Organization versus Activation: The Role ofEndocrine-disrupting Contaminants (EDCs)during Embryonic Development in WildlifeLouis J. Guillette, Jr., D. Andrew Crain, Andrew A. Rooney,and Daniel B. PickfordDepartment of Zoology, University of Florida, Gainesville, Florida

Many environmental contaminants disrupt the vertebrate endocrine system. Although they may be no more sensitive to endocrine-disruptingcontaminants (EDCs) than other vertebrates, reptiles are good sentinels of exposure to EDCs due to the lability in their sex determination. This isexemplified by a study of alligators at Lake Apopka, Florida, showing that EDCs have altered the balance of reproductive hormones resulting inreproductive dysfunction. Such alterations may be activationally or organizationally induced. Much research emphasizes the former, but a completeunderstanding of the influence of EDCs in nature can be generated only after consideration of both activational and organizational alterations. Theorganizational model suggests that a small quantity of an EDC, administered during a specific period of embryonic development, can permanentlymodify the organization of the reproductive, immune, and nervous systems. Additionally, this model helps explain evolutionary adaptations tonaturally occurring estrogenic compounds, such as phytoestrogens. - Environ Health Perspect 103(Suppl 7):157-164 (1995)

Key words: endocrine disrupting contaminants, reptiles, steroids, estrogen, activation, organization, embryo

IntroductionEnvironmental contaminants have posed a

major threat to wildlife health since theonset of the industrial age. The focus ofour concern on the health consequences ofenvironmental pollution have, in the lastthree decades, been on lethal, carcinogenic,and extreme teratogenic manifestations.Evidence from a number of sources sug-

gests that another mechanism, endocrine-disruption, must also be examined (1).

There is good evidence that man-madefactors in the environment act as hormonesor antihormones. It was this concern thatled the National Institute of Environmental

This paper was presented at the Symposium onEstrogens in the Environment, III: Global HealthImplications held 9-11 January 1994 in Washington,DC. Manuscript received: March 15, 1995; manu-script accepted: April 4, 1995.

The St. John's Water Management District,*BEECS Program, University of Florida, and U.S.Department of Agriculture grant #92-37203-7919 pro-vided partial funding for the alligator work describedin this manuscript. We thank Howard Bern, Fred vomSaal, Wade Welshons, and Theo Colbom for valuablediscussions concerning the topics presented in thispaper.

Address correspondence to Dr. Louis J. Guillette,Jr., 223 Bartram Hall, University of Florida,Gainesville, FL 32611. Telephone: (904) 392-1098.Fax: (904) 392-3704. E-mail: LJG@ZOO.UFL.EDU

Abbreviations used: DDT, dichlorodiphenyltri-chloroethane; DES, diethylstilbestrol; EDCs,endocrine-disrupting contaminants; PCBs, polychlori-nated biphenyls; DDE, dichlorodipheny1dichloroethyl-ene; E2, 17j3estradiol; IGF-1, insulinlike growth factorI; SBG, sex steroid-binding globulin; T, testosterone.

Health Sciences to organize a sympo-sium-"Estrogens in the Environment"in 1979 to address growing concern thatmany estrogenic compounds, such asdiethylstilbestrol (DES) used in the animalscience industry and the pesticide DDT,were being released into the environment(2). Over a decade later, the World WlldlifeFund (3) organized a conference thataddressed the premise that xenobiotic com-pounds did not act solely as estrogens butwere also agonists and antagonists of otherhormones that disrupted the endocrine sys-tem of developing embryos. Disruption ofembryonic development produced perma-nent modifications in the reproductive,immunological, and neurological capabili-ties of future populations and did so bymechanisms other than gene mutation.The World Wildlife Fund meeting gener-ated a consensus statement by the attend-ing scientists which read, "We are certainof the following: a large number of man-made chemicals that have been releasedinto the environment... have the potentialto disrupt the endocrine systems of ani-mals, including humans" (3). To furtherassess the damage done to wildlife popula-tions, it is essential that we expand thewildlife models currently in use. Onegroup of vertebrates, the reptiles, can pro-vide important new insight due to a num-ber of attributes associated with theirsystem of sex determination.

Reptiles as Models for theStudy of Endocrine-disruptingXenobioticsMany crocodilian (induding alligators) andturtle species exhibit environmentally (tem-perature) determined sex, and this environ-mental influence can be overcome (sexreversal) by treating eggs with estrogen.Genetic factors associated with sex determi-nation in reptiles, and thus gonadal develop-ment, are still under study, but the role oftemperature and sex steroids has been exten-sively examined (4-7). In alligators, thetemperature of incubation at specific criticalperiods of embryonic development triggersthe determination of sex (4,7). Incubationtemperature induces an all-or-none responseso that embryos are either males or females,with few intersexes produced (8).

Studies have shown that alligators andseveral turtle species can exhibit sex reversal(male to female) if developing embryos areexposed to an estrogenic compound duringa specific period of development, usuallythe second third of the embryonic period.This critical period represents a window ofheightened vulnerability to estrogenic com-pounds. Thus, compounds with a shorthalf-life in the environment need only bepresent during this window to have perma-nent developmental effects. For instance, asingle pulse of estradiol treatment given toalligator (9) or turtle (10,11) eggs incu-bated at male-producing temperatures caninduce the production of apparently normal

Environmental Health Perspectives 157

GUILLETTE ETAL.

females. Intersex individuals are producedwhen the concentration of an estrogeniccompound is below a specific threshold.That is, with intermediate treatment levels,an embryo may exhibit Miillerian (female)ducts and a testis or even an ovotestis.Turtle eggs exposed (painted or injected)to estradiol or other estrogenic compounds(polychlorinated biphenyls [PCBs]) showsuch partial reversal. Bergeron et al. (12)observed that 2',4',5'-trichloro-4-bipheny-lol induced 100% male to female sex rever-sal (based on histological examination ofgonads and internal ducts) in the red-earedturtle (Trachemys scripta elegans), whereastreatment with 2',3',4',5'-tetrachloro-4-biphenylol stimulated total sex reversal in50% of the embryos and partial sex reversal(intersex) in 21% of the embryos.Interestingly, red-belly turtle neonates(Pseudemys nelsoni) from Lake Apopka,Florida, contaminated with a number ofendocrine-disrupting contaminants (EDCs),the most common being p,p'-DDE, haveprimarily normal ovaries or abnormalovotestes-very few animals exhibit mor-phologically normal testes (TS Gross andLJ Guillette, unpublished data). Moreover,chorioallantoic fluid concentrations of1713-estradiol (E2) and testosterone (T)indicate that no turtle hatchling from LakeApopka produces a normal androgen syn-thesis pattern (TS Gross and LJ Guillette,unpublished data). Similar observationshave been reported for the alligators livingin Lake Apopka (13,14). Both males andfemale juvenile alligators exhibited abnor-mal plasma sex steroid concentrations, withmales from Lake Apopka having greatlyreduced plasma T concentrations similar tothat of females from either the contami-nated (Lake Apopka) or control (LakeWoodruff) lakes. The juvenile female alli-gators from Lake Apopka had elevatedplasma E2 concentrations compared to thefemales from the control lake. Testes fromLake Apopka juvenile males make elevatedlevels of E2 but normal levels ofT in vitro,suggesting that the aromatase enzyme sys-tem, essential for estrogen synthesis, isupregulated in males (14). Treatment ofalligator eggs with either DDE or estradiolinduces similar results, with complete sexreversal or intersexes and abnormal plasmasteroid concentrations observed (TS Grossand LJ Guillette, unpublished data).

These studies suggest that reptiles rep-resent excellent model species to determinethe extent of contamination of an ecosys-tem due to the lability of their sex determi-nation in response to the presence of

steroids, steroid-mimicking compounds, orsteroid synthesis-disrupting compounds(12,15,16). In addition to their labile sexdetermination, reptiles such as turtles andalligators represent species that feed at vari-ous levels in the food web and live forextensive periods. Many species also showstrong site affinity, allowing the examina-tion of specific wetland regions. Theseattributes, and those of other species, willhelp provide general answers on how EDCsaffect wildlife populations. However, thesestudies will not address the full extent ofthe EDCs threat to wildlife health unlessall life stages are examined, especially theperiod of embryonic development.Organization versusActivationIf embryos are a major life stage affected byEDCs, it is important to address why thisstage is so susceptible to factors characterizedas weak estrogens and antiestrogens whencompared to the native estrogen 17p-estra-diol. In this discussion, we address a num-ber of biological principles we feel shouldprovide a framework for future EDCresearch. Obviously, embryos are small andexhibit high rates of mitosis, which makesthem sensitive to environmental perturba-tions, but additional factors contribute toan embryo's organizational response toEDC exposure: a) critical periods of sensi-tivity during embryonic organization, b)bioaccumulation versus degradation andsecretion, and c) free versus bound hor-mone. These principles are important asthey suggest that all EDCs have the poten-tial to significantly modify the organizationof a developing embryo. We will addressthese issues in relation to the organizationalversus activational framework and provideexamples from wildlife.

It was only recently that studies in thefield of developmental biology began toappreciate the diverse roles of hormonesduring early embryonic development (17).The pervasive view has held that hormonescould not influence embryonic develop-ment until a source of hormones was pre-sent. Early studies, for example, clearlyshowed that androgens and Mullerianinhibiting hormone were essential fordevelopment of mammalian, male repro-ductive duct systems, but this influenceoccurred only after a source of these hor-mones, the testes, formed (18). However,recent studies have begun to show thatmany embryonic cells exhibit receptors forvarious hormones significantly before thesehormones are synthesized by embryonic

sources. For example, developing chick gas-trulas exhibit receptors for insulinlikegrowth factor I (IGF-I) on the day of laying(19), but an embryonic source of this hor-mone is not present until several days later(20). Similar observations have now beenmade in frogs (21). These data suggest thatearly embryos would be capable ofresponding to a hormonal signal if a signalwere present.

Is there a source of hormone that mightinteract with the receptors present beforean embryonic source of hormone? IGF-I ofmaternal origin has been identified in theyolk of chicken eggs (22) and in the yolkand albumen of alligator eggs (23). Like-wise, the yolk of teleost fish is a rich sourceof maternal thyroid (24) and varioussteroid (25) hormones, and alligator yolkhas abundant steroid hormone concentra-tions at the time of ovulation (Guillette etal., unpublished data). Eggs of most verte-brate species have been shown to possesssignificant quantities of various environ-mental contaminants, many ofwhich act ashormone mimics and bind with specifichormone receptors (26,27). This is espe-cially true for turtles and alligators that laylarge, yolky eggs (28-32). The ability tobioaccumulate and mimic hormones makesendocrine-disrupting contaminants potentmodifiers of embryonic development.

The Importance of ContaminantEposure duwing Embryonic Organiation.The concepts of organization versus activa-tion have been useful in explaining the roleof hormones in the differentiation of verte-brate sexual dimorphism, whether mor-phological, physiological, or behavioral(33). As originally defined, organizationaleffects occur early in an individual's life-time and induce permanent effects,whereas activational effects usually aretransitory actions occurring during adult-hood (34). For example, androgens orga-nize mammalian embryos by stimulatingthe development of the male reproductiveduct/glandular system as well as the exter-nal genitalia (Figure 1). Likewise, andro-gens stimulate growth and secretoryactivity of glands associated with the malereproductive tract, an activational effect(Figure 1). In their original presentation ofthe organizational/activational concept,Phoenix et al. (34) examined the role ofprenatal testosterone treatment on subse-quent adult behavior of guinea pigs andsuggested that three main criteria definedthis concept. Organizational effects a) werepermanent, b) occur early in life, usuallyjust before or after birth, and c) occur

Environmental Health Perspectives158

ORGANIZATION VERSUS ACT!VATIONAND EDCS

Organizational

Testosterone $

OrganizationalActivational t.*-,TestosteroneTestosterone +f

Figure 1. A simple representation of the organizationalversus activational model. The presence of testos-terone early in development induces the differentiationof the Wolffian duct so as to form the male internalreproductive duct system ( 18). The source of testos-terone for this transformation is the testis. In mam-

mals, testicular formation is apparently due to agenetically expressed signal that does not involve a

steroid hormone. In contrast, in reptiles, steroids or

enzyme inhibitors (antiaromatase) can modify embry-onic development and gonadal formation, thus havingan organizing effect at this level (16,35). Thus, in mam-mals and reptiles, testosterone-induced reproductiveduct development is said to be an organizing effectwhereas the stimulation of reproductive duct growthand secretion by testosterone after puberty is activa-tional. When initially presented, organizational effectswere believed to occur before birth, but it is clear thatthese effects can continue throughout life, especiallyprior to puberty [e.g., stimulation of Sertoli cell numberduring neonatal development (36)].

during a critical or sensitive period. Thethree original criteria have been extendedby subsequent researchers. The first exten-

sion states that permanent organizationaleffects imply structural changes in tissuesor organs; the second extension concludesthat organizational effects produce sexualdimorphism (33). However, it is simplisticto assume that an individual is completelyorganized at birth. Current research indi-cates that steroid hormones have organiz-ing effects on the neonate [e.g., stimulationof Sertoli cell number (36)] and even intoadulthood [e.g., changes in mammary tis-sue with pregnancy (37)].

Much of the work, both theoretical andexperimental, examining the organizationalversus activational roles of hormones hasbeen focused in the field of neurobiologyand behavior (33). However, these same

concepts are also central to the role of envi-ronmental EDCs. Currently, many studiesexamine the activational effects of hor-mone-mimicking environmental contami-

nants, primarily in adult model systems. Incontrast, we must emphasize the organiza-tional effects of these endocrine mimics ifwe are to assess their complete impact on

environmental health. It is, in fact, the

reorganization of the embryo by exposureto EDCs that is a major concern. There arewell-known examples of this phenomenon.

The best known and most intensivelystudied example is the DES daughter/sonsyndrome. The DES syndrome was par-tially recognized and understood becauseof a preexisting model using prenatally orperinatally estrogen-exposed laboratoryrodents (38). In rodents, exposure to estro-genic compounds during in utero develop-ment or immediately after birth results inpathological changes of the reproductivetract, as well as functional differences atpuberty and throughout adulthood. Similarestrogenic effects have also been demon-strated on the immune (39) and neuroen-docrine (40) systems.

Many of the observed DES-inducedmodifications to the reproductive system aremorphologically subtle but result in majorfunctional changes. A number of studieshave documented that the reproductive tractof both male and female mice exposedneonatally to DES exhibit protein and geneexpression patterns that are unique whencompared to controls (41-43). Exposure ofneonatal male mice to DES produced asignificant organizational effect so that, atsexual maturity, the seminal vesicle stimu-lated secondarily with estrogen could syn-thesize lactoferrin, a protein not normallyproduced by this gland (44). Secretion oflactoferrin was at a level similar to that seenin normal uterine tissue. Perinatal exposureof mice to DES also induces organizationalmodifications of a very subtle type; that is, itmodifies the types and abundance of recep-tors on various tissues of the reproductivesystem (38). Although changes in receptortype and abundance can be considered sub-tle, developmental abnormalities of this typemay be the basis for infertility or reproduc-tive cancers (38).

Similar to the reproductive effectscaused by DES, immune modificationresulting from neonatal DES exposureoccurs during a critical period. Whereasadult exposure to estrogens temporarilyinhibits many aspects of the immune system(45), neonatal exposure to DES causes apersistent impairment of several immuneparameters, including reduced delayed-hypersensitivity response, decreased in vitromitogen response, and depressed graft versushost reaction (46,47). The developmentalstage during which exposure occurs seemsto be critical for determining persistence ofestrogenic effects. For example, DES treat-ment of mice during the first 5 days afterbirth causes immune depression evident at

17 months of age, whereas treatments ondays 6 through 10 had no discernibleeffects later in life (47). The mechanismsof the action of estrogens on the immunesystem appear to involve both lymphoidand nonlymphoid tissues. Estrogen recep-tors are present at low levels in lymphoidcells and near uterine levels in thymicepithelium (39). EDCs may alter estrogenreceptor levels in immune tissues, similar toreceptor changes of the reproductive system.However, to our knowledge, modificationsin steroid receptor number in immune tis-sues following embryonic exposure to EDCshave not been investigated.A complication when examining the

immune or reproductive systems is theinterwoven nature of their function. Thebidirectional communication between theneuroendocrine and immune systems hasbeen the subject of many recent reviews(48). One such manifestation of thisrelationship may be the relationshipbetween immunotoxicity and estrogenicity.Immunotoxicity of estrogens correlates withestrogenicity as measured by uterotrophicactivity (39). The interactions between theneuroendocrine and immune systemsappear to be temporally restricted in a man-ner similar to the increased sensitivity toEDCs during windows in the organiza-tional phase of development. Neonatallythymectomized mice, as well as congeni-tally athymic mice, display a profoundlyaltered endocrine system (49). Endocrineirregularities associated with this T-lym-phocyte deficiency include delays inpuberty in females, alterations in normaladenohypophysis formation, alterations inadrenal gland formation, and abnormal cir-culating levels of gonadal hormones. Theseconditions are often developmentally fixedand fail to normalize despite successfulrestoration of immunocompetence by lym-phoid replacement later in life.

Interestingly, it is also important tonote that many organizational modifi-cations do not become apparent until laterin life. For example, exposure to elevatedconcentrations of estrogens during pubertyand adult life appears to contribute to anincrease in the severity of DES-inducedembryonic abnormalities. For example, amajor symptom of the DES daughter syn-drome is the occurrence of vaginal clear-cell carcinoma in young women (50). Thisphenomenon only occurs after pubertywhen plasma estrogen concentrations areelevated and stimulate reproductive tractgrowth. In mice, postnatal estrogen expo-sure significantly increases the severity of

Volume 103, Supplement 7, October 1995 159

GUILLETTE ETAL.

the reproductive abnormalities induced byprenatal estrogen exposure (42,51,52).

Examples of organizational effects inwildlife species have been noted as well(Table 1). The pioneering studies of Fryand Toone (57) demonstrated that expo-sure of seabird eggs to DDT concentra-tions, comparable to those found incontaminated eggs obtained in wild nests,induced the development of ovarian tissueand oviducts in male embryos (57). SimilarEDC-induced gonadal abnormalities havebeen noted in some fish populations (62).A recent study demonstrated that PCBs willinduce sex reversal-male to female-infreshwater turtles with the development ofapparently normal ovaries and oviducts(12). Organizational modifications havealso been reported for alligators collectedfrom a lake, Lake Apopka, known to havewildlife and eggs with elevated p>p'-DDEconcentrations. Gonads of male and femalejuvenile alligators are permanentlymodified morphologically and physiologi-cally after exposure to this EDC (13,14).Although the specific time of embryonicexposure is unknown, as these are wild eggspossessing elevated contaminant levels atovulation, the nature of the reproductiveabnormalities (polyovular follicles, poorlyorganized testes, small phalli, and alteredsteroid secretion) indicates that exposureprobably occurred throughout the in ovoperiod, inducing multiple organizationaldefects as observed. Other wildlife speciesfeeding in contaminated food webs exhibitdevelopmental abnormalities characteristicof endocrine disruption, and many adultsshow reproductive, immune, and neuro-logical abnormalities that may representorganizational or activational effects (Table1). A specific cause-effect relationship isoften difficult to recognize in wildlifespecies, but laboratory studies confirm thatin ovo or in utero exposure to EDCs cancause irreversible alterations to the repro-ductive systems of wildlife species.

It was not until the medical and sci-entific community began to appreciate theorganizational influences of DES on thedeveloping mammalian embryo that thetrue magnitude of the DES daughter/sonsyndrome became clear (38). If we are toassess the threat of endocrine-disruptingcontaminants to wildlife and human popu-lations, then we must examine both activa-tional and organizational effects of EDCson developing embryos.

Bioaccumulation and EvolutionaryAdaptation to Endoerine-disruptingContaminants. One of the legacies of

Table 1. Organizational and presumed activational effects of environmental endocrine-disrupting contaminants(EDCs) in wildlife species.a

Species Organizational effectb Activational effectb EDCsC References

FishSalmon Premature sexual maturity -I Fertility PCBs (53)

Loss of sexual dimorphism .1I1 7)-Estradiol DDTT Embryo mortality 4- Dihydrooxyprogesterone Dioxins

FuransMetals

Mosquito-fish ? Masculinization of females Kraft mill (54)Anal fin modifications effluentMating behavior

Trout ? T Plasma vitellogenin (yolk Sewage works (55)protein) in males effluent

White croaker I Embryonic mortality 4 Fecundity and fertility DDT (56)1 Ovarian follicular atresia

ReptilesSnapping turtle I Embryonic deformities ? PCBs (30)

DioxinFurans

American alligator 4-Testosterone (male) Poor quality eggs DDE (13)Abnormal testicular cells1 17,B-Estradiol (female)I Polyovular follicles andpolynuclear oocytes

BirdsWestern gulls Retained Mullerian ducts Female-female pair bonds DDT (57)

Abnormal gonadal Abnormal mating behavior DDEmorphology

Bald eagle T Embryonic mortality and 4 Fertility PCBs (58)deformities DDE

DieldrinJapanese quail ? 4 17)-Estradiol before PCBs (59)

sexual maturationDelayed oviposition4 Laying capacity

MammalsDall's porpoises ? 4 Plasma testosterone p,p'-DDE (60)Beluga ? I Follicular activity DDT (61)

Mammary carcinoma MirexPCBs

Partial, representative listing only. bAs the mechanisms underlying many of these effects are unknown or understudy, the phenomenon listed as activational may be due to an organizational effect not apparently obvious atbirth. CThe contaminant(s) listed has been found to compose the greatest body burden in the animals studied.

environmental pollution is the bioaccumu-lation and biomagnification of contami-nants within the animals feeding at variouslevels of the food chain. Lipid-soluble pol-lutants are stored in fat reserves and, uponmobilization during reproductive events,developing embryos are exposed to thebioaccumulated contaminants. Femaleswith large, yolky eggs use the energy storesin fat reserves to synthesize and store vari-ous compounds in the oocyte, and thesecompounds are later used for embryonicdevelopment and growth. Hormones suchas growth factors (e.g., insulinlike growthfactor-I), steroids (17 -estradiol, testos-terone), thyroxin, and vitamins aredeposited in the yolk for use during embry-onic development, but EDCs such as

PCBs, DDT, DDE, and dioxin, to name afew, also are deposited in the eggs becauseall females have bioaccumulated EDCs.Similarly, embryos developing in utero areexposed to hormones and EDCs via uter-ine secretion and placental transfer.

It is clear that developing embryos canbe exposed to EDCs, but it has beenargued that this exposure is innocuousbecause embryos are normally exposed toexogenous (plant) estrogenic sources. Thisargument is supported by the observationthat herbivores grazing on various plantscontaining phytoestrogens can transportthese plant secondary compounds to theirdeveloping young. A number of these phy-toestrogens are known to interact with theendocrine system and disrupt reproduction.

Environmental Health Perspectives160

ORGANIZATION VERSUSACTIVATIONANDEDCS

For example, phytoestrogens in dryinggrasses decrease reproductive output inwild populations of California quail (63)and deer mice (64). Similarly, clover dis-ease in livestock, caused by animals feedingon phytoestrogen-containing subterraneanclove (Trifolium subterraneum), inducesinfertility in adults (65). These examples,which represent extremes in the action ofphytoestrogens, and studies demonstratingthat some plant compounds are antiestro-gens, are used by some to suggest thatenvironmental estrogens represent no gen-eralized health threat because organisms areexposed to estrogen mimics or antiestro-gens everyday via natural foods. There is aflaw in this argument as it ignores thecentral paradigm of biology-evolution.

Plant-animal interactions have existedfor eons, and evolutionary theory predictsthat plants should battle herbivores becausethey represent predators (66-68).Compounds that would limit herbivorereproduction would decrease predationpressure on plants. Thus, clover disease ispredicted as a natural ongoing evolutionaryevent. Why then do animals not respondto more plant secondary compounds insuch a dramatic fashion? Evolutionary the-ory also predicts that those animals that arenot affected by the plant phytoestrogenswill reproduce at a greater rate and thus,given numerous generations, they will pro-duce a population in which most individu-als either avoid or adapt to any givenchemical. Indeed, the phytoestrogen genis-tein only elicits estrogenic effects duringspecific windows of development (69) and,thus, chronological adaptation is a plausi-ble mechanism of avoiding the estrogeniceffects of phytoestrogens. Additionally,humans physiologically respond to theexposure of the phytoestrogen genistein bystimulating sex-hormone binding globulinproduction (70) and suppressing aromataseactivity (71), both of which reduce theamount of bioavailable natural estrogen.

Physiological adaptation to an environ-mental toxic compound (acquired resis-tance) is seen in insect pests responding topesticide exposure and bacteria respondingto antibiotic treatments. We can recordthese examples of evolutionary responsedue to these organisms' very short genera-tion times relative to our own generationlength. An extension of the acquired resis-tance model suggests that vertebratesshould show a similar pattern, and it is clearthat we do (72). We eat many plant com-pounds that are toxic, and several act asendocrine disruptors, such as antithyroidal

goitrogens found in many flowering vascu-lar plants of the family Brassicae (e.g., cab-bage, brussel sprouts, rutabaga, turnips)(73). But we readily degrade many of thesesubstances metabolically so that the com-pound ingested and the subsequent break-down products of this metabolism haveminimal or no affect on our bodilyprocesses. Moreover, we do not bioaccu-mulate these compounds. However, onecannot expect acquired resistance to evolvein vertebrates within a generation or two ormaybe ever. The vertebrates living on earthtoday are exposed to a greater range of for-eign chemicals than probably at any timein our evolutionary history. Althoughmany of these synthetic compounds aremetabolized and flushed from the body,many bioaccumulate. Thus, a situationexists where a developing egg or embryo isexposed to chemicals stored over amother's lifetime. If these bioaccumulatedcompounds act as hormonal mimics, thenembryonic development can be modified.These changes may be subtle, as discussedabove, but can lead to catastrophic changeslater in life.

Free versus Bound Hormone. The toxi-cokinetics of any given EDC are extremelycomplex, but several generalities can bemade about the distribution and activity ofsuch contaminants. As discussed earlier,lipid-soluble contaminants are mobilizedfrom fat stores during energetically expen-sive reproductive events. These contami-nants are mobilized to the extracellularfluid surrounding fat stores and are eitherlocalized into an adjacent organ or trans-ported into the circulatory system. In thecase of direct organ exposure, passage intothe cell's cytoplasm and nucleus is unhin-dered as the lipophyllic compound passeseasily through the cell and nuclear mem-branes. The kinetics are much more com-plicated for EDCs entering the circulatorysystem because the biological activity of aparticular compound is not dependent onthe amount of compound in the circula-tion, but rather the amount of compoundavailable to cells.

When dealing with vertebrate hor-mones, most of a given hormone is usuallybound to plasma proteins while a verysmall amount circulates in a free form. Theplasma protein-hormone complexes arelarge and cannot cross capillary walls.Thus, only the unbound portion of hor-mone in the blood determines the hor-mone's physiological activity-an ideatermed the free hormone hypothesis (74).In essence, plasma proteins in mammals

act as storage depots for steroid hormones,but do these same plasma proteins protecttissues from EDCs? An answer can onlycome from analyses of specific contami-nant-protein complexes, but most researchon the binding of xenobiotic chemicals toplasma proteins has been conducted ondrugs, not contaminants (75). This researchhas revealed basic information that can beused, such as the binding characteristics ofalbumin. Albumin has six binding regionsand acts as a transport protein for numerousendogenous and exogenous compounds(76). However, albumin's equilibrium con-stant for steroid complexes is relatively low(77); thus, albumin offers little protectionas a storage protein. Conversely, someplasma proteins avidly bind to contaminantsas exemplified by the organochlorine pesti-cide dieldrin; 99% of dieldrin in circulationbinds to plasma proteins (78). A new tech-nique can assess the cellular availability ofspecific contaminants in the blood (79).This assay allows one to determine ifplasma constituents bind specific EDCsand restrict their availability to a cell.Using this assay, it has been demonstratedthat o,p'-DDT and its structural relativemethoxychlor have very different access-abilities. o,p'-DDT in plasma is readilyavailable to the cells whereas methoxychloris not. This study emphasizes that plasmaconstituents (proteins or lipids) can modifythe accessibility ofan EDC to a cell and thateven structurally related compounds mayact and interact in plasma in significantlydifferent ways. It is clear that futureresearch should investigate the binding ofplasma proteins or lipids to specific conta-minants. Ifwe hope to compare circulatinglevels of hormones and contaminants, it iscritical that all factors involved in cellularexposure be considered.

Future Research NeedsThe activational effects of steroid hormoneson vertebrate reproductive, immune, andneurological systems are well established,but specific mechanisms by which steroidsand endocrine-disrupting contaminants(EDCs) elicit organizational effects on thesesystems are lacking for most wildlife. This isespecially apparent when one examinesspecies other than mammals and birds. Forinstance, the mechanisms by which EDCsalter thyroid hormone concentrations infish (53,80) and common seals (81) are asyet unknown. Both the magnitude andtiming ofEDC exposure should be consid-ered when organizational effects are exam-ined. Future studies must examine which

Volume 103, Supplement 7, October 1995 161

GUILLE1TE ETAL.

receptors bind EDCs and their subsequentendocrine action. For example, does aspecific EDC bind to the estrogen orandrogen receptor and is it an antagonistor agonist? In particular, this informationis needed for those chemicals that arefound commonly in wildlife exhibitingsymptoms of EDC exposure. Informationon receptor-related mechanisms will allowthe development of assays for specific bio-markers and direct laboratory-based studiesto link symptoms with contaminant expo-sure. Additionally, data must be obtainedto determine if current models usingsteroid receptors obtained from mam-malian species are characteristic of recep-tors from other species. Given the largedegree of homology among steroid recep-tors from different species, it is likely thatthe mammalian model will provide a veryuseful tool, but this must be determined.

One aspect of the EDC problem thatmust be studied in more depth in non-

mammalian vertebrates is the degradation,recycling, and plasma partitioning of vari-ous chemicals. For example, we know thatreptiles bioaccumulate and biomagnifycontaminants, and these contaminants aretransferred to developing eggs. However,do reptiles, or any nonmammalian verte-brate, degrade various EDCs in a mannersimilar to traditional mammalian models?What is the difference in degradation orrecycling ofEDC between sexes? Oviparousfemales mobilize large amounts of lipid andprotein during follicular development. Theyolk protein, vitellogenin, is synthesized inthe liver and circulates in the blood at veryhigh concentrations. Does vitellogeninserve as a protein carrier of EDCs, thustransporting these contaminants into theeggs, as suggested that it does for naturalsteroids? Do other plasma proteins or lipidstransport EDCs and protect them fromhepatic degradation? What is the affinity ofEDCs, if any, with sex hormone-binding

globulin (SBG), corticosteroid-bindingglobulin, or plasma albumin? AlthoughSBGs have been identified in reptiles, as inother nonmammalian vertebrates, theirrole is still poorly identified under evennatural endocrine functioning. We mustdetermine what role, if any, SBGs andother plasma constituents play in protect-ing and transporting EDCs. The biologicalactivity of various EDCs will be dictated bythe amount that is free or unbound andavailable to the developing embryo.Finally, additional research to thoroughlyexamine the physiological adaptations tovarious phytoestrogens, especially in anevolutionary context, would provide greatinsight into mechanisms by which verte-brates have adapted to naturally occurringendocrine-disrupting chemicals. Suchadaptations may suggest methods wherebywe can maximize the health of exposedwildlife populations.

REFERENCES

1. Colborn T, vom Saal FS, Soto AM. Developmental effects ofendocrine-disrupting chemicals in wildlife and humans.Environ Health Perspect 101:378-384 (1993).

2. McLachlan JA. Functional toxicology: a new approach to detectbiologically active xenobiotics. Environ Health Perspect101:386-387 (1993).

3. Colborn T, Clement C, eds. Chemically-induced Alterations inSexual and Functional Development: The Wildlife/HumanConnection. Princeton, NJ:Princeton Scientific Publishing,1992.

4. Deeming DC, Ferguson MWJ. Environmental regulation ofsex determination in reptiles. Philos Trans R Soc Lond, SerBiol Sci 322:19-39 (1988).

5. Deeming DC, Ferguson MWJ. Effects of incubation tempera-ture on growth and development of embryos of Alligator missis-sippiensis. J Comp Physiol Biochem Syst Environ Physiol159:183-193 (1989).

6. Janzen FJ, Paukstis GL. Environmental sex determination inreptiles: ecology, evolution, and experimental design. Q RevBiol 66:149-179 (1991).

7. Lang JW, Andrews HV. Temperature-dependent sex determi-nation in crocodilians. J Exp Zool 270:28-44 (1994).

8. Bull JJ. Sex determination in reptiles. Q Rev Biol 55:3-21(1980).

9. Bull JJ, Gutzke WHN, Crews D. Sex reversal by estradiol inthree reptilian orders. Gen Comp Endocrinol 70:425-428(1988).

10. Gutzke WHN, Bull JJ. Steroid hormones reverse sex in turtles.Gen Comp Endocrinol 64:368-372 (1986).

11. Crews D, Wibbels T, Gutzke WHN. Action of sex steroid hor-mones on temperature-induced sex determination in the snap-ping turtle (Chelydra serpentine). Gen Comp Endocrinol76:159-166 (1989).

12. Bergeron JM, Crews D, McLachlan JA. PCBs as environmentalestrogens: turtle sex determination as a biomarker of environ-mental contamination. Environ Health Perspect 102:780-781(1994).

13. Guillette LJ Jr, Gross TS, Masson GR, Matter JM, PercivalHF, Woodward AR. Developmental abnormalities of the

gonad and abnormal sex hormone concentrations in juvenilealligators from contaminated and control lakes in Florida.Environ Health Perspect 102:680-688 (1994).

14. Guillette LJ Jr, Gross TS, Gross D, Rooney AA, Percival HF.Gonadal steroidogenesis in vitro from juvenile alligatorsobtained from contaminated and control lakes. Environ HealthPerspect 103(Suppl 4):31-36 (1995).

15. Crews D, Bull JJ, Wibbels T. Estrogen and sex reversal in tur-tles: a dose-dependent phenomenon. Gen Comp Endocrinol81:357-364 (1991).

16. Lance VA, Bogart MH. Disruption of ovarian development inalligator embryos treated with an aromatase inhibitor. GenComp Endocrinol 86:59-71 (1992).

17. Bern HA. The "new" endocrinology: its scope and its impact.Am Zool 30:877-885 (1990).

18. vom Saal FS, Montano MM, Wang MH. Sexual differentiationin mammals. In: Chemically-induced Alterations in Sexual andFunctional Development: The Wildlife/Human Connection(Colborn T, Clement C, eds). Princeton, NJ:PrincetonScientific Publishing, 1992;17-83.

19. Girbau M, Gonzalez-Guerrero PR, Bassas L, De Pablo F.Insulin receptors and insulin-like growth factor I receptors inembryos from gastrula until organogenesis. Mol CellEndocrinol 90:69-75 (1992).

20. Serrano J, Shuldiner AR, Roberts CT Jr, LeRoith D, De PabloF. The insulin-like growth factor I (IGF-I) gene is expressed inchick embryos during early organogenesis. Endocrinologl127:1547-1549 (1990).

21. Scavo L, Shuldiner AR, Serrano J, Dashner R, Roth J, DePablo F. Genes encoding receptors for insulin and insulin-likegrowth factor I are expressed in Xenopus oocytes and embryos.Proc Natl Acad Sci USA 88:6214-6218 (1991).

22. Scavo L, Alemany J, Roth J, De Pablo F. Insulin-like growthfactor I activity is stored in the yolk of the avian egg. BiochemBiophys Res Commun 162:1167-1173 (1989).

23. Guillette LJ Jr, Williams PE. Detection of insulin-like growthfactors I and II in the egg albumen of the American aligator(Alligator mississippiensis) using microcapillary electrophoresis.Biol Reprod 44(Suppl 1):58 (1991).

162 Environmental Health Perspectives

ORGANIZATION VERSUS ACTIVATIONAND EDCS

24. Brown CL, Bern HA. Thyroid hormones in early development,with special reference to teleost fishes. In: Development,Maturation, and Senescence of Neuroendocrine Systems: AComparative Approach (Schreibman MP, Scanes CG, eds).New York:Academic Press, 1989;289-306.

25. Feist G, Fitzpatrick MS, Redding JM, Schreck CB. Sex steroidprofiles of coho salmon, Oncorhynuchus kisutch, during earlydevelopment and sexual differentiation. Gen Comp Endocrinol79:233-240 (1990).

26. Colborn TE, Davidson A, Green SN, Hodge RA, Jackson CI,Liroff RA. Great Lakes Great Legacy. Washington:TheConservation Foundation, 1990.

27. Cooper K. Effects of pesticides on wildlife. In: Handbook ofPesticide Toxicology. Vol 1 (Hayes WJ, Laws ER Jr, eds). NewYork:Academic Press, 1991 ;463-495.

28. Olafsson PG, Bryan AM, Bush B, Stone W. Snapping turtles: abiological screen for PCBs. Chemosphere 12:1525-1532(1983).

29. Bryan AM, Stone WB, Olafsson PG. Disposition of toxic PCBcongeners in snapping turtle eggs: expressed as toxic equivalentsofTCDD. Bull Environ Contam Toxicol 39:791-796 (1987).

30. Bishop CA, Brooks RJ, Carey JH, Ng P, Norstrom RJ, LeanDRS. The case for a cause-effect linkage between environmen-tal contamination and development in eggs of the commonsnapping turtle (Chelydra s. serpentine) from Ontario, Canada. JToxicol Environ Health 33:521-547 (1991).

31. Heinz GH, Percival HF, Jennings ML. Contaminants inAmerican alligator eggs from lakes Apopka, Griffin andOkeechobee, Florida. Environ Monit Assess 16:277-285(1991).

32. Struger J, Elliott JE, Bishop CA, Obbard ME, Norstrom RJ,Weseloh DV, Simon M, Ng P. Environmental contaminants ineggs of the common snapping turtle (Chelydra serpentine ser-pentina) from the Great Lakes-St. Lawrence River Basin ofOntario, Canada (1981, 1984). J G Lakes Res 19:681-694(1993).

33. Arnold AP, Breedlove SM. Organizational and activationaleffects of sex steroids on brain and behavior: a reanalysis. HormBehav 10:469-498 (1985).

34. Phoenix CH, Goy RW, Gerall AA, Young WC. Organizingaction of prenatally administered testosterone propionate onthe tissues mediating mating behavior in the female guinea pig.Endocrinology 65:369-382 (1959).

35. Wibbels T, Crews D. Putative aromatase inhibitor inducesmale sex determination in a female unisexual lizard and in aturtle with temperature-dependent sex determination. JEndocrinol 141:295-299 (1994).

36. Sharpe RM. Regulation of spermatogenesis. In: The Physiologyof Reproduction (Knobil E, Neill JD, eds). New York:RavenPress, 1994;1363-1434.

37. vom Saal FS, Finch CE, Nelson JF. Natural history and mecha-nisms of reproductive aging in humans, laboratory rodents, andother selected vertebrates. In: The Physiology of Reproduction(Knobil E, Neill JD, eds). New York:Raven Press,1994; 1213-1314.

38. Bern H. The fragile fetus. In: Chemically-induced Alterationsin Sexual and Functional Development: The Wildlife/HumanConnection (Colborn T, Clement C, eds). Princeton,NJ:Princeton Scientific Publishing, 1992;9-15.

39. Luster MI, Hayes HT, Korach K, Tucker AN, Dean JH,Greenlee WF, Boorman GA. Estrogen immunosuppression isregulated through estrogenic responses in the thymus. JImmunol 133:110-116 (1984).

40. Dohler KD, Jarzab B. The influence of hormones and hormoneantagonists on sexual differentiation of the brain. In:Chemically-induced Alterations in Sexual and FunctionalDevelopment: The Wildlife/Human Connection (Colborn T,Clement C, eds). Princeton, NJ:Princeton ScientificPublishing, 1992;231-259.

41. Uchima FDA, Vallerga AK, Firestone GL, Bern HA. Effects ofearly exposure to diethylstilbestrol on cellular protein expres-sion by mouse vaginal epithelium and fibromuscular wall. Proc

Soc Exp Biol Med 195:218-224 (1990).42. Newbold RR, Pentecost BT, Yamashita S, Lum K, Miller JV,

Nelson P, Blair J, Kong J, Teng C, McLachlan JA. Female geneexpression in the seminal vesicle of mice after prenatal exposureto diethylstilbestrol. Endocrinology 124:2568-2576 (1989).

43. Takamatsu Y, Iguchi T, Takasugi N. Effects of neonatal expo-sure to diethylstilbestrol on protein expression by vagina anduterus in mice. In Vivo 6:1-8 (1992).

44. McLachlan JA, Newbold RR, Teng CT, Korach KS.Environmental estrogens: orphan receptors and geneticimprinting. In: Chemically-induced Alterations in Sexual andFunctional Development: The Wildlife/Human Connection(Colborn T, Clement C, eds). Princeton, NJ:PrincetonScientific Publishing, 1992;107-112.

45. Kalland T. Decreased and disproportionate T-cell populationin adult mice after neonatal exposure to diethylstilbestrol. CellImmunol 51:55-63 (1980).

46. Kalland T, Forsberg J-G. Delayed hypersensitivity response tooxazolone in neonatally estrogenized mice. Cancer Lett4:141-146 (1978).

47. Kalland T, Strand 0, Forsberg J-G. Long-term effects ofneonatal estrogen treatment on mitogen responsiveness ofmouse spleen fymphocytes. J Natl Cancer Inst 63:413-421(1979).

48. Bateman A, Singh A, Krak T, Solomon S. The immune-hypo-thalamic-pituitary-adrenal axis. Endocr Rev 10:92-112 (1989).

49. Besedovsky HO, Sorkin E. Network of immune-neuroen-docrine interactions. Clin Exp Immunol 27:1-12 (1977).

50. Herbst AL, Bern HA, eds. Developmental Effects ofDiethylstilbestrol (DES) in Pregnancy. New York:Thieme-Stratton, 1981.

51. Mori T, Mills KT, Bern HA. Sensitivity of the vagina anduterus of mice neonatally exposed to estrogens or an rogen topostnatal treatment with estrogen or androgen. Proc Soc ExpBiol Med 199:466-469 (1992).

52. Ostrander PL, Mills KT, Bern HA. Long-term responses of themouse uterus to neonatal diethylstilbestrol treatment and latersex hormone exposure. J Natl Cancer Inst 74:121-135 (1985).

53. Leatherland JF. Endocrine and reproductive function in GreatLakes salmon. In: Chemically-induced Alterations in Sexualand Functional Development: The Wildlife/HumanConnection (Colborn T, Clement C, eds). Princeton,NJ:Princeton Scientific Publishing, 1992; 129-146.

54. Davis WP, Bortone SA. Effects of kraft mill effluent on the sex-uality of fishes: an environmental early warning? In:Chemically-induced Alterations in Sexual and FunctionalDevelopment: The Wildlife/Human Connection (Colborn T,Clement C, eds). Princeton, NJ:Princeton ScientificPublishing, 1992;1 13-127.

55. Purdom CE, Hardiman PA, Bye VJ, Eno NC, Tyler CR,Sumpter JP. Estrogenic effects of effluents from sewage treat-ment works. Chem Ecol 8:275-285 (1994).

56. Hose JE, Cross JN, Smith SG, Diehl D. Reproductive impair-ment in a fish inhabiting a contaminated coastal environmentoff of Southern California. Environ Pollut 57:139-148 (1989).

57. Fry DM, Toone CK. DDT-induced feminization of gullembryos. Science 213:922-924 (1981).

58. Bowerman WB, Best D, Kubiak T, Postupalsky S, Tillett D.Bald eagle reproduction impairment around the Great Lakes:association with organochlorine contamination. In: Cause-Effect Linkages II. Symposium Abstracts (Schneider S,Campbell R, eds). Ann Arbor, MI:Michigan Audubon Society,1991;31-32.

59. Biessmann A. Effects of PCBs on gonads, sex hormone balanceand reproduction processes of Japanese quail Coturnix coturnixjaponica after ingestion during sexual maturation. EnvironPollut, SerA Ecol Biol 27:15-30 (1982).

60. Subramanian AN, Tanabe S, Tatsukawa R, Saito S, MiyazakiN. Reduction in the testosterone levels by PCBs and DDE inDall's porpoises of Northwestern North Pacific. Mar PollutBull 18:643-646 (1987).

61. Beland P, DeGuise S, Plante R. Toxicology and pathology of

Volume 103, Supplement 7, October 1995 163

GUILLETTE ETAL.

St. Lawrence marine mammals. Washington:WildlifeToxicology Fund, World Wildlife Fund, 1992; 110 pp.

62. Munkittrick KR, Port CB, Van Der Kraak GJ, Smith IR,Rokosh DA. Impact of bleached kraft mill effluent on popula-tion characteristics, liver MFO activity, and serum steroids ofthe Lake Superior white sucker (Catostomus commersoni) popu-lation. Can J Fish Aquat Sci 48:1-10 (1991).

63. Leopold AS, Erwin M, Oh J, Browning B. Phytoestrogens:adverse effects on reproduction in California quail. Science191:98-100 (1976).

64. Berger PJ, Sanders EH, Gardner PD, Negus NC. Phenolicplant compounds functioning as reproductive inhibitors inMicrotus montanus. Science 195:575-577 (1977).

65. Bennetts HW, Underwood EJ, Sheir FLA. A specific breedingproblem of sheep on subterranean clover pastures in WesternAustralia. Aust Vet J 22:2-12 (1946).

66. Ehrlich P, Raven PH. Butterflies and plants: a study of coevo-lution. Evolution 18:586-608 (1964).

67. Jansen DH. When is it coevolution? Evolution 34:611-612(1980).

68. Hughes CL Jr. Phytochemical mimicry of reproductive hor-mones and modulations of herbivore fertility by phytoestro-gens. Environ Health Perspect 78:171-175 (1988).

69. Levy JR, Faber KA, Ayyash L, Hughes CL. The effect of prena-tal exposure to the phytoestrogen genistein on markers of sex-ual differention in rats. Proc Soc Exp Biol Med 208:60-66(1995).

70. Mousavi Y, Aldercreutz H. Genistein is an effective stimulatorof sex hormone binding globulin production in hepatocarci-noma human liver cancer cells and suppresses proliferation ofthose cells in culture. Steroids 58:301-304 (1993).

71. Adlercreutz H, Bannwart C, Wahala K, Makela T, Brunow G,Hase T, Arosemena PJ, Kellis TT, Vickery LE. Inhibition of

human aromatase by mammalian lignans and isoflavonoid phy-toestrogens. J Steroid Biochem Mol Biol 44:147-153 (1993).

72. Howe HF, Westley LC. Ecological Relationships of Plants andAnimals. New York:Oxford University Press, 1988.

73. Yamada T, Kajihara A, Takemura Y, Onaya T. Antithyroid com-pounds. In: Handbook of Physiology. Sec. 7: Endocrinology.Baltimore:Williams and Wilkins, 1974;345-357.

74. Mendel CM. The free hormone hypothesis: a physiologicallybased mathematical model. Endocr Rev 10:232-274 (1989).

75. Klaassen CD, Rozman K. Absorption, distribution, and excre-tion of toxicants. In: Casarett and Doull's Toxicology: TheBasic Science of Poisons (Amdur MO, Doull J, Klaassen CD,eds). New York:McGraw-Hill, 1993; 12-49.

76. Kragh-Hansen U. Molecular aspects of ligand binding to serumalbumin. Pharmacol Rev 33:17-53 (1981).

77. Tait JF, Tait SAS. The effect of plasma protein binding on themetabolism of steroid hormones. J Endocrinol 131:339-357(1991).

78. Heath DF, Vandekar M. Toxicity and metabolism of dieldrinin rats. Br J Ind Med 21:269-279 (1964).

79. vom Saal FS, Nagel SC, Palanza P, Boechler M, Parmigiani S,Welshons WV. Estrogenic pesticides: binding relative to estra-diol in MCF-7 cells and effects of exposure during fetal life onsubsequent territorial behavior in male mice. Toxicol Lett (inpress).

80. Sinha N, Lal B, Singh TP. Pesticide induced changes in circu-lating thyroid hormones in the freshwater catfish Clarias batra-chus. Comp Biochem Physiol 100C:107-110 (1991).

81. Brouwer A, Reijnders PJH, Koeman JH. Polychlorinatedbiphenyl (PCB)-contaminated fish induces vitamin A and thy-roid hormone deficiency in the common seal (Phoca vitalina).Aquat Toxicol 15:99-106 (1989).

164 Environmental Health Perspectives