indirect enzyme-conjugated antibody method).
blood collected at delivery, rinsed again in a buffer solution, and fixed in formalin. Each sample was treated with the enzyme-conjugated antibody method described above, and the localization of transferrin was observed by light microscopy. Localization of ferritin in placental villous tissue. The enzyme-conjugated antibody method (indirect method) was performed similarly on placental vil lous tissue obtained at normal term delivery. For the first antibody, antihuman placental rabbit IgG: whole (Hoechst Corporation) was used. For the second anti body, peroxidase-labeled antirabbit IgG and goat IgG, F(ab')2: whole (Cappel Corporation) was used. The control test was similarly performed with transferrin, and the. localization of ferritin was observed by light microscopy. Determination of ferritin concentration in placental villous tissue. Thirty samples of placentas at each stage of gestation were used. After the decidua, the fetal mkmbranes, and the umbilical cord were removed, the placenta was cut into small pieces and rinsed several times in physiologic saline solution to remove blood. The fetal vasculature was also removed as completely as possible. To the villous tissue thus obtained, 1.5 times as large an amount of distilled water was added, and the placenta was homogenized. A part of the homog enate was used for protein determination, and the re mainder was centrifuged at 3000 gm for 30 minutes.
346 Okuyama et al.
June I , 1985 Am J Obstet Gynecol
Q
L£ '~"N" LJ ~
placental section
foriOmin . •
0. SM ammonium
- - - - - - - - - • Removal of Transferrin
l
Wash in PBS
Ul
thiocyanate Solution
•
u
incubate for IOmin .
Human Trans f err in Solut ion
~
Peroxifse antibody method
-Replacement of Transferr in
Wash in PBS
Fig. 2. Removal and replacement of transferrin in the trophoblast. Table I. Serum ferritin and related hematologic values in pregnant women Stage of gestation
Nonanemic group (hemoglobin ;;;.JJ gm/dl) First trimester Second trimester Third trimester Anemic group (hemoglobin < llg/dl) First trimester Second trimester Third trimester
N o. of women
Serum ferritin (nglml)
H emoglobin (gmldl)
Total iron-binding capacity (JLgldl)
32 17 17
34.1 ± 19.0 22.9 ± 12.2 19.4 ± 11.7
12.3 ± 0.7 11.8 ± 0.7 11.7±0.8
394 ± 51 434 ± 81 443 ± 53
10 15 22
15.1 ± 7.8 10.2 ± 6.7 5.2 ± 2.2
9.9 ± 1.1 9.9 ± 0.6 9.8 ± 1.1
464 ±II 501 ± 52 509 ± 82
Values are expressed as mean ± SD. Table II. Iron, ferritin, and total iron-binding capacity of maternal and fetal blood at delivery (n = 35) Iron (1-Lg/dl) Ferritin (ng/ml) Total iron-binding capacity (1-Lg/dl)
M atemal blood
Cord blood
59.8 ± 26.2* 9.6 ± 7.2* 495.2 ± 100.3*
161.5 ± 42.1* 160.5 ± 67.2* 177.7 ± 33.9*
*p < 0.05. Values are expressed as mean ± SD. The supernatant fraction was heated at 80° C for 10 minutes, then cooled suddenly, and again centrifuged at 3000 gm for 30 minutes. The supernatant thus ob tained was used for the determination of ferritin by the ferritin test (Research Laboratories, Yamonouchi Phar maceutical) with the use of antihuman placental ferritin antibody. Protein concentration was determined by the method of Lowry et al. 6
Results Hemoglobin volume, serum ferritin level, and total iron-binding capacity in the blood of the pregnant woman at successive stages of gestation. Pregnant
women at each stage of gestation were classified into a nonanemic group if hemoglobin volume was ;;;<11 gm/ dl and into an anemic group if hemoglobin volume was < 11 gm/dl. In both anemic and nonanemic patients the serum ferritin level decreased gradually with the progress of gestation whereas the total iron-binding capacity increased gradually with gestation (Table I). Serum iron level, serum ferritin level, an-i total iron-binding capacity in maternal blood and umbilical blood at delivery. Both the serum iron level and the serum ferritin level in umbilical blood were greater than in maternal blood, whereas the total iron-binding capacity in maternal blood was significantly greater than that in umbilical blood (p < 0.05) (Table II). There was no significant correlation between the con centrations of any of these constituents in maternal and fetal blood. Localization of transferrin in placental villous tis sue. Deposits of reaction products indicating the lo calization of transferrin were detected by light micros copy on the site facing the intervillous space over the surface of the trophoblast and by electron microscopy on the surface of microvilli of syncytiotrophoblasts (Figs. 3 and 4). In the incubation experiments conducted simulta
Transferrin and ferritin in fetal-maternal-placental unit 347
Volume 152 Number 3
I l I .
"'
.i!'
.f•"f "
~
' 1
Fig. 3. Localization of transferrin in placental villous tissue ( x 400). Reaction products of transferrin are found on the surface of the trophoblasts (arrow).
Fig. 4. The localization of transferrin in placental ·villo\}s tissue ( x 7200). Reaction products of transferrin are found on the surface of microvilli of syncytiotrophoblasts (arrow). N, Nucleus; IVS, intervillous space.
neously, the removal of transferrin from the surface of trophoblasts by treatment with thiocyanate and the sub sequent rebinding of transferrin to the surface of tro phoblasts were observed (Fig. 5). Localization of ferritin in placental villous tissue. Deposits of reaction products indicating the localization of ferritin were found ·in all ·layers of the trophoblast and especially in the syn~;;ytiotrophoblast (Fig. 6). Ferritin concentrations in placental villo~s tissue. The concentration of ferritin in placental villous tissue increased as gestation progressed from 0.07 ± .0.03 ~Jog/ mg protei~ in the first trimester to 0.18 ± 0.03 ~J.g/mg protein in the second and 0.25 ± 0.09 ~J.g/mg protein in the third trimester (Table III and Fig. 7). .
Comment Pommerenke' observed that radioactive iron admin istered to a pregnant woman was transferred to the fetus very quickly. In rabbits and in humans iron trans port is one-way from the mother to the offspring, and jn the latter stage of gestation iron is transferred against a concentration gradient. 3 In humans the amount of iron passing through the placenta increases as gestation progresses. 8 In rabbits iron is taken UP, by the placenta even if the umbilical cord is ligated and the traffic be tween the placenta and the fetus is interrupted. 9 In contrast, Woehler 10 found that a high dose of iron administered to pregnant rats did not result in an ex cessive amount of iron in the fetus.
348
Okuyama et al.
June I, 1985 Am J Obstet Gynecol
Fig. 5. A, Removal of transferrin with use of ammonium thiocyanate. Reaction products of transferrin are not visible on the surface of the trophoblasts. The surface seems smooth ( x 200). B, Replacement of transferrin by maternal serum transferrin. Reaction products of transferrin are evident on the surface of the trophoblasts.
Fig. 6. The localization of ferritin in placental villous tissue ( X 200). Reaction products of ferritin are present in all layers of the trophoblasts, especi
Table III. Concentration of ferritin in placental tissue at first, second, and third trimester Stage of gestation
First trimester
Second trimester
Third trimester
No. of cases Ferritin content (J.Lg/mg of protein)
10 0.07 ± 0.03*
5 0.18 ± 0.03*
15 0.25 ± 0.09*
*p < 0.05. Values are expressed as mean ± SD.
In animals with hemochorial placentas such as the human, rabbit, and rat, iron transport from mother to fetus may be one-way. Some specific mechanism may actively take up iron needed by the fetus against a con-
centration gradient and control the iron concentration so that there is no excess of iron in the fetus . Serum ferritin" has been considered to reflect the amount of iron stored in the body. Its concentration
Volume 152 Number 3
characteristically decreases as gestation progresses. The concentration of maternal hemoglobin is also greatly reduced in the latter part of pregnancy even in women who were not anemic. These findings indicate that iron is mobilized from the stored irc:~n pool in the mother as gestation pro gresses in order to meet the increasing demands for iron for hetn;1topoiesis in the fetus and placenta. The amount of stored iron decreases. Latent iron deficiency is probably always present in the mother, even if the hemoglobin level remains within the normal range. At delivery the umbilical cord blood had a signifi cantly higher content of serum iron than the maternal blood. This provides evidence that at least in the ter minal stage of gestation iron is actively transported from mother to child against a concentration gradient. The umbilical cord blood also contains a significantly higher amount of serum ferritin than does the maternal blood. The total iron-binding capacity was significantly higher in the maternal blood than in the umbilical cord blood. This indicates that there is little or no transfer of ferritin and transferrin between mother and child. The mother is in a state of iron deficiency, with low serum ferritin level and high total iron-binding capac ity, and the fetus is in a state of iron sufficiency, with high serum ferritin level and low iron-binding capacity. Thus iron Storage in the mother and ircin metabolism in the fetus are not directly related. !his suggests that mother and fetus have independent systems for the control of iron metabolism. The transport of iron from mother to fetus by the placenta may be under some separate controlling mechanism. Iron is present in maternal serum mainly bound to transferrin and ferritin. When a large volume of fer ritin was loaded on placental villous tissue, ferritin was incorporated into the placenta by pinocytosis in the trophoblasts. 12 However, the amount of serum ferritin in maternal blood is small, and its iron content is also very small. Even though the turnover of iron bound to ferritin is very rapid, the significance of serum ferritin as the source of iron to the fetus may be minimal. The major source of iron to the fetus is the iron bound to transferrin in maternal blood. Our present studies of the localization of transferrin in placental villous tissue by the enzyme-conjugated an tibody method confirmed its localization on the micro villi of the syncytiotrophoblast. The subsequent incu bation experiments showed that this was identical to the transferrin in maternal blood. Receptors for transferrin are present on the surface of placental villous epithelial cells.' This may indicate that transferrin in maternal blood binds specifically to the surface of the microvilli of syncytiotrophoblasts as the first step in the mechanism by which iron is trans ported across the placenta.
Transferrin and ferritin in fetal-maternal-placental unit 349
0.40
c
·a;
"§
Q.
•
0.35
••
0.30
•
•• • •• •
llD
E ....._
llD
3
2c:
c0 4>
•
0. 25
c: :;::; ·c
•• •
•
0.20
0
•
0.15
~
•
•
••
0.10
••• • • •
0.05 (
I
4
I
I
I
I
l
I
I
I
.!
I
8 12 16 20 24 28 32 36 40 42 gestational weeks
Fig. 7. Concentration of ferritin in placental tissues.
The process by which iron is incorporated from the transferrin in maternal blood bound on the surface of microvilli into the cytoplasm of the synEytiotrophoblast remains unclear. Harding et al. 13 demonstrated in rat reticulocytes that the transferrin bound tci transferrin receptors on the cell membrane was incorporated into the cytoplasm by pinocytosis. Thus they demonstrated the existence of receptor-mediated endocytosis, both biochemically and morphologically. We .suggest that a similar mechanism exists on the syncytiotrophoblast of the placenta. It has been maintained 2 that the amount of trans ferrin transported from maternal to fetal blood is very small compared to the amount of iron transported. Therefore the transferrin incorporated into syncytio ti-ophoblasts is considered to transfer iron to other me diators rather than being transferred completely across the placenta to the fetal blood. Oski and Naiman 14 re ported that iron bound to transferrin in maternal blood was incorporated into trophoblasts first and then was bound to ferritin. Our experiments indicate that the ferritin concentration in the placenta increased with the progress of gestation. The amount of iron passing daily through the placenta is increased as gestation pro gresses.9Our experiments show that ferritin is localized in the entire trophoblast. These findings indicate that the amount of ferritin within the trophoblasts rose with the increase in iron transport and that ferritin plays an important role in iron transport across the placenta. Ferritin consists of H and L subunits. Acidic ferritin, in which the H sub unit is dominant, can incorporate iron more efficiently
350
Okuyama et al.
ahd can meet the demand for iron ions more rapidly. Basic ferritin , in which the L subunit is dominant, is more stable metabolically and is associated with the stor age of iron. The ratio of H and L subunits varies with the amount of iron in the tissue. When an adequate amount of iron is present, ihe L subunit becomes dom inant, whereas in iron deficiency the H subunit becomes dominant. Ferritin may be involved not only in the storage of surplus i~on but also more actively in iron metabolism than had been considered previously. The ferritin in human placenta is acidic ferritin with the H subunit dominant. 15 Acidic ferritin characteris tically incorporates iron efficiently and responds quickly to the demands for iron. When an excess amount of iron was administered to rats, their placenta ferritin became basic and the L subunit became dom inant. It can be postulated from these findings that human placental ferritin has the two functions of trans poriing irCil} when it is defiolism, regulating the transport of iron be tween mother and fetus and ensuring a constant supply of iron to the feius. REFERENCES 1. Giblin D, Perricelli A. Synthesis of serum albumin, preal bumin, a-fetoprotein, a 1-antitrypsin and transferrin by the human yolk sac. Nature 1970;228:995. 2. Burman D. Iron in biochemistry and medicine. New York: Academic Press, 1974:543.
. June I , 1985 Am J Obstet Gynecol
3. Furuya H. Studies of anemia ' in pregnant women. Acta Obstet Gynecol 1965; 17:683. 4. Nakane PK, Kawa:oi A. Peroxidase labeled antibody: a new method of conjugation. J Histochem Cytochem 1974;22 : 1084.. . 5. Galbraith GMP, Galbraith RM, Temple A,.faulk WF. Demonstration •pf transferrin receptors on human pla cental ttophoblasts. Blood 1980;55:240. 6. Lowry OH, Rosenbrough NJ , Farf AL, Randall .RJ. PI;O tein measurement with the Folin phenol reagenL J Bioi Chern 1951 ; 193:265 . . . 7. Pommerenke WT. Transmission of radioactive iron to the human fetus .. Amj Physioi 1942;137 :164: . 8. FletcherJ , Suter PEN. The transport of iron by the humari placenta. Clin Sci 1969;36:209. 9. Bothwell TH, Pribilla WF, Mebust W; Finch CA. Iron metabolism in the pregnant rabbi.t: iron transport across the placenta. Am J Physiol 1958;193:615. · 10. Woehler F. Intermediary iron metabolism of the placenta with special consideration of the transport of therapeu tically administered iron through this organ. Curr Ther Res 1964;6:464. 11. Walters GO, Miller TM, Worwood M. Serum ferritin con centration and iron stores in normal subjects. J Clin Pathol 1973;26:770. 12. Yamaguchi R. Studies of placental physiology, concerning fetomaternal transport and metabolism. Acta Obstet Gy necol 1963; 15:743. 13. Harding C, Heuser J, Stahl P. Receptor mediated endo cytosis of transferrin and recycling of the transferrin re ceptor in rat reticulocytes. J Cell Bioi 1983;97:329. 14. Oski FA, Naiman JL. Hematologic problem in the new born. Clin Pediatr 1972;4: l. 15. Brown PJ, Johnson PM, Ogbimi AO, Tappin JA. Char acterization and localization of human placental ferritin . Biochemj 1979; 182:763.