The mixture was incubated for 90 min at room temperature. ..... studies of Fissore et ai (1989) which showed that injection of EDTA in the perivitelline space.
How does transferrin overcome the in vitro block to development of the mouse preimplantation embryo? M. H. Nasr-Esfahani and M. H. Johnson
'Department of Anatomy, Downing Street, Cambridge CB2 3D Y; and2Assisted Conception Research Unit,
Department of Obstetrics and Gynaecology, St Thomas' Hospital, London SEI 7EH, UK
Summary. Transferrin promotes development of mouse embryos through the two-cell block in vitro. Uptake of transferrin into blastocysts was shown to occur by both receptor-mediated and nonspecific pathways, but neither pathway was used to a detectable extent by embryos before the eight-cell stage. Conversely, the dialysis of culture medium, non-permissive for development through the two-cell block, against a solution of transferrin rendered it capable of supporting development. It was therefore concluded that transferrin exerts its supportive effect on development in vitro via its chelating effects. Keywords: hydrogen peroxide; transferrin; embryo; reactive oxygen species; mouse
Introduction
Embryos from most outbred and inbred mouse strains do not develop to blastocysts when cultured a chemically defined medium but arrest at the late two-cell stage when there is a rise in reactive oxygen species such as hydrogen peroxide or lipid peroxides. This rise in reactive oxygen species is not observed in embryos of the same age recovered directly from the female tract (Nasr-Esfahani et ai, 1990a). Hydrogen peroxide appears to result from dismutation of Superoxides, produced via the xanthine-xanthine-oxidase system (Nasr-Esfahani & Johnson, 1991). The rise in concentration of reactive oxygen species appears to be a response to in vitro culture that depends on the stage of development and might be related in some way to the two-cell block. Reactive oxygen species such as H,02 in conjunction with Superoxide (02·) can cause cell damage by promoting hydroxy radical (HO·) formation, a reaction catalysed by transitional metals such as iron (Halliwell, 1987; Halliwell & Gutteridge, 1990). Iron can also act directly on lipids to magnify peroxidative damage once this has been initiated by free hydroxy radicals (Minotti & Aust, 1989). We have previously shown that the addition of certain iron chelators such as bovine apo-transferrin (BAT), ethylene diamine tetra acetic acid (EDTA) and diethylene triamine-penta acetic acid (DETAPAC) to the culture medium promotes development through the two-cell block, whereas addition of iron retards development (Nasr-Esfahani et ai, 1990b; Nasr-Esfahani & Johnson, 1991). Bovine apo¬ in
transferrin
was
the most effective agent
over a
wide concentration range
(1-1-22-2 µ
-1) and
required only immediately after recovery from the oviduct until the end of the two-cell stage, precisely the period over which the two-cell block and the peroxide rise occur. However, the addition of bovine apo-transferrin to T6 medium (BAT6 medium) had no effect on the increase in peroxide concentration in vitro during the two-cell stage (Nasr-Esfahani et ai, 1990b). Three possible mechanisms of action of transferrin were suggested: (i) transferrin sequesters free traces of iron and thereby promotes development through the two-cell block via inhibition of hydroxy radical formation and lipid peroxidation; (ii) transferrin facilitates the was
Corresponding author.
transport of iron (and possibly other transitional metals) into embryos, thus overcoming a two-cell
block due to iron deficiency (Rudland et ai, 1977; Yabe et ai, 1987), or (iii) transferrin serves as a growth factor in a manner that is independent of its iron carrying properties (Sanders, 1986; Shapiro & Wagner, 1989; Clover et ai, 1991).
Materials and Methods
Embryo handling MF1 female mice, 3-4 weeks old (Olac, Bicester, UK) were superovulated by intraperitoneal injection of 5 or 10 iu of pregnant mares' serum gonadotrophin (PMS; Intervet, Cambridge, UK) at midday followed 48 h later by 5 or 10 iu of human chorionic gonadotrophin (hCG; Intervet). One-cell fertilized zygotes were obtained by pairing each female overnight with an HC-CFLP male (Interfauna, Wyton-Huntingdon, UK) and females were inspected for vaginal plugs the next day as an indication of successful mating. Fertilized oocytes at the pronuclear stage were recovered from mated females at between 26 and 30 h after hCG treatment by release from the oviduct into warmed medium H6 plus BSA (a modified form of T6 medium; Nasr-Esfahani el ai, 1990b). Two-, four- and eight-cell embryos and blastocysts were recovered at 36-48, 50-54, 69-70 and 96 h after hCG, respectively. Embryos were released from the oviduct into warmed H6 plus BSA under paraffin oil (FSA Laboratories, Loughborough, UK) in Falcon tissue culture dishes. All manipulations were carried out at 37°C on heated stages, pads or in incubators. Zonae were removed by acid Tyrode's solution.
Scoring of embryos Zygotes were cultured in medium T6 plus BSA in 5% CO, in air and inspected at regular intervals as follows: at 47-50 h after hCG treatment the few one-cell or abnormal oocytes and zygotes (0-3%) were removed and were not included in the totals; the remaining embryos were scored as two-, three- or four-cell. At 69-72 h after hCG treatment embryos were scored as dead, two-cell, three-cell, four-cell, five- to eight-cell precompact, or compact. At 98-100 h after hCG treatment embryos were scored as dead, noncompact, compact, early blastocyst or expanded blastocyst (see Chisholm et ai, 1985, for definitions of blastocyst subtypes); at 116-119 h after hCG treatment embryos were scored as dead, preblastocyst, early blastocyst or expanded blastocyst. Chemicals and reagents
Dialysis tubing was obtained from Media Scientific (London, UK). EDTA and formaldehyde were obtained from BDH (Poole, UK). Sephadex G-25 was obtained from Pharmacia-LKB (Uppsala, Sweden). The remaining chemicals used in these experiments were all obtained from Sigma (UK). Monoclonal antibody (RI7208) against mouse transferrin receptor was a gift from I. Trowbridge (The Salk Institute, San Diego, California).
Dialysis Dialysis tubing was prepared according to Maniatis et ai (1982). Initially a 10-20 cm length of dialysis tubing was boiled in a large volume of 2% sodium bicarbonate and 1 mmol EDTA 1 ', rinsed thoroughly in distilled water, boiled again in 1 mmol EDTA I-1, allowed to cool and thoroughly washed and boiled several times in a larger volume of distilled water, to remove any trace of EDTA. FITC
conjugation of transferrin
Bovine apotransferrin, 50 mg was dissolved in 4-5 ml of 015 mol NaCl 1 ', to which 0-5 ml of 0-5 mol bicarbonate buffer 1 ' (pH 90) was added, followed by 1 mg fluorescein isothiocyanate (FITC) dissolved in 2 ml acetone. The mixture was incubated for 90 min at room temperature. Unconjugated fluorochrome was separated from the apotransferrin by gel filtration on Sephadex G-25 (Bridges & Fona, 1982). The amount of apotransferrin recovered (88%) was determined by Bio-Rad protein assay (Bio-Rad, Hertfordshire, UK). In a preliminary series of experiments on blastocysts the optimal protocol for measuring uptake of FITCtransferrin (FITC-TF) was determined; this involved removal of zona by acid Tyrode's solution (to prevent subzonal trapping), followed by 1 h in culture before incubation with FITC-TF conjugate (1-2-9-8 mol 1 ') for up to 550 min. Embryos were then rinsed in phosphate-buffered saline (PBS) and fixed in 4% formaldehyde in PBS for 20 min at room temperature, and rinsed through two washes of PBS before permeabilization with 0-2% Triton X-100 in PBS for 5 min to reduce background fluorescence, followed by a short rinse in PBS and neutralization using 0-26% ammonium chloride in PBS for 5 min, followed by a further rinse in PBS. Embryos were mounted in Citifluor (City
University, London).
In a second set of competition experiments, freshly recovered blastocysts were freed from zonae in acid Tyrode's then assigned to six groups. Embryos in groups 1-5 were exposed to 4-9 pmol FITC-TF ' plus 00, 4-9, 9-8, 19-7 and 39-5 pmol unconjugated bovine apo-transferrin 1 ', respectively. Embryos in group 6 were exposed to 4-9 pmol FITCTF plus 58 pmol BSA 1 '. After 300 min the blastocysts in each group were washed, fixed, mounted and FITC emissions measured.
Nonspecific and receptor-mediated transferrin uptake The extents of nonspecific and receptor-mediated transferrin uptake were determined, by culturing zona-free blastocysts in FITC-TF (1-2-9-8 pmol 1 ') for 300 min in the presence or absence of a 40-fold (370 pmol 1 ') excess of unconjugated transferrin. The specific uptake of transferrin was determined by subtracting nonspecific uptake from the total uptake (Fig. 4). In a further test for receptor mediated uptake, freshly recovered zona-free blastocysts were divided into three ' groups and exposed for 300 min to 4-9 pmol FITC-TF 1 ', 4-9 pmol FITC-TF 1 + 395 pmol bovine apo¬ transferrin 1 ' or 4-9 pmol FITC-TF 1 + 544 pg ml monoclonal anti-transferrin receptor antibody (MATRA). '
'
"
Effect of removal of iron One-cell MF1 zygotes were recovered from the oviducts and cultured in the following media: (1) T6 + BSA (3 ml) which had been dialysed against 50 ml of T6 + BSA + TF (22-2 pmol 1 '). (2) T6 + BSA (3 ml) that had been dialysed against T6 + BSA (50 ml), (3) 3 ml of T6 + BSA + TF (111 pmoll ') that had been dialysed against T6 + BSA (50 ml), (4) non-dialysed T6 + BSA, and (5) T6 + BSA + TF ( 11 1 pmol 1 '). ·
Fluorescence microscopy The fluorescence emissions of the embryos were measured immediately by photocytometry on a Leitz Ortholux II microscope fitted with the photomultiplier housing of a Leitz MPV-1. Fluorescent emission was deflected to the amplifier/discriminator (Model 1140B, SSR Instruments Co, USA) of a quantum photometer (Model 1140A, SSR Instruments). For each data point in each experiment the fluorescent emissions of 10-20 embryos were measured and their mean value was expressed as the 'mean reading' (McConnell et ai, 1990). In some experiments two neutral density filters were used so that a reading higher than 10 could be measured on scale. The values for the mean reading between experiments for the same treatment may differ due to different settings of the quantum photometer.
Results FITC-transferrin
uptake
Transferrin is taken up by cells by receptor-mediated endocytosis (Trowbridge & Domingo, 1981; Hopkins, 1985; reviewed by May & Cuatrecasas, 1985). The results of studies on early mouse embryos suggest that endocytic activity matures fully by the blastocyst stage (Fleming & Pickering, 1985). There was an increase in FITC-TF uptake by blastocysts followed by a secondary slower phase (Fig. 1). An example of a blastocyt showing accumulation of FITC-TF is shown
(Fig. 2a).
A decrease in FITC-TF emission was observed with increase in concentration of unlabelled bovine apo-transferrin concentration (Fig. 3). The FITC-TF emission was also decreased by BSA, although not to the same extent as by lower unconjugated bovine apo-transferrin concentrations. The specific and non-specific uptake of FITC-TF were calculated (Fig. 4). MATRA had no effect on FITC-TF uptake by blastocysts, although the addition of 39-5 µ bovine apo-transferrin f reduced the FITC-TF emission (Fig. 5). After characterizing transferrin uptake in blastocysts, zona-free two-cell embryos were incubated with 9-8 µ FITC-TF 1 for up to 24 h, fixed, permeabilized, and then mounted. There was little FITC emission from a zona-free two-cell embryo apart from a few surface flecks (Fig. 2b). Quantitative analysis revealed no uptake over background regardless of the duration of culture. Addition of various concentrations of MATRA to fertilized MF1 zygotes recovered from the oviduct and cultured in bovine apo-transferrin (0-11 1 µ 1_1) did not impair development to the blastocyst stage (data not shown).
"
200
300
400
Duration of culture (min)
Fig. 1. Zona-free freshly recovered mouse blastocysts were cultured in fluorescein isothiocyanatetransferrin (FITC-TF) conjugate ( ) limoli"1; (O) 2-4 µ 1; ( ) 4-9 µ 1; ( ) 9-8 mmol 1 for up to 550min, fixed, permeabilized and FITC-TF emission from each blastocyst measured. Each point on the graph presents the mean reading for 10-20 blastocysts (±sd).
Fig. 2. Photomicrographs show fluorescent labelling of a zona-free freshly recovered (a) mouse blastocyst and (b) mouse two-cell embryo that had been cultured in conjugated fluorescein isothiocyanate-transferrin (FITC-TF) at 9-8 µ 1 ', fixed and permeabilized. Note the somewhat brighter staining in the upper region of the blastocyst associated with the position of the inner cell
mass,
600.
30
fluorescein Fig. 3. Zona-free freshly recovered mice blastocysts were exposed to 4-9 µ 19-7 or (D) ( ) 0-0, (Q) 4-9, (S) 9-8, (0) isothiocyanate-transferrin (FITC-TF) plus "' ' 39-5 µ unconjugated bovine apo-transferrin (BAT) or (0) 58 µ 1 bovine serum albumin (BSA) for 300 min, fixed, permeabilized and the FITC-TF emission from each blasto¬ cyst measured. Each bar represents the mean reading for 10-20 blastocysts (±SD).
2
4
6
FITC-TF concentration
8
(µ
10
1)
Fig. 4. The total uptake (UJ) by zona-free blastocysts from mice of fluorescein isothiocyanatetransferrin (FITC-TF) (1-2-9-8 µ F1) over 300 min was measured and compared with the nonspecific FITC-TF uptake ( ) in the presence of 40-fold (370 µ 1"') excess of unconju¬ gated bovine apo-transferrin. The specific uptake ( ) was calculated by subtraction, hence no error bars. Each point on the graph represents the mean reading for 10-20 blastocysts (±SD).
during the two- to four-cell stage by traces of iron If transferrin and other metal chelators function by chelation of iron from the culture medium and thereby inhibit lipid peroxidation and the formation of hydroxy radicals, dialysis of T6 plus BSA against T6 plus BSA plus TF should remove any trace of iron from the T6 + BSA. The results of this experiment (Table 1) suggest that removal of iron by dialysing T6 + BSA against apo¬ transferrin overcomes the two-cell block and embryos in this medium develop at a rate similar to those in medium containing transferrin. In order to confirm that transferrin had not leaked from the dialysing medium to the dialysate, an SDS one-dimensional gel was run with the dialysate and with calibration levels of transferrin (data not shown). Conversely, when medium containing transferrin was dialysed against an excess of medium, it became less effective in supporting development. Inhibition of development
Fig. 5. Zona-free freshly recovered mouse blastocysts were exposed to 4-9 µ 1_1 fluorescein isothiocyanate-transferrin (FITC-TF) plus (S3) 00 or ( ) 39-5 pmol unconjugated bovine apo-transferrin (BAT) or (D) 544pgmF' monoclonal anti-transferrin receptor antibody (MATRA) for 300 min, fixed, permeabilized and the FITC emission from each blastocyst measured. Each point on the graph represents the mean reading for 10-20 blastocysts ( + sd). Table 1. Effect of dialysing T6 + BSA against T6 + BSA + 2 mg ml ' bovine apo-transferrin (BAT) on development of MF1 one-cell zygotes in vitro Number of Treatment
Dialysate (l)T6(3ml) (2) T6 (3 ml) (3) BAT6 (3 ml) (4) T6 + BSA
(5)BAT6(lmgml')
Dialysed against BAT (50 ml) T6(50ml) T6(50ml)
embryos 48 53 55 43 40
%
embryos at 70 h cells Compact
> Three
10(1 15 55
56 93
00 00 00 00 00
% blastocysts at 100 h 119 h 50 00 00 02 33
85
02 29 14 90
Discussion of FITC-TF at the blastocyst stage is concentration dependent and shows an initial rise followed by a secondary slower phase (Fig. 1). The FITC-TF uptake was decreased when blasto¬ cysts were co-cultured with increasing concentrations of bovine apo-transferrin (Fig. 3). However, part of this inhibition may have been nonspecific since equivalent or excess molar concentrations of BSA also reduced FITC-TF uptake slightly. Two components of transferrin uptake were indicated in the experiments in which blastocysts were co-cultured in various concentrations of FITC-TF plus a 40-fold excess of unlabelled transferrin. The specific uptake, which is inhibited by an excess of unlabelled transferrin, approaches a characteristic saturation plateau, whereas the nonspecific uptake shows a linear increase with increasing concentration of FITC-TF (Fig. 4). Similar results were obtained by Sorokin et ai (1988) using myoblast cells. Attempts to block the specific uptake using the monoclonal antibody to transferrin receptor (MATRA) were not successful. Lesley & Schulte (1985, 1986) used various MATRAs (including that used in this study) to inhibit growth of some but not all lymphoma cells with TF receptors. They concluded that cross-linking of the cell surface receptor is a function of both antibody valency and receptor density and that an antibody with high valency (IgM), like the MATRA which was used here, can be rendered non-inhibitory by a reduction in cell surface TF receptors. It is therefore possible that mouse embryos at the blastocyst stage have a low density of transferrin receptors. Indeed embryonic cells at later stages of development than the blastocyst, such as the giant trophoblast cells which are derived from blastocysts, have a very low transferrin receptor density
Uptake
(Adamson, 1986).
After demonstrating transferrin uptake at the blastocyst stage, we analysed FITC-TF uptake earlier stages but found that it was negligible at the two-cell stage when TF is active in over¬ coming the two-cell block. This result confirms that of Fleming & Pickering (1985) who showed that maturation of the endocytic apparatus begins at about the eight-cell stage. Not surprisingly, addition of MATRA did not prevent the positive effects of transferrin on the development of M F1 embryos in vitro. These results do not support a role for transferrin as a growth factor or an iron-carrier during the transition from the one-cell to the four-cell stage. In contrast, the result from the dialysis experiments suggests that exposure of the medium to chelators relieved the block and allowed blastocyst formation. Most laboratory chemicals and reagents contain traces of iron (1-6-19-4µ 1') that are sufficient to catalyse oxygen radical formation (Gutteridge, 1987). An extracellular role for iron chelators confirms the preliminary studies of Fissore et ai (1989) which showed that injection of EDTA in the perivitelline space of the mouse zygote, unlike intracellular microinjection, resulted in normal development to the blastocyst stage. These authors also suggested that the beneficial effect of EDTA is exerted at or outside the vitelline membrane. Transferrin could act here to prevent autocatalytic peroxidative damage to lipids, to prevent free radical formation by Superoxides and hydrogen peroxide leaking from the cells, and to reduce levels of free iron and other transitional metals available for entry into the cells. Transferrin has been identified as one of the many protein constituents of follicular and ampullary fluids but has not been shown to have any clear functional role during folliculogenesis, ovarian steroidogenesis or oocyte development (Viriji et ai, 1990; Nagy et ai, 1989; Aleshire et ai, 1989), although the transferrin concentration in the follicular fluid does correlate with follicular maturity and steroidogenesis. The concentration of transferrin in follicular fluid ranges from 0-37 l-1 with a mean of 2-5 µ l_l (Entman et ai, 1987), which is within the same effec¬ to 72-5 µ tive range of bovine apo-transferrin used in BAT6 (2-4 to 24 µ 1"'). It has been suggested that such be involved as in the normal process of ovu¬ oxygen-derived radicals, Superoxide, may since the of matured follicles be reduced can ovulated lation, percentage by Superoxide dismutase (Miyazaki et ai, 1991). Transferrin might sequester free traces of iron (with or without other transitional metals such as copper) to prevent damage to the oocyte by the conversion of the required Superoxide to more damaging radicals. In conclusion, our data suggest that transferrin acts both in vivo and in vitro as a chelator to prevent highly toxic hydroxy radical production and thereby free-radical-mediated damage to oocytes and preimplantation embryos, and that this function may be especially important during the peroxide rise observed in vitro during the two-cell stage. at
We wish to thank M. George and B. Doe for their technical assistance, I. Trowbridge for the MATRA, and J. Aitken for advice and discussion. The work was supported by a programme grant
from the Medical Research Council to M. H. Johnson and P. R. Braude. References Adamson, E.D. (1986) Cell lineage-specific gene expres¬
sion in development. In Experimental Approaches to Mammalian Embryonic Development, pp. 338-339. Eds J. Rossant & R. A. Pedersen. Cambridge University Press, New York. Aleshire, S.I., Osteen, K.G., Maxson, W.S. & Enlaman. S.S. (1989) Localization of transferrin receptor in ovarian follicular development. Fertility and Sterility
51,444-449.
Bridges, C.B. & Fona, S.L. (1982) Use of lectin investigate photoreceptor membrane. Methods Enzymology 81, 65-77.
to
in
Chisholm, J.C., Johnson, M.H., Warren, P.D., Fleming, T.P. & Pickering, S.J. (1985) Developmental vari¬ ability within and between mouse expanding blasto¬
cysts and their ICMs. Journal of Embryology and
Experimental Morphology 86, 311-336. Clover, R.M., Howe, A.M., McDonough, P.G. & Boldt, J. (1991) Influence of growth factors in defined
cultured medium on in vitro development of mouse embryos. Fertility and Sterility 55, 194-199. Entman, S.S., Maxson, W., Bradley, C.A., Osteen, K., Webster, B., Vaughn, W.K. & Wentz, A.C. (1987) Follicular fluid transferrin levels in preovulatory
human follicles. Journal of in vitro Fertilization and Embryo Transfer 4, 98-103. Fissore, R.A., Jackson, K.V. & Kiessling, A.A. (1989) Mouse zygotes development in culture medium without protein in the presence of ethylenediaminetetraacetic acid. Biology ofReproduction 41, 835-841. Fleming, T.P. & Pickering, J.S. (1985) Maturation and polarization of the endocytotic system in outside blastomeres during mouse preimplantation develop¬ ment. Journal of Embryology and Experimental Morphology 89, 175-208. Gutteridge, J.M.C. (1987) A method for removal of trace iron contamination from biological buffers. FEBS Letters 214, 362-364. Halliwell, B. (1987) Superoxide-dependent formation of hydroxy radicals in presence of iron chelates. FEBS Utters 92, 321-326. Halliwell, B. & Gutteridge, J.M.C. (1990) In Free Radicals in Biology and Medicine (3rd Edn), Oxford University Press, Oxford. Hopkins, C.R. (1985) The appearance and internalization of transferrin receptors at the margins of spreading human tumor cells. Cell 40, 199-208. Lesley, J.F. & Schulte, R.J. (1985) Inhibition of cell growth by monoclonal anti-transferrin receptor anti¬ bodies. Molecular and Cellular Biology 5, 1814-1821. Lesley, J.F. & Schulte, R.J. (1986) Selection and characterization of receptor mutants using receptorspecific antibodies. Immunogenetics 24, 163-170. McConnell, J., Pickering, S.J., Johnson, M.H. & Bashford, J. (1990) A technique for quantifying the amount of macromolecule injected into cells of the
early embryo. Journal of Reproduction and Fertility
88,375-381. Maniatis, T., Fritsch, E.F. & Sambrook, J. (1982) Molecular Cloning: a Laboratory Manual, p. 456.
Cold Spring Harbor Laboratory, NY. S.W. & Cuatrecasas, P. (1985) Transferrin recep¬ tor: its biological significance. Journal of Membrane Biology 88, 205-215. Minotti, G. & Aust, S.D. (1989) The role of iron in oxygen radical mediated lipid peroxidation. Chemical and Biological Interactions 71, 1-19.
May,
Miyazaki, T., Sueoka, ., Dharmarajan, A.M., Atlas, S.J., Bulkley, G.B. & Wallach, E.E. (1991) Effect of inhibition of oxygen free radicals on ovulation and progesterone production by the in-vitro perfused rabbit ovary. Journal of Reproduction and Fertility 91, 207-212. Nagy, B., Pulay, T., Szaraka, G. & Csomor, S. (1989) The serum protein content of human follicular fluid
and its correlation with the maturity of oocytes. Ada Physiologica Hungarica 73, 71-75. Nasr-Ésfahani, M.H. & Johnson, M.H. (1991) The origin of reactive oxygen species in mouse embryos cultured in vitro. Development 113, 551-560. Nasr-Esfahani, M.H., Aitken, R. J. & Johnson, M.H. (1990a) Hydrogen peroxide levels in mouse oocytes and early cleavage stage embryos developed in vitro or in vivo. Development 109, 501-507. Nasr-Esfahani, M.H., Aitken, R.J. & Johnson, M.H. (1990b) The effect of iron and iron chelators on the in vitro block to development of the mouse pre¬ implantation embryo: BAT6 a new medium for improved culture of mouse embryos in vitro. Human Reproduction 5, 997-1003. Rudland, P.S., Durbin, II., Clingan, D. & de-Asua, L.J. (1977) Iron salts and transferrin are specifically required for cell division of cultured 3T6 cells. Bio¬ chemical Biophysical Research Communications 75, 556-562. Sanders, E.J. (1986) Changes in the transferrin require¬ ment of cultured chicken embryo mesoderm cells during early differentiation. Journal of Embryology and Experimental Morphology 95, 81-95. Shapiro, L.E. & Wagner, N. (1989) Transferrin as auto¬ crine growth factor secreted by Reuber H-35 cells in serum-free culture. In Vitro Cell and Developmental Biology 25, 205-215. Sorokin, L.M., Morgan, E.H. & Yeoh, G.C.T. (1988) Difference in transferrin receptor function between normal developing and transformed myogenic cells as revealed by differential effect of phorbol ester on receptor distribution and rates of iron uptake. Journal of Biological Chemistry 263, 14128-14133. Trowbridge, I.S. & Domingo, D.L. ( 1981 ) Anti-transferrin receptor monoclonal antibody and toxin-antibody conjugates affect growth of human tumour cell. Nature 194, 171-173. Viriji, N., Phillips, D.M. & Dunbar, B.S. (1990) Identifi¬ cation of extracellular protein in the rat cumulus oophorus. Molecular Reproduction and Development 15, 339-344.
Yabe, N., Kato, M., Matsuya, Y., Yamane, I., Ilzuka, M., Takayoshi, H. & Suzuki, K. (1987) Role of iron
chelators in growth promoting effect on mouse hybri¬ doma cells in a chemically defined medium. In Vitro Cell and Developmental Biology 23, 815-820.
Received 19 June 1991
