Acting Director: B.A. Schwetz, D.V.M., Ph.D.

Introduction

Increasing recognition of the importance of women's health issues reemphasizes the need for better identification of developmental toxicants and improved assessment of their risk. Congenital malformations recognized at birth affect one in 14 infants (7%); this doubles when later-recognized deficits are included. Some experts estimate that at least one child in three has a birth defect. Additionally, another 7% of infants have low birth weights and at least 25% of recognized pregnancies end in spontaneous abortion.

Birth defects cause over 20% of all infant deaths and are the fifth leading cause of potential years of life lost. More money is spent by states on developmental disabilities (including mental retardation) than on any other category of chronic disease. Over one dozen chemicals, the majority of which are FDA-regulated, are recognized as human teratogens; many more agents are suspected human teratogens. However, no chemical regulated by FDA has been tested for developmental toxicity in pregnant women; only recently have non-pregnant women been included in clinical trials, and some consideration is now being given to also including pregnant women. This puts a heavy burden on laboratory animal research.

FY 97 Goals

1. Develop improved methods and new strategies for detection and prediction of developmental toxicity in laboratory animals and the human population, focusing on reproductive tract development, central nervous system development, whole embryo development, pharmacokinetics during development, and the molecular biology aspects of development.
2. Develop new concepts on how xenobiotics produce developmental toxicant effects.
3. Develop a knowledge base for the estrogenic action of xenobotics during development.

The availability of natural and synthetic estrogens, as well as antiestrogens (each with different pharmacological and toxicological properties), provides opportunities for development of methods and mechanistic approaches to predict risk. Estrogens are etiological agents in female reproductive tract toxicity, a major human health problem. Exposure to FDA-regulated estrogens and antiestrogens occurs in tens of millions of women. There is oral contraceptive exposure in over 100,000 pregnancies each year. In the U.S., about 5% of women will receive tamoxifen sometime during their lifetime. The fertility drug, clomiphene, is responsible for 1% of the live births. Phytoestrogen exposure of the human population via food is virtually universal; infants consuming soy formula are exposed to the highest doses. Estrogens are studied both with respect to their varying pharmacological and toxicological properties and their common mechanism of action. The division is constructing an estrogen knowledge base to predict hormonal activity of untested xenobiotics and to help generate hypotheses identifying gaps in regulatory data. These strategies are important in providing FDA with human and computational expertise and experimental flexibility in dealing with regulatory issues in these areas.

Hyperactivity, mental retardation and other neurological birth defects are another major public health concern. Since pregnant women are exposed to neuroactive xenobiotics, the FDA is responsible for assuring safety. Because functional brain damage can be difficult to detect anatomically (e.g., mental retardation, schizophrenia, depression), functional testing is a necessity. Research is accordingly focused on the development of new strategies, concepts, and more sensitive and interpretable functional, anatomical, and neurochemical methods for detecting developmental neurological insult. This activity provides FDA with expertise regarding the use and interpretation of the newest techniques for assessing developmental neurotoxicity.

Women and their embryos/fetuses are exposed to a number of xenobiotics during pregnancy; most drugs are necessary to maintain maternal health and well-being. Mechanistic studies provide strategies and new concepts to help identify at­risk pregnancies as well as suggest possible intervention therapies (e.g., the FDA issue of folate supplementation) that could circumvent developmental toxicity. Species and strain differences can be investigated in vivo (e.g., Segment II developmental toxicity studies), in which maternal physiological factors can be monitored to determine any maternal effects of a chemical. Maternal plasma and embryonic/fetal drug levels can be measured to estimate embryonic exposure. Toxicity assessments in an in vitro whole embryo culture system allow for the evaluation of the effects of a chemical (or metabolite) in the absence of possible confounding maternal effects.

As part of NCTR's strategic move into the molecular biology of development, efforts toward identifying potential gene biomarkers critical to development are underway. One such effort involves insulin-like growth factors (IGFs), their binding proteins, and receptors. Since diabetes increases the risk of birth defects even in women on insulin therapy, these molecular probes may be important in the etiology of such birth defects.

FY 96 Accomplishments and FY 97 Plans

Over the past 15 to 20 years, NCTR has been a leader in defining the normal and estrogen-altered reproductive tract developmental profile in the rat. This expertise provided the foundation for the reproductive and developmental toxicology involvement with the FDA Women's Health Issues initiatives. This same expertise and the well-defined estrogenic database created over the past 20 years has led to the initiation of a project to create and validate a computerized knowledge base utilizing experimental data to aid in the regulatory decision process, funded by a series of grants from FDA's Office of Women's Health. Additionally, scientists within the division are collaborators with Dr. Fred vom Saal (University of Missouri - Columbia) in a research project on endocrine disruptors funded by the NIEHS for four years.

Center studies have characterized the effects of two newer antiestrogens, droloxifene and toremifene, on developmental endpoints in the rat uterus. The results for toremifene are compared to those we previously described for the related drug tamoxifen in a submitted manuscript.

Work continues on the developmental effects of phytoestrogens. Alterations in reproductive tract morphology and biochemistry in rats treated neonatally with phytoestrogens at times soon after treatment were previously published. Results of follow-up experiments with sacrifices at six and ten months are being written for publication.

The Third International Phytoestrogen Conference, organized and sponsored by the DRDT and NCTR, was held in Little Rock, Arkansas, December 3-5, 1995. The papers from the symposium will be published in the Proceedings of the Society for Experimental Biology and Medicine.

Major studies, involving several NCTR research divisions, on several endocrine disruptors are in the planning stage. Estrogenic chemicals in foods, devices, drugs, veterinary medicines, and other FDA- regulated products are a developing concern. NCTR has taken a leadership role in the area, both within and outside FDA.

Assessment of the sensitivity and validity of behavioral and neuroanatomical measures of developmental brain damage continues. Scientists in this division have improved our ability to detect and assess minor functional abnormalities in rodents. Following a FY96 move into a larger, consolidated laboratory, we now have one of the best equipped rodent behavioral laboratories in the U.S. This laboratory permits the in-depth assessment of as many as ten different behaviors in large numbers of rodents. We are especially well-equipped for detection of mild developmentally-induced hyperactivity and/or mental retardation in rodents. We routinely monitor 24-hour ambulatory and running wheel activity in as many as 16 pairs of animals simultaneously. We can assess short-term activity and response to stimulants in 16 to 24 rodents daily, providing the ability to sensitively detect even modest hyperactivity in large-scale rodent screening studies. Addition of the Morris water maze and expansion of the NCTR complex maze has substantially improved our ability to detect the equivalent of mild mental retardation in rodents.

FY96 saw an expansion of division research on the effects of prenatal retinoid exposure. It has shown that exposure in rodents at a period equivalent to the 4th to 6th week of human pregnancy causes abnormalities in brain development and function at doses far below those which cause morphological abnormalities. In FY96, the division utilized computerized image analysis to detect very specific retinoid effects on the pontine and olivary cerebellar relay nuclei during this sensitive period. It also initiated a major study (still on-going) of the neuroanatomical and developmental effects of retinoid exposure at three separate developmental stages. While still preliminary, this study suggests that major behavioral abnormalities occur with retinoid exposure well into the rodent equivalent of the 6th to 8th week of human pregnancy, again in the absence of major malformations. These findings suggest that the vitamin supplementation often begun at the time of pregnancy detection in humans may have the potential to create "silent" neurological effects, perhaps manifesting as mild hyperactivity. This disturbing possibility clearly merits further experimental scrutiny, and in FY97 the division will continue the above project while initiating a number of collaborative projects to more closely examine these effects. These collaborative efforts will include: 1) work with Dr. Jane Adams (University of Massachusetts) to attempt to determine whether cerebellar functional abnormalities seen in rodents might also occur in isotretinoin-exposed children; 2) work with Dr. Mark Stanton (EPA) to better characterize the functional cerebellar deficits in retinoid-exposed rodents; and 3) work with Dr. Frank Scalzo (UAMS/Arkansas Children's Hospital) to better characterize and understand feeding problems caused by retinoid exposure. This research will help to better clarify the neuroanatomical and functional effects of retinoid exposure in rodents and in humans.

Attention deficit disorder/hyperactivity still afflicts some 2% of children in the U.S. Over the past six years, division research has suggested that this syndrome is seen in rats after developmental exposure to a variety of compounds. Common to these syndromes is an unexpected 10% reduction in cerebellar weight. Neonatal exposure to antimitotics or dexamethasone, or prenatal exposure to low doses of retinoids, reduces adult cerebellar weight and is accompanied by behavioral hyperactivity. Consequently, they now believe that the cerebellum may be an especially sensitive target for developmental neurotoxicants. During FY97, the division with our collaborators at the University of Tennessee, are examining such effects with state-of-the-art neuroanatomical and molecular biology techniques. To date, we have seen dexamethasone effects in all major regions of the cerebellum, as late as six months after a single dexamethasone injection on postnatal day seven have been seen. The division is currently assessing retinoid effects on survivors at this and earlier ages.

Microdialysis in various brain regions has become an established neurochemical assessment technique in our laboratory. In FY96, the Division of Reproductive and Developmental Toxicology (DRDT) initiated a project to assess the neurochemical response to FDA-regulated compounds such as fenfluramine in rodent neonates. The division staff also published a report on changes in the brain over the first 24 hours following microdialysis probe implantation. These findings have important implications for the interpretation of all results collected by this procedure. That work will continue in FY97 while also expanding the ability to use this technique to track cerebral metabolism of pharmaceuticals. This will be done in part by applying the technique of mass spectrometry to analyze methylphenidate (ritalin) levels in the brain and to trace the cerebral metabolism of this compound. It is also hoped that a better understanding of cerebral nitric oxide (NO) release with mass spectrometry will be achieved. If this application of mass spectrometry to cerebral microdialysis succeeds, the NCTR will have taken a major step forward in development of this important tool.

A classical Segment II teratology study of fumonisin B₁ (FB₁) in rabbits was completed. Decreased fetal body and organ weights were observed but only at the higher FB₁ doses. These doses also produced a significant amount of maternal toxicity, so it is unclear if the effects on fetal weight were secondary responses to maternal toxicity. The ratio of sphingonine to sphingosine, which was used as a biochemical marker of FB₁ exposure, was increased in a variety of maternal tissues but was not altered in fetal tissues. This suggests that FB₁ may not have crossed the placenta and further suggests that, in the absence of maternal toxicity, this compound does not appear to be a significant developmental toxicant.

The anticonvulsant drug, carbamazepine, is believed to cause neural tube defects in humans. The drug does produce developmental toxicity in animal models, and it is also capable of producing neural tube defects in a rodent whole embryo culture system in which rodent embryos are cultured directly in serum containing carbamazepine. Addition of rat or human hepatic metabolizing fractions increased the drug's embryotoxicity, somewhat suggesting that a metabolite might be the developmental toxicant. The tripeptide glutathione is able to detoxify a number of reactive metabolites. Experiments to manipulate embryonic concentrations of glutathione, either by decreasing its concentration by the action of the inhibitor buthionine sulfoximine or by increasing its concentration by addition of exogenous glutathione, have been inconclusive. During the next year, DRDT plans to develop methods to measure embryonic concentrations of glutathione to insure that the treatments are having the desired effect. It also plans to test additional scavenging compounds to determine if they are able to decrease carbamazepine-induced embryotoxicity.

Another anticonvulsant drug, valproic acid (VPA), is known to produce neural tube defects in 1 to 2% of exposed human offspring as well as in animal models. The mechanism whereby VPA produces these defects is unknown. A number of anticonvulsants, including VPA, decrease the concentration of the vitamin folic acid. Evidence has indicated that this vitamin decreases the incidence of neural tube defects in humans. However, using a rodent whole embryo culture system, it has been demonstrated that neither folic acid nor a variety of other compounds involved in folic acid metabolic steps are able to decrease the frequency of neural tube defects produced by valproic acid. Preliminary evidence suggests that VPA might decrease the concentration of the primary donor of methyl groups for various methylation reactions, S-adenosylmethionine (SAM)(a product of one-carbon transfer reactions which utilize folic acid), and may decrease overall methylation of embryonic DNA. During the next year DRDT plans to further examine the effects of VPA on embryonic methylation reactions. And they plan to look at other compounds which produce neural tube defects to determine if alterations in SAM concentrations and methylation reactions might be a common mechanism leading to neural tube defects.

Based on its suspected morphoregulatory role in nervous system development, expression of the cell adhesion molecule (N-CAM) was examined in mouse embryos treated with a dose of VPA known to produce neural tube defects. Using a high sensitivity, high resolution Western blot system, N-CAM was detected in embryonic heads. No difference could be detected between VPA- and control-treated mice; however, localized differences in expression would not be detectable by this technique. Work has begun to use immunohistochemistry and in situ hybridization to determine if more localized alterations in expression could occur following treatment with VPA.

Additional mechanisms of developmental toxicity will be examined over the next year. These include alterations in stress protein synthesis as a mechanism for various developmental toxicants, including retinoic acid, heat and VPA.

A molecular biology capability has now been brought to this research area which enables developmental toxicity to be measured in terms of effects on fetal gene expression. This approach has been applied to determine the effects of maternal diabetes on fetal expression of eight insulin-like growth factors (IGFs) and binding protein mRNAs. The division has been able to identify a single binding protein, previously demonstrated to function as a growth inhibitor in other systems, that is upregulated in the growth retarded fetuses of diabetic dams and downregulated by insulin treatment. The extent of regulation of this binding protein gene is currently being quantitated by extremely precise molecular techniques.

The mechanism of retinoic acid-induced limb malformations will be analyzed at two different levels over the coming year using similar molecular techniques. First, the fetal distribution of nine distinct retinoic acid (RA) receptors and binding proteins will be determined to identify which are located in limb and other target tissues for RA-induced birth defects. This information could potentially allow the selection of retinoid derivatives which are more appropriate for therapeutic application due to a lack of binding to the "teratogenic" RA receptor or binding protein. Second, the effects of RA exposure on production of IGFs and IGF binding proteins in fetal limbs will be determined in order to identify which might be involved in RA-induced limb growth defects.

Investigations utilizing a laser scanning confocal microscope coupled with a Silicone Graphics Workstation have electronically imaged mouse embryos. The 3-D reconstruction of gestational day 9 to 11 embryos has provided insight into logistical problems of gestational age identification markers, processing shrinkage of tissues and organs, and storage, identification, and retrieval of the vast amounts of electronic data being generated. This next year will continue this effort to collect developmental data on normal embryonic and fetal development in the mouse and rat.

Significance to the FDA

Reproductive and developmental toxicology is investigating the effects of drugs and other xenobiotics regulated by FDA which collectively have extensive human exposure during pregnancy. By improving methods to detect and characterize developmental toxicants as well as determining the mechanisms for their effects, the FDA will be in a better position to predict the human developmental toxicity of regulated products and to advise regulated industry of appropriate procedures. This research area utilizes an integrated research approach by emphasizing molecular, endocrine, behavioral, limb and whole embryo in vitro, and pharmacokinetic techniques.

This combination of techniques and expertise is unequalled by any other single group in the FDA for studies in developmental toxicity and positions the individual scientist to be able to best contribute to the FDA regulatory arena.