Syngenta Central Toxicology Laboratory, Alderley Park, Macclesfield, Cheshire SK10 4TJ, United Kingdom
Received July 17, 2002; accepted October 7, 2002
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key Words: genistein; postnatal development; phytoestrogen.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The intrinsic estrogenic activity of genistein has been assessed in the immature mouse and rat uterotrophic assay following both oral and subcutaneous injection of doses up to 300 mg/kg (Ashby, 2000). It was active at lower doses following subcutaneous injection than following oral gavage. These studies indicated that there were no effects on uterine growth in the rat or the mouse at 4 mg/kg body weight/day, the estimated level of human infant exposure (MAFF, 1998
).
The present studies were designed to evaluate the effects on neonatal rats of oral exposure to genistein during the period from birth to postnatal day (PND) 21. As the overall purpose of this work was to generate data for use in assessing human risk following oral ingestion of genistein, the oral route of administration was preferred. However, direct oral gavage of large numbers of rat pups is not technically practical from day 1 postpartum (the day of birth), accordingly two potential exposure protocols were investigated: administration of the chemical to the pup via the dams milk and direct dosing of the neonatal rat pup (by subcutaneous administration to day 7 of age and by oral gavage after this time). The targeted doses were 4 mg/kg/day genistein (equivalent to the average exposure of infants to total isoflavones in soy milk) and a dose 10 times higher than this. The studies were preceded by experiments to evaluate if genistein is secreted in milk following oral administration to lactating rats. Failure to demonstrate significant exposure of the pups by this route led to adoption of subcutaneous injection of the pups during the early preweaning phase. The doses selected for the injection experiments was based on pharmacokinetic studies that enabled equivalent doses to be administered during both the injection and the gavage phases of the study. The definitive study involved assessment of the sexual development of the neonates, including determination of effects on the development, organization, and function of the reproductive neuroendocrine system.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Secretion of genistein into milk following po administration to lactating rats.
A group of four pregnant female rats was allowed to litter normally, after which the litter size was reduced randomly to six. Once lactation had been established, each dam was given a single po dose of 16 mg genistein/kg body weight and milk and plasma samples were taken at 24 h intervals until 5 days after dosing. Each day a terminal blood sample was taken from one pup from each litter. The concentration of genistein was then determined in plasma and milk. The experiment was repeated using [14C]-genistein (50 mg/kg). In this experiment the concentration of genistein was determined in maternal plasma and milk samples and in pup plasma and a metabolite profile obtained.
In the above genistein assays the acidified plasma was extracted with methyl-tert-butyl-ether and the acidified milk was extracted with ethyl acetate. Genistein in the extracts was quantified by liquid chromatography using an ultraviolet detection system (LC-UV). The analytical column was packed with Alltech Alltima C18 5 µm packing and the mobile phase was acetonitrite/water/acetic acid (50:50; 0.1 v/v/v). The metabolites of genistein were characterized by LC-mass spectroscopy following enzymic digestion by ß-glucuronidase/sulphatase for 60 min at 37°C. Radiolabeled peaks were identified by co-chromatography with an authentic genistein standard.
Comparison of systemic exposure in rat pups following administration by either the sc or po route.
A group of pregnant rats was allowed to litter, and on PND 7 two male and two female pups from each of 64 litters were dosed po with 0, 0.4, 4, or 40 mg genistein/kg and a further two male and two female rat pups from each litter were dosed sc with 0, 0.4, 4, or 40 mg genistein/kg. Two litters from each dose group (dosed as described above) were terminated 0.5, 1, 2, 4, 6, 8, 12, and 24 h after dosing and samples of blood collected. In the case of animals dosed with 4 and 40 mg/kg genistein, plasma samples were also taken at termination and analyzed for genistein. In addition to these experiments using unlabeled genistein, 12 litters each comprised of four male and four female rat pups were dosed with 0, 0.4, 4, or 40 mg [14C]-genistein/kg (two male and two female pups dosed sc and two male and two female pups dosed po). One litter from each dosage group was terminated 2, 4, and 8 h after dosing and terminal blood samples collected and the concentration of radioactivity determined for each sample in order to assess both parent compound and metabolites. These terminal blood samples were pooled by sex, time point, and dose route and analyzed for metabolites according to the LC-UV method described above.
Assessment of the effects of genistein on the general postnatal development in the rat.
Time mated rats were allocated to three groups of 20 and allowed to litter. Pups were dosed sc each day from PND 1 (the day of birth) to PND 6, and po each day from PND 7 to 21. From the results of the previous study on comparative exposure by different routes, sc doses were scaled down to provide equivalent exposure to the po route and were set at targets doses of 0.2 or 4 mg/kg (equivalent to 4 or 40 mg/kg by po route) to reflect the increased exposure to genistein by this route. Due to technical error the actual sc doses received on days 16 was 0.2 mg/kg and 2.0 mg/kg. The oral doses given from days 7 to 21 were 0 (control), 4 (low dose), or 40 (high dose) mg/kg body weight/day.
On PND 5 litters were standardized to a maximum of three to four males and three to five females. When standardization was complete individual pup weights and sex were recorded for each litter.
At PND 22 one male and one female pup from each litter were killed and serum samples taken for the analysis of testosterone, LH, and FSH in males, and estradiol, LH, FSH, and progesterone in females. The uterus of each female pup killer was weighed at this time.
All males were examined each day from PND 22 for the time of testes descent. At the time that the developmental landmark was reached the animal was weighed.
Retained F1 animals.
On PND 29, pups were weaned and 30 males and 40 females per group were retained (by selecting up to two males and two females per litter). These were designated as the F1 generation and were weighed at selection and once each week until termination. All animals were examined for vaginal opening or preputial separation from the time of weaning (day 29); at the time that the developmental landmark was reached the animal was weighed. Daily vaginal smears were taken and examined from 20 retained F1 females per group from the time of vaginal opening until the second pro-estrus occurred. At this time (approximately 13 weeks of age), they were killed and serum samples taken for the analysis of estradiol, LH, FSH, and progesterone. All remaining retained males were killed at approximately 13 weeks of age and blood samples taken for the analysis of testosterone, LH, and FSH. Epididymides, prostate and seminal vesicles and testes were weighed.
Assessment of the effects on the development, organization, and function of the reproductive neuroendocrine system.
Groups of five pregnant rats were allocated to four experimental groups and allowed to litter. One group of five litters was given DES (10 µg/kg/day) sc on PND 16 and by po gavage on days 721. A further five litters per group were treated with genistein sc (0.2 and 2.0 mg/kg/day) on PND 16 and by oral gavage (4.0 and 40 mg/kg/day) on days 721. A further group of 5 litters were given carboxy methyl cellulose vehicle alone. At PND 22, between 10 and 12 males and females from each group were retained for surgical procedures. The remaining pups were killed and the uterus and testes weights were recorded. Ovariectomy and castration were performed on PND 2224 and iv catheters were inserted between PND 42 and 54. A two to three day recovery period was then allowed before pups received an iv gonadotropin releasing hormone (GnRH) challenge test, followed immediately by perfusion fixation of the upper body and removal of the brain, for subsequent tissue processing. As the target higher dose of 40.0 mg/kg/day was not achieved in the initial study (an sc dose equivalent to a po dose of 20 mg/kg/day was achieved for days 16, with the target of 40 mg/kg/day achieved for days 721), a second investigation was carried out. In this study only the SDN-POA at the high dose was evaluated where the achieved dose was equivalent to 40.0 mg/kg/day for the whole of the dosing period.
Surgical procedures and blood collection.
Rats were ovariectomized and castrated using standard surgical techniques. Between PND 42 and 54 an intra-atrial cannula was inserted via the external jugular vein using standard surgical procedures and 200 µl aliquots of blood were withdrawn at 15 and 0 min before and 5, 10, 15, and 30 min after iv administration of GnRH (50 ng/kg in sterile saline, 1 ml/kg). Serum was stored at 20°C prior to radioimmunoassay for LH.
Following withdrawal of the final blood sample, rats were deeply anesthetized with Halothane and perfused via the aorta with 4% paraformaldehyde fixative. Approximately 250 ml of fixative was required to achieve good fixation of all brain structures. The brain was then carefully removed and stored for at least a week in 4% paraformaldehyde before processing and analysis.
Measurement of the SDN-POA.
Brains were transferred to a 20% sucrose solution (dissolved in 0.01M phosphate buffered saline, pH 7.4) and left for two days at 4°C. The brain was then trimmed and mounted on a Peltier freezing stage to allow cryostat sectioning using a base sledge microtome. When frozen, 60 µm sections were cut in the coronal plane throughout the region containing the SDN-POA. Sections were mounted on gelatin coated glass slides (six sections/slide) and left overnight at 37°C. Sections were then stained with cresyl violet, dehydrated in alcohol, and coverslipped. The SDN-POA is identified as a darker stained group of cells on either side of the third cerebral ventricle. Using an image analysis system (Olympus BX-50 microscope, and Image ProPlus software) the area of the SDN-POA was measured in each section. The volume of the nucleus was then calculated by multiplying the area by the number and thickness of the sections. The volume of the SDN-POA referred to in the present study is the sum of the left and right measurements.
Statistical methods.
Mean pup weight, total litter weight, mean day of vaginal opening, mean day of preputial separation, and mean day of testes descent were considered by ANOVA. Subsequent mean pup body weights were considered by analysis of covariance on day 1 mean pup bodyweight. Day 1 body weights for selected F1 animals were considered by ANOVA, separately for males and females. Subsequent body weights were considered by analysis of covariance on day 1 body weights, separately for males and females.
Organ weights were considered by ANOVA and analysis of covariance on final body weight, separately for males and females. Mean anogenital distance and blood hormone concentrations were considered by ANOVA, separately for males and females.
Analyses of day 22 blood hormone concentrations and day 22 uterus weights were carried out using both the pup and the litter as the unit of analysis.
LH levels were considered by a repeated measures ANOVA, separately for males and females. SDN-POA volume was considered by ANOVA.
All analyses were carried out in SAS (1996). For Fishers exact tests the proportion in each treated group was compared to the pooled control group proportion. ANOVA and analysis of covariance were carried out using the MIXED procedure in SAS (1996). Least-squares means for each group were calculated using the LSMEAN option in SAS PROC MIXED. Unbiased estimates of differences from control were provided by the difference between each treatment group least-squares mean and the control group least-squares mean. Differences from control were tested statistically by comparing each treatment group least-squares mean with the control group least-squares mean using a two-sided Students t-test, based on the error mean square in the analysis. All statistical tests were two-sided.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
The maximum concentration of genistein in both enzyme treated and untreated milk was reached between 13 h from dosing. Maximum concentrations were 0.04 µg/ml and 0.17 µg/ml in untreated and treated milk respectively.
Radioactivity Content of Milk and Plasma
The mean concentration of radioactivity in plasma, red blood cells, and milk from rats administered a single po dose of 50 mg [14C]-genistein/kg, expressed as µg equivalents/g, are presented in Figure 2. The peak concentration of radioactivity in plasma, red blood cells, and milk was reached at 8 h after dosing with concentration values of 7.1 µg equivalents/g, 0.8 µg equivalents/g, and 3.7 µg equivalents/g in plasma, red blood cells, and milk respectively. At 24 h after dosing, concentrations in plasma, red blood cells, and milk were 2.0 µg equivalents/g, 0.3 µg equivalents/g, and 3.2 µg equivalents/g respectively.
|
|
|
|
The results from this study show that secretion of genistein into milk is low (approximately 0.04% of a dose of 50 mg genistein/kg in milk at 8 h), thus making the required top dose for the assessment of developmental effects in pups (target of 40 mg/kg/day) difficult to achieve. Furthermore because unidentified metabolites of genistein are also secreted into milk this confounds the accurate calculation of dose and hence value in risk assessment were genistein to be administered to the pups through maternal milk.
Comparison of Systemic Exposure in Rat Pups following Administration by Either the Subcutaneous or Oral Route
Plasma concentrations of genistein administered by po and sc routes at 4.0 mg/kg and 40 mg/kg are shown in Figure 6. Mean plasma concentrations of total radioactivity obtained following administration of po and sc doses of 0.4 mg, 4.0 and 40.0 genistein/kg, are presented in Tables 1, 2, and 3
respectively. The analysis of plasma samples obtained from pups dosed at 4 and 40 mg/kg either po or sc showed that plasma concentrations of genistein increased with increasing dose and concentrations were higher in samples taken from pups dosed sc than in pups dosed po. Plasma samples from control pups or from pups dosed 0.4 mg/kg either po or sc, were not analyzed for genistein because analysis of 4 mg/kg samples showed that levels of genistein were below the limit of quantitation within 1 h following po dosing.
|
|
|
|
General Postnatal Development Study
In this study target dose levels of 4.0 and 40.0 mg genistein/kg/day were selected. The 0.4 mg/kg/day dose evaluated in the bioequivalence study was omitted as plasma levels following this dose were below the limit of detection.
There were no effects on clinical condition, body weight gain, or on pup survival in the early postnatal period (up to day 22 postpartum). Anogenital distance was no different in immature pups when assessed at day 2 postpartum and showed no biologically significant difference from control values at day 22. There were no consistent treatment related effects on hormone levels in males or female offspring at day 22. In female pups in the high dose group the uterus weight was two-fold higher than that of the control pups at day 22 (Fig. 7). There was no effect on uterus weight in pups in the low dose group. When this endpoint was assessed in mature females at 12 weeks of age there was no difference between the experimental groups (Fig. 7
). The time of vaginal opening was four days earlier in females in the high dose group. There was no effect on the time of vaginal opening in females in the low dose group (Fig. 8
). The time of preputial separation in males was not affected by administration of genistein.
|
|
|
|
Assessment of the Effects on the Development, Organization, and Function of the Reproductive Neuroendocrine System
Uterine weights.
Uterine weights (with and without adjustment for body weight) were significantly increased by the top dose of genistein and by DES (p < 0.01; Fig. 11). The low dose of genistein had no effect on uterine weight. The effect of DES on uterine weight was significantly greater than that seen for the top dose of genistein. Testis weights (analyzed by litter, with adjustment for body weight) were significantly (p < 0.01) reduced by DES, but were unaffected by either dose of genistein.
|
Sexually dimorphic nucleus of the preoptic area.
The effects of DES and genistein on the volume of the SDN-POA in two studies are shown in Figure 12. Study 1: In control rats the volume of the SDN-POA was significantly (p < 0.05) larger in males compared with females. The SDN-POA in male rats was not affected by any of the treatments. In females, DES treatment caused a significant increase in SDN-POA volume compared with female controls. The SDN-POA volume in DES treated female rats was not significantly different to the volume in DES treated males. The low dose of genistein had no influence on the SDN-POA volume in female (or male) rats. The SDN-POA volume in females treated with the top dose of genistein was higher than in controls but lower than in DES treated females, although these differences were not statistically significant. The SDN-POA volume in females treated with the top dose of genistein remained significantly smaller than corresponding males. Study 2: For DES, a statistically significant increase in volume was seen for females (p < 0.01), although they remained statistically significantly lower than the treated or control males (p < 0.01). No effect was seen in males. For genistein, a small but statistically significant increase was seen in females (p < 0.05). Volumes remained statistically significantly lower than in males (p < 0.01). No effect was seen in males. The volume for females was less than that for the females dosed with DES. A summary of the outcome of the assessments is given in Tables 4 and 5
.
|
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Marked effects on postnatal development were observed for the high dose of genistein. These were characterized by doubling of uterus weight at day 22 and a four-day advance in vaginal opening for the treated animals compared to the control group. These effects are consistent with an estrogenic effect of genistein.
Abnormal vaginal smear patterns of persistent cornification were seen for the high dose animals at the time of vaginal opening, and these persisted for at least 10 weeks after the end of the dosing period, with no sign of a normal pattern becoming established. This finding is the same as the permanent estrus observed by Whitten et al.(1995) for animals exposed to coumestrol for 21 days from birth. It is well established that exposure of fetal and/or neonatal female rats to estrogens is capable of abolishing the capacity to induce an LH surge. This is thought to result from permanent functional changes in the sexually dimorphic regions of the hypothalamus, which are responsible for integrating the positive feedback effects of estradiol to elicit a preovulatory GnRH surge. It is likely therefore, that the persistent estrus induced in rats exposed to the high dose of genistein is the result of an inability to produce an LH surge. LH concentrations were assessed in the females in the present study but the data do not show any decrease in values in the high dose group. However as samples were only taken at a single time point from each animal it is probable that sampling did not coincide with the LH surge. In immature males there were no effects on developmental landmarks or on FSH or LH concentrations.
Body weights of the high dose female pups began to diverge from control values by eight weeks of age. It is interesting that this difference was not seen until four weeks after genistein administration was complete. Fluctuations in body weight are known to be related to the stage of the estrus cycle (Simpson and May, 1973 and from our own unpublished observations), with a decrease in body weight during estrus. This decrease in body weight has been related to increases in activity and decreases in food and water consumption (Brobeck et al., 1947
). Therefore it seems likely that the weight difference is related to the permanent estrus state (persistent cornification) of the females. This conclusion is further supported by Simpson and May (1973)
who noted that the body weight loss during estrus is first evident in their Hooded Lister rats when they reach 170 g at about 5470 days of age. This matches well with the weight difference seen in our study when the rats were approximately 180190 g (78 weeks of age).
Animals exposed to DES, but not those exposed to genistein, had reduced testis weights on PND 22. This effect may be related to a reduction in Sertoli cell numbers, as a consequence of suppressed concentrations of FSH. A marked sex difference was observed in basal LH secretion in animals that received DES neonatally, with LH being significantly lower in females compared with males. This difference was not observed for either dose of genistein. In contrast, the LH response to GnRH was similar in DES treated females compared with controls, providing evidence for an effect on tonic rather than stimulated secretion. It is possible that this reflects an organizational effect occurring within the pituitary gland itself, although a hypothalamic site of action cannot be ruled out. In male and female rats, GnRH elicited a small but significant rise in LH secretion, the characteristics of which were unaffected by either DES or either dose of genistein. Therefore neonatal administration of genistein has no effect on the ability of the pituitary to respond to a GnRH challenge despite the fact that DES caused a permanent reduction in basal LH. Collectively, these data demonstrate that at the doses administered genistein is without effect on basal or GnRH stimulated LH secretion whereas the more potent estrogen, DES, has sexually distinct effects on pituitary physiology (Faber and Hughes, 1991; Register et al., 1995
).
The observation of a sex difference in the volume of the SDN-POA in control rats is in agreement with previous findings, and results from exposure of the male brain to elevated concentrations of testosterone (from the neonatal testis) that is aromatized to produce high local concentrations of estrogen within the SDN-POA. It is thought that these high concentrations of estrogen are neuroprotective, acting either directly, or via the production of neurotrophic factors that leads to the prevention of apoptosis and cell survival (Torran-Allerand, 1996). It is well established that neonatal administration of DES to female rats will result in "masculinization" of the brain and this is reflected by the fact that the SDN-POA is similar in size compared with males (Dohler et al., 1984
; MacLusky and Naftolin, 1981
). A similar masculinizing effect of DES on the female SDN-POA was observed in the present study. In contrast the low dose of genistein had no effect on the SDN-POA volume, which is perhaps not surprising given its inability to increase uterine weights at PND 22. However, the top dose of genistein caused a small but permanent elevation in SDN-POA volume in females only. It is possible that this lack of a significant effect on the SDN-POA volume reflects a difference in the sensitivities of these two endpoints (SDN-POA and uterine weight), with complete masculinization of the SDN-POA requiring a higher concentration of estrogen than that required to elicit a uterotrophic response. Whether this is a true reflection of estrogenic sensitivities or differences in the technical complexities of the two endpoints is not established. Alternatively, the lack of a complete masculinization of the SDN-POA by genistein may be due to the timing of administration. Previous studies have shown that for complete masculinization to take place, as measured by a change in SDN-POA volume, exposure must span the entire critical window in development, which for a rat is between gestational day 18 and PND 27 (Dohler et al., 1984
; MacLusky and Naftolin, 1981
). In the present study exposure was limited to the neonatal period, and this may have resulted in a reduced ability to cause complete masculinization.
In the developmental toxicity study it was noted that in animals that received the top dose of genistein, cyclicity was disturbed and the animals remained in persistent estrus. The most plausible explanation for this is that neonatal administration of genistein caused masculinization of hypothalamic function and this prevented the animals from exhibiting normal cyclical patterns of gonadotrophin secretion. It is therefore likely that the top dose of genistein does have a biological effect on neuroendocrine function, but this is not consistently reflected by measuring changes in the SDN-POA volume. This therefore calls into question the usefulness of this endpoint for evaluating the organizational effects of weak estrogens on neuroendocrine function.
In summary, administration of 40 mg/kg genistein increased uterus weights at day 22, advanced the mean day of vaginal opening and induced permanent estrus in the developing female pups. Progesterone concentrations were also decreased in the mature females. There were no effects in females dosed with 4 mg/kg genistein, the predicted exposure level for infants drinking soy-based infant formulas. There were no consistent effects on male offspring at either dose level of genistein. Although genistein is estrogenic at 40 mg/kg/day, as illustrated by the effects described above, this dose does not have the same repercussions as DES in terms of the organizational effects on the SDN-POA.
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
NOTES |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Ashby, J., Tinwell, H., Odum, J., Kimber, I., Brooks, A. N., Pate, I., and Boyle, C. C. (2000). Diet and the aetiology of temporal advances in human and rodent sexual development. J. Appl. Tox. 20, 343347.[ISI]
Bluck, L. J. C., and Bingham, S. A. (1997). Isoflavone content of breast milk and soy formulas. Benefits and risks; Letter to the Editor. Clin. Chem. 43, 851852.[ISI][Medline]
Brobeck, J. R., Wheatland, M., and Strominger, J. L. (1947). Variations in regulation of energy exchange associated with estrus, diestrus and pseudopregnancy in rats. Endocrinology 40, 6572.[ISI]
Dohler, K. D., Hancke, J. L., Srivastava, S. S., Hoffman, C., Shryne, J. E., and Gorski, R. A. (1984). Participation of estrogens in female sexual differentiation of the brain: Neuroanatomical, neuroendocrine and behavioural evidence. Prog. Brain Res. 61, 99117.[ISI][Medline]
Essex, C. (1996). Phytoestrogens and soy based infant formula. Risks remain theoretical. Br. Med. J. 313, 507508.
Faber, K. A., and Hughes, C. L. (1991). The effect of neonatal exposure to diethylstilbestrol, genistein, and zearalonone on pituitary responsiveness and sexually dimorphic nucleus volume in the castrated adult rat. Biol. Reprod. 45, 649653.[Abstract]
Franke, A. (1997). Isoflavone content of breast milk and soy formulas: Benefits and risks; Letter to the Editor. Clin. Chem. 43, 850851.
Irvine, C. H. G., Fitzpatrick, M. G., and Alexander, S. L. (1998). Phytoestrogens in soy-based infant food, daily intake and possible biological effects. Proc. Soc. Exp. Biol. Ed. 217, 247253.
MacLusky, N. J., and Naftolin, F. (1981). Sexual differentiation of the central nervous system. Science 211, 12941303.[ISI][Medline]
MAFF UK (1998). Plant oestrogens in soy-based infant formulae. Ministry of Agriculture, Fisheries, and Food. MAFF Food Surveillance Information Sheet 167, 18.
MAFF UK (1996). Statement by the COT on phytoestrogens. Ministry of Agriculture, Fisheries, and Food. MAFF Food Surveillance Paper 57, 5981.
Register, B., Bethel, M. A., Thompson, N., Walmer, D., Blohm, P., Ayyash, L., and Hughes, C. (1995). The effect of neonatal exposure to diethylstilbestrol, coumestrol, and ß-sitosterol on pituitary responsiveness and sexually dimorphic nucleus volume in the castrated adult rat. Proc. Soc. Exp. Biol. Med. 208, 7277.[Abstract]
SAS Institute. (1996). SAS/STAT Software: Changes and Enhancements through Release 6.11. SAS Institute, Cary, NC.
Setchell, K. D. R., Zimmer-Nechemias, L., Cai, J., and Heubi, J. E. (1997). Exposure of infants to phyto-estrogens from soy-based infant formula. Lancet 350, 2327.[ISI][Medline]
Sheehan, D. M. (1997). Isoflavone content of breast milk and soy formulas: Benefits and risks; Letter to the Editor. Clin. Chem. 43, 850.
Simpson, K. B., and May, D. (1973). Some effects of the oestrus cycle in the female rat. J. Inst. Anim. Technicians 24, 2529.
Torran-Allerand, C. D. (1996). Mechanisms of estrogen action during neural development: Mediation by interactions with the neurotrophins and their receptors? J. Steroid Biochem. 56, 169178.[ISI]
Whitten, P. L., Lewis, C., Russell, E., and Naftolin, F. (1995). Phytoestrogen influences on the development of behaviour and gonadotropin function. Proc. Soc. Exp. Biol. Med. 208, 8286.[Abstract]