Male Reproductive Health and Environmental Xenoestrogens

Home , Isoflavonoid, Reproductive toxicity

Male reproductive health and environmental xenoestrogens

The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters

Citation Toppari J, Larsen JC, Christiansen P, Giwercman A, Grandjean P, Guillette LJ Jr, Jégou B, Jensen TK, Jouannet P, Keiding N, Leffers H, McLachlan JA, Meyer O, Müller J, Rajpert-De Meyts E, Scheike T, Sharpe R, Sumpter J, Skakkebaek NE. 1996. Male reproductive health and environmental xenoestrogens. Environmental Health Perspectives 104 (Supp 4).

Citable link http://nrs.harvard.edu/urn-3:HUL.InstRepos:34787314

Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of- use#LAA Male Reproductive Health and Introduction In 1992, a research group at the Department Environmental Xenoestrogens of Growth and Reproduction, the National University Hospital (Rigshospitalet), and the Jorma Toppari,12 John Chr. Larsen,3 Peter Christiansen,' Panum Institute, Copenhagen, Denmark, Aleksander Giwercman,1 Philippe Grandjean,4 published in the British MedicalJournal a interna- Louis J. Guillette Jr.,5 Bernard Jegou,6 Tina K. Jensen, metaanalysis of the data from the tional literature that revealed a significant Pierre Jouannet,7 Niels Keiding,8 Henrik Leffers,l decrease in sperm concentration and semen John A. McLachlan,9 Otto Meyer,10 J0rn MIlIer,' volume in otherwise normal men over the Ewa Rajpert-De Meyts,l Thomas Scheike,1'8 Richard period 1938 to 1990. During the same time period, the incidence of testicular Sharpe,"1 John Sumpter,12 and Niels E. Skakkebaekl"13 cancer had markedly increased in many 'Department of Growth and Reproduction, Juliane Marie Center, countries. These and other observations provided a clue that this apparent decline National University Hospital, Copenhagen, Denmark; 2Departments of in male reproductive health might be Pediatrics and Physiology, University of Turku, Turku, Finland; 3National caused by some common environmental Food Agency, S0borg, Denmark; 4Department of Environmental factors. It was recognized that similar abnor- Medicine, Odense University, Odense, Denmark; 5Department of malities of the male reproductive system Zoology, University of Florida, Gainesville, Florida; 6GERM-INSERM U were caused by administration of estrogens 435, Campus de Beaulieu, 35042 Rennes Cedex, France; 7Universite Paris during pregnancy in humans and experi- V, Groupe Hospitalier Cochin, Paris, France; 8Department of Biostatistics, mental animals; therefore, a hypothesis was University of Copenhagen, The Panum Institute, Copenhagen, Denmark; put forward that environmental chemicals having estrogenic effects were contributing of 9Tulane/Xavier Center for Bioenvironmental Research, Department agents. In particular, it was suggested that Pharmacology, Tulane University, New Orleans, Louisiana; 10lnstitute of fetal exposure to an excess of estrogenic Toxicology, Department of General Toxicology, S0borg, Denmark; compounds was a key risk factor. 1MRC Unit of Reproductive Biology, Edinburgh, United Kingdom; After an intense public debate in the 12Department of Biology and Biochemistry, Brunel University, Uxbridge, Danish news media on the possible role of Middlesex, United Kingdom environmental chemicals, such as pesticides, detergents, plasticizers, and other industrial Male reproductive health has deteriorated in many countries during the last few decades. In the chemicals, the Danish Environmental 1990s, declining semen quality has been reported from Belgium, Denmark, France, and Great Protection Agency (DEPA) of the Ministry Britain. The incidence of testicular cancer has increased during the same time. Incidences of of Environment and Energy in September hypospadias and cryptorchidism also appear to be increasing. Similar reproductive problems 1994 decided to support the preparation of occur in many wildlife species. There are marked geographic differences in the prevalence of a review summarizing the current knowl- male reproductive disorders. While the reasons for these differences are currently unknown, both edge on male reproductive disorders and clinical and laboratory research suggest that the adverse changes may be inter-related and have a environmental chemicals with estrogenic common origin in fetal life or childhood. Exposure of the male fetus to supranormal levels of effects. In addition, the review was to iden- estrogens, such as diethlylstilbestrol, can result in the above-mentioned reproductive defects. tify gaps in knowledge and address research The growing number of reports demonstrating that common environmental contaminants and needs and requirements in order for natural factors possess estrogenic activity presents the working hypothesis that the adverse researchers to perform adequate risk assess- trends in male reproductive health may be, at least in part, associated with exposure to estro- ments. The DEPA asked Professor Niels E. genic or other hormonally active (e.g., antiandrogenic) environmental chemicals during fetal and Skakkebaek, Department of Growth and An is needed to understand the extent of childhood development. extensive research program Reproduction at the National University the problem, its underlying etiology, and the development of a strategy for prevention and inter- Hospital and John Chr. Larsen, Division vention. Environ Health Perspect 104(Suppl 4):741-803 (1996) Head, Institute of Toxicology at the Danish Key words: male reproduction, reproductive disorders, semen quality, testicular neoplasms, National Food Agency to prepare the environment, estrogenic chemicals, endocrine disruptors, exposure, pesticides report. The project received additional

Manuscript received 27 April 1995; manuscript accepted 15 March 1996. Supported by EU Contract BMH4-CT96-0314. Address correspondence to Dr. Niels E. Skakkebaek, Dept. of Growth and Reproduction, Rigshospitalet, Section GR 5064, 9 Blegdamsvej, DK-21 00 Copenhagen 0, Denmark. Telephone: 45 35 45 50 85. Fax: 45 35 45 60 54. E-mail: [email protected] Abbreviations used: ADI, acceptable daily intake; AFP, ax-fetoprotein; Ah, aryl hydrocarbon; APE, alkylphenol polyethoxylate; BBP, butylbenzyl phthalate; BHA, butylated hydroxyanisole; BHC, benzene hexachloride; bw, body weight; CIS, carcinoma in situ; CNS, central nervous system; CP, cyclophosphamide; 2,4-D, 2,4-dichlorophenoxyacetic acid; DBCB, dibromochloropropane; DBP, di-n-butyl phthalate; DDD, 1,1-dichloro-2,2-bis(p-chlorophenyl)ethane; DDE, 1,1-dichloro-2,2- bis(p-chlorophenyl)ethylene; DDT, 1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane; DEHP, di(2-ethylhexyl)phthalate; DEPA, Danish Environmental Protection Agency; DES, diethylstilbestrol; EBDC, ethylenebisdithiocarbamate; EHC, Environmental Health Criteria; EU, European Union; FDA, U.S. Food and Drug Administration; FSH, follicle-stimulating hormone; GAP, Good Agricultural Practice; GLP, Good Laboratory Practice; HCB, hexachlorobenzene; HCH, hexachlorocyclohexane; (continued)

Environmental Health Perspectives - Vol 104, Supplement 4 - August 1996 741 TOPPARI ET AL.

financial support from the Ministry of occupational toxicants were shown to rather than a true biological phenomenon. Health through its National Research affect the semen quality of the workers A systematic metaanalysis of 61 studies Centre for Environmental Medicine. through a toxic action on the gonads, with- that included 14,947 normal men revealed Dr. Jorma Toppari, Departments of out any apparent estrogenic effects. Such a significant decrease in sperm concentra- Pediatrics and Physiology, University of toxic effects are not the object of the pre- tion (113 million/ml vs 66 million/ml; Turku, Finland, prepared a draft of the sent report, but should be kept in mind in Figure 1) and semen volume (3.40 ml vs report. This draft was then discussed by a any consideration or scientific investigation 2.75 ml) over the period of 1938 to 1990 group of Danish and international experts of the adverse effects of environmental (5). This report stimulated extensive dis- invited to a one-week workshop that was chemicals on male reproductive health. cussion and some criticisms on the basis of held at Rigshospitalet in Copenhagen, possible technical errors and known limita- January 23-27, 1995. The participants Secular Trends in the tions of metaanalysis (6,7). Carlsen and who are the authors of the present publica- Incidence of Male co-workers responded to these criticisms tion actively contributed to the endeavor Reproductive Disorders (8). Although the data for 1970 to 1990 both during the workshop and afterward. were compatible with a decrease as well as The final manuscript was edited by Jorma Trends in Semen Quality with no change or an increase in semen Toppari together with Niels E. Skakkebak Several reports in the literature have sug- quality, the cautious general conclusion was and John Chr. Larsen. The present review gested a possible decline in human semen that a real decline occurred during the 50- is a revised version of the official report quality during the last 50 to 60 years year period (9). The findings of Carlsen et (1), which was printed by the Danish (2-4). However, most of these reports al. (5) were also compared (6) to those of Ministry of Environment and Energy were based on data from men attending MacLeod and Wang (10) from the United mainly for circulation in Denmark. infertility clinics or from very selected States. This comparison is not relevant, The review addresses the possible groups of fertile men and, therefore, the however, because the metaanalysis was effects of environmental chemicals known decline in sperm counts was presumed just based on semen analyses of normal men, to possess estrogenic activity on male to reflect changes in the policy of infertility whereas the American study examined men reproductive health. The term xenoestro- treatment or a bias in selection of patients who were clients ofan infertility clinic. gen is often used for such compounds, whereas the term synthetic estrogens refers to medical drugs mainly used for contra- 150 - ception and treatment ofvarious diseases. A number of other environmental l 10~~~~~~~~~ 0 chemicals have been implicated as environ- 0 mental hormones or endocrine disruptors. 0 Although not shown (and in many 100-0 instances not adequately tested) to possess A 0 Q. 9~~~~~~~~~~~~~~ direct estrogenic activity, some of these 0 00 compounds may in some cases also affect C a, 0 0 .0 male reproduction. The mechanisms of 0c o ~ ~ o 8 action are not known in detail but they 0 may involve, for example, antiandrogenic 50 - activity; modulatory effects on enzymes controlling sex hormone metabolism; or direct influence on the hormone-producing organs such as the thyroid gland, pituitary gland, and adrenal glands. These compounds may also affect estrogen levels 0 through indirect feedback mechanisms. 1940 1950 1960 1970 1980 1990 The authors are well aware that the decline in semen quality and the increase Year in the incidence of testicular cancer may Figure 1. Linear regression of mean sperm density reported in 61 publications (represented by circles the area of be caused by many other environmental, which is proportional to the logarithm of the number of subjects in the study), each weighted according to number life-style, or genetic factors. For example, of subjects, 1938-1990. The figure is based on the data reported by Carlsen et al. 15). A corresponding figure in some chemicals that are now known as that paper was incomplete.

Abbreviations used (continued): HPLC, high-performance liquid chromatography; JMPR, Joint FAO/NHO Expert Meeting on Pesticide Residues; LH, luteinizing hormone; LOAEL, lowest observed adverse effects level; MIS, Mullerian inhibiting substance; MRL, maximal residue limits; NOAEL, no observed adverse effect level; NOEL, no observed effect level; NP, nonylphenol; OECD, Organisation for Economic Cooperation and Development; OP, octylphenol; OPPT, Office of Pollution Prevention and Toxics; OPPTS, Office of Prevention, Pesticides and Toxic Substances; PCB, polychlorinated biphenyl; PCDD, polychlorinated dibenzo-p-dioxin; PCDF, polychlorinated dibenzofuran; PCP, pentachlorophenol; PCT, porphyria cutanea tarda; PE, proliferative efficiency; PNS, peripheral nervous system; 2,4,5-T, 2,4,5- trichlorophenoxyacetic acid; RfD, reference dose; RPE, relative proliferative efficiency; RPP, relative proliferative potency; SHBG, sex hormone-binding globulin; STW, sewage treatment water; TBT, tributyltin; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; TCDD-TEQ, TCDD equivalency; TDI, tolerable daily intake; TEF, toxic equivalency factor; TWI, tolerable weekly intake; U.S. EPA, U.S. Environmental Protection Agency.

742 Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

The metaanalysis of Carlsen et al. (5) sperm count of samples from 302 healthy A 500 I prompted several laboratories to evaluate fertile donors was 83 million/ml (15). I their data on the quality of semen obtained Furthermore, the sperm concentration in i.I 100 - i;- during recent years. In a French study of semen of Finnish men has remained , ., E II iI 1351 healthy men volunteering for sperm unchanged between 1958 and 1992 (111 50 - Ii X, * donation, a 2.1% decrease in sperm con- million/ml vs 124 million/ml) and is :: * !, I. from million/ml in higher than elsewhere in Europe (16). It is . centration per year 89 CD i. *: 8 1973 to 60 million/ml in 1992 (p<0.001) of interest that the incidence of testicular 10 - was found (11). Furthermore, the percent- cancer, and perhaps also hypospadias, in z

5- . ages of motile and normal spermatozoa also Finland is much lower than that in other I decreased significantly (Figure 2), whereas Nordic countries (below), suggesting that semen volume remained unchanged (3.8 these phenomena may be related in some ml). It is notable that the year of birth of way. The reason remains unknown, but the study subjects contributed significantly further examination may provide impor- 1973 1982 1992 to the results. Multiple regression analysis tant clues to the etiology of decreasing B 1o0- (which allows for separate effects of age sperm quality worldwide. Urban areas

to a . . . . . at (e.g., Paris) appear have declining . showed . . . and calendar year birth) yearly 80 - . . decreases of 2.6% in sperm concentration, trend in sperm counts, whereas rural areas . 0.3% in motility percentage, and 0.7% in (e.g., Toulouse or Finland) seem to have the percentage of normal spermatozoa stable sperm concentrations in semen. 60 - fE ...... according to the year of birth of the men V1) . . . . . Incidence of Testicular Cancer -2 . . . . . (all changes, p < 0.001) (11). Similar results 40 - ...... were obtained in a Scottish study (12) of Testicular cancer is now the most com- 577 semen donors where a correlation was mon malignancy of young men in many found between the median sperm count and countries; and although it is still rare com- 20 - the year of birth; the median sperm concen- pared to the malignant diseases most preva- tration decreased from 98 million/ml lent in old age, the lifetime risk of o - among donors born before 1959 to 78 mil- developing testicular cancer now approaches 1973 1982 1992 lion/ml among donors born after 1970 1% in a country such as Denmark. The (p=0.002). The total number of sperm in incidence of testicular cancer has increased C 100- the ejaculate fell from 301 million to 214 for several decades (17). On the basis of million (p = 0.0005)(12). The association data from cancer registries, increases in inci- 80 - between declining semen quality and a dence are evident in England and Wales E more recent year of birth lends support to (18,19), Scotland (20), the Nordic and 60- the concept that adverse prenatal factors Baltic countries (21,22), Australia (23), may influence the sperm production capac- New Zealand (24,25), and the United ity in adult life. Deterioration of sperm States (26). The observed increase has 40 - counts as well as motility among semen been approximately 2 to 4% per annum in donor candidates during the past two men under 50 years of age (Figures 3, 4) 20- decades was also observed in a smaller and occurred in the same age group in Belgian study (13). Ginsburg and Hardiman which testicular cancer incidence peaks, 0 - (14) reported a decrease in sperm concentra- i.e., young adults (Figure 5). Table 1 dis- 1973 1982 1992 tions (105 million/ml in 1978-1983 vs 76 plays the changes that have occurred dur- million/ml in 1984-1989) of the partners of ing the last 25 years. There are marked Year of donation women treated for infertility and living in racial and geographic differences. For Figure 2. Changes in the sperm concentration (A),the the Thames water supply area of London, example, Denmark has a 4-fold higher percentage of motile sperm (B), and the percentage of whereas no decrease was found among those incidence of testicular cancer than does morphologically normal sperm (C) in 1351 fertile men, who lived in other water supply areas of nearby Finland, and Caucasians are 3-fold 1973-1992. Linear regression analysis revealed a London. However, the mean percentage of more susceptible to this disease than are decrease of 2.1% per year in the mean sperm concen- abnormal spermatozoa increased in all African Americans in the United States. tration, from 89x 106 per milliliter in 1973 to 60x 106 areas during the study period (18-19% vs Nevertheless, it is obvious that there is a per milliliter in 1992. The concomitant decreases in the 30-32%) (14). The data in all the studies worldwide trend toward an increased inci- mean percentages of motile and normal spermatozoa cited above originated in individual labo- dence of testicular cancer as illustrated in were 0.6 and 0.5% per year, respectively. Reproduced ratories that used consistently the same Figures 3-5. The incidences of both semi- with permission from Auger et al. (11). methods for semen analysis throughout nomas and nonseminomas have increased the period. (17). Mortality due to testicular cancer of testicular cancer and cannot therefore The decreasing trend in semen quality increased from the beginning of this cen- develop any preventive measures. may not be global. In contrast to the Paris tury until the early 1970s when, because of area, no change in sperm concentration the development of good medical treat- Incidence ofCryptorchidism was found in the Toulouse area in France ment, mortality began to decline (17). Birth data from several reports have indi- during 1977 to 1992 (15). The mean However, we still do not know the etiology cated a substantial increase in the incidence

Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 743 TOPPARI ET AL.

A t'1. I P. I5

IV i i i I P g ia Ia ulII5 L.

Ia a II I i .ia I

B E

c F

Figure 3. Secular, racial, and geographic trends in the incidence of testicular cancer, 1953-1987. Compilation of data from IARC (318-323).

of cryptorchidism (maldescent of the testis). (27-48); 0.16 to 13.3% in surveys from difference. Racial and ethnic data are pooled However, estimates of the prevalence of school, army, etc. (36,38,49-57); and 2 to in most studies. Unfortunately, very few cryptorchidism obtained from different 4.7% in cohort studies based on discharge studies exist examining temporal changes in studies are difficult to compare. It is often diagnosis (41,58,59). A few studies include the incidence of cryptorchidism, confined not clear how a cryptorchid testis was ethnic data on non-Caucasians: birth data to the same population and geographic defined, and inclusion of different propor- from India (33), Formosa (Taiwan) (38), areas and using an identical definition of tions of boys with retractile testes could and Korea (44) indicated prevalences of the condition. account for the reported differences. The 1.6, 1.4, and 0.7% of cryptorchidism, Discharge data from the Hospital sources of data used in these reports also respectively. A school survey from Nigeria Inpatient Enquiry from England and differ considerably. The prevalence rates (56) indicated a prevalence of 0.5%. The Wales showed that the proportion of boys have varied between 0.03 and 13.4% on incidence of cryptorchidism among African undergoing orchidopexy (operation to the basis of data from birth to 1 year of age Americans was reported to be only one- bring the testis into the scrotum) before 15 from hospital or central registers (including third that among whites (34), although years of age increased from 1.4% for a different proportions of preterm babies) another study (48) did not find a significant 1952 birth cohort to 2.9% for a 1977 birth

744 Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

10.0 indicated an incidence of cryptorchidism Denmark ofapproximately 2% (59). Lithuania In a study from the late 1950s that v Estonia examined more than 3500 male infants - Poland ,- delivered in a hospital in London and fol- 7-5- Norway lowed up to 1 year of age, Scorer (31) Saar-Filand found that the incidences of cryp- at LO S aGermany torchidism 3 months of age in boys with CD .------Sweden birth weights <2500 g and >2500 g were a) 5.0- Latvia 1.74 and 0.91%, respectively. Scorer used very accurate definitions of the positions of yg7+r.--testes, and therefore, this well-conducted study has served as a reference for later 2.5 ~ ...... research. In a large study from the 1980s, 2.5 ------comprising 7441 male infants from Oxford (47), the very same examination technique ,,,,,_ and definitions of cryptorchidism were used; the rates of cryptorchidism at the age oo9 of 3 months in boys with birth weights 1943 1948 1953 1958 1963 1968 1973 1978 1983 1988 < 2500 g and > 2500 g were 5.2 and 1.61%, respectively, indicating a significant 3-year moving average was used increase compared to Scorer's figures. In another study from the late 1980s com- Fi gure 4. Trends in age-standardized (world standard population) incidence rates of testicular cancer. From Adami posed of 6935 male infants from New York (8 (us ing altstudyNes et al. (22); reprinted with permission from Wiley-Liss, Inc. York (48) (using identical study techniques and case definition), prevalence rates of cryptorchidism at the age of 3 months in 25 A Sweden boys with birth weights <2500 g and v Denmark > 2500 g were 1.94 and 0.91%, respec- /Z' Norway tively. However, the study population was 20 * Finland racially and ethnically heterogeneous. o Poland From these three large studies, one can * East Germany conclude that there has been a significant 15 - increase in the incidence of cryptorchidism a1) in England, but the incidence in the a1) racially and ethnically mixed population of a) 10 , \ < w vNew York is similar to that reported in the 10\w < \ \ 1950s in England. `1g \ ~N \The epidemiological data indicative of an increasing incidence of cryptorchidism 5 o G 1\ ~ \ :g are not unequivocal. This important issue necessitates large regional prospective studies in which standard criteria are 0 adopted for an accurate description of 20,24 2529 3034 35-39 4044 45-49 5054 5559 60-4 6569 70-74 75-79 80+ cryptorchidism.

Age Incidence ofHypospadias Birth data from several reports have indi- Figure 5. Age-specific incidence of testicular cancer 1985-1989 in the Nordic countries, Poland, and former East Germany. From Adami et al. (22); reprinted with permission from Wiley-Liss, Inc. cated a substantial increase in prevalence of hypospadias (Figures 6, 7). Figures of birth prevalence of hypospadias in the world lit- cohort (58). However, it is not known of cryptorchidism at birth in male infants erature vary considerably-from 0.37 to 41 whether this is a reflection of a true increase weighing > 2500 g varied between 1 and per 10,000 infants (61,62)-and are in the prevalence of cryptorchidism or 1.8% in three different data sets obtained difficult to compare. There are several fac- whether a considerable number of boys in the late 1950s (60). School surveys sug- tors that may contribute to the reported with retractile testes were undergoing orchi- gested higher prevalence rates up to 7% differences: different levels of ascertain- dopexy. In Scotland, the annual number of during 1940 to 1966, but these figures ment, different inclusion of minor forms of discharges of boys 0 to 14 years of age with apparently included retractile testes (55). hypospadias or differences in ethnical the diagnosis of cryptorchidism also Cohort analysis of data from the Danish origin of the population. As reported for showed a substantial increase during 1961 National Register of Hospital In- and cryptorchidism, very few longitudinal to 1985 (41). In Denmark, the prevalence Outpatients, from the period 1982 to 1985, studies confined to the same population

Environmental Health Perspectives - Vol 104, Supplement 4 - August 1996 745 TOPPARI ET AL.

Table 1. The incidence of testicular caancer, 1953 to 1987.a and geographic area exist. The increasing incidence of hypospadias has been reported 19160-1963 1964-1968 1969-1972 1973-1977 1978-1982 1983-1987 primarily in England and Wales (39), Nordic countries Hungary (63,64), Sweden (65-67), Denmark 3.8 4.5 4.9 6.7 7.8 8.4 Norway (67,68) and Denmark (66,67). Finland 1.2 0.8 1.1 1.7 1.5 1.8 No increasing trend was noticed in Iceland 3 - - - 3.2 4.7 Finland, Spain, New Zealand, Australia, or Norway 3.3 3.6 4.4 4.4 5.9 6.6 Czechoslovakia (67). England and Wales. The data from England and Wales are based on the United Kingdom national register that includes the whole England and Wales population. Analysis of the data indicated a Brimingham region 2.1 2.1 2.7 2.9 3.1 - steady increase in the prevalence of Birmingham and West Midlands - - - - - 3.5 hypospadias from 7.3 per 10,000 births in Liverpool region Southern Metropolitan region 2.4 2.2 24 - - 1964 to approximately 16 per 10,000 Southwest region 2.1 2.4 2.6 - 3.8 5 births in the early 1980s, when the number Northwest - - - 3.2 3.3 4 of cases stabilized (Figure 6). In 1990 the Urban - - - 3.3 - - prevalence of hypospadias showed a Rural -- - 2.9 - - decrease to 11.7 per 0,000. Sheffield region - 2.1 2.4 - - - Hungary. The Hungarian data are Oxford -2.5 2.6 3.7 3.9 4 based on the national register of the whole South Thames - - - 3.4 3.5 3.5 Trent - - - 3 3.4 3.8 population. As shown in Figure 6, there Mersey - - - 2.3 4 - was a rapid increase in the prevalence of Yorkshire - - - - - 4.1 hypospadias from 5.5 per 10,000 births in Scotland - 2.5 - - 4.4 5 1964 to 23.9 per 10,000 in 1978. Since East Scotland - - - 3.6 3.8 5.3 then, the prevalence of hypospadias, North Scotland - - - 4.3 4.2 4.4 although fluctuating, has remained at Northeast Scotland - - - 4 5.5 5.3 approximately the same level. Southeast Scotland - - - 4.3 5.3 5.1 Scandinavian Countries. The data West Scotland - - - 2.9 3.9 4.8 on the incidence of hypospadias in Scandinavia (Figure 7) are all based on the Western Europe national registers that include the whole France populations. The Danish data from 1970 Bas-Rhin - - - 3.5 4.3 5.9 to 1981 indicated a significant increase for Urban - - - 4.2 - - this period (from approximately 7.5 to 12 Rural - - - 2.6 - - per 10,000 births) (66). A further increase Doubs - - - 2.3 3.4 3.8 Urban - - - 2 3.3 - was noticed during the period 1982 to Rural - - - - 3.8 - 1988 (67). However, this increase may be Germany difficult to interpret as a new registration Saarland - - 3 3.4 5.4 6.2 system was introduced. Nevertheless, this Urban - - - - 5.5 - increase was similar to that reported for the Rural - - - - 5.4 - years 1970 to 1981. Hamburg 2.6 4 4.7 5.4 5.7 - The data from Sweden also indicated a Italy marked increase in the early 1970s: preva- Varese - - - 2.3 3.3 4 lence of hypospadias at birth was 40% Parma Ragusa - - - - 0.6 12.4 higher between 1974 and 1982 than for the period 1965 to 1968 (66,69). Netherlands However, the data obtained in the earlier 3 provinces 2.8 - - - - - Eindhoven 2 - - - 3.2 3 period could be more complete, because Spain they included both hospital records and Navarra - 1.3 1.2 1.4 registry data. Zaragoza - - 1.1 1.7 1.3 1.5 Also in Norway the prevalence of Urban - - - 1.6 - - hypospadias at birth increased from 7 to 8 Rural - - - 2 - - per 10,000 births between 1967 and 1971 Switzerland to 13 per 10,000 in 1973 (68). In 1988, the Basel - - - - 8.3 8.4 prevalence was 20.7 per 10,000 births (67). Geneva - - 4.4 5.1 6.3 6.2 Ethnic Differences. The incidence of Neuchatel - - - 9.1 5.4 7 in St. Gall hypospadias the United States is higher Vaud - - - 9.9 7.7 8.5 in Caucasians than in African Americans Urban - - - 8.9 9.9 - (ratio of 1.3-3.9:1) (70-73). In British Rural - - - 10.5 6.5 - Columbia, Canada, Native Americans were (continued)

746 Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

Table 1. (continued) reported to have a lower incidence of hypospadias than the general Caucasian 1960-1963 1964-1968 1969-1972 1973-1977 1978-1982 1983-1987 population. with a ratio of 1:6.7 (74,75). Zurich - - - - 7.4 8.8 Geographic Variation. Considerable Urban - - - 1.9 - - variation exists in the prevalence of Rural - - - 1.4 hypospadias among different countries. Interestingly, some populations with a low Eastern Europe - - - - - incidence of testicular cancer (e.g., Finland) Czechoslovakia - - - 2.4 2.7 3.8 (22) have a very low prevalence ofhypospa- Urban - - - 3.6 4.8 - dias (Figure 7). Furthermore, there seems Rural - - - 2.2 2.4 - to be considerable geographic variation in East Germany (GDR) - 2.6 3.3 4 - 7 the prevalence of hypospadias within differ- Hungary ent countries (66,76). Szabolcs - 1 1.8 1.5 1.2 2.7 Urban - - - - 1.7 - Incidence of Male Breast Cancer Rural - - - - 1.1 - Vas - 1.2 1 3.4 3.3 6.1 As xenoestrogens were implicated as possi- Urban - 4.1 4.6 4.7 6.6 - ble factors involved in the pathogenesis of Rural - 3.3 4.3 4.2 5.5 breast cancer in women (77), the trends in Poland the incidence of this disease in males Cieszyn - - - 2 - should be monitored. Male breast cancer is Urban - - - 2.8 a rare disease; only a few studies exist on Rural - - - 1.5 - - geographical and temporal trends. Ewertz Cracow city - 1 1.4 1.5 2.4 2.6 et al. (78) studied the incidence of male Katowice - 0.9 1.2 1.1 - - Urban - - - 1.1 breast cancer in four Nordic countries and Rural - - - 0.9 - - found a weak increase with calendar time Nowy Sacz - - - 1.1 1.2 1.2 in Denmark (1943-1982) but no change Warsaw city - 0.8 1.7 2.4 2 3.3 in Finland, Norway, and Sweden over the Warsaw rural - - 0.6 0.9 - 1.9 period 1953 (1958 for Sweden) to 1982. Romania Ewertz et al. noted a remarkable geographi- County Cluj - - - 1.3 1.3 1.4 cal trend, with Denmark having an inci- Urban - - - - 1.7 - dence about twice that of Finland. Yugoslavia Slovenia 1.3 1.6 1.9 2.1 2.8 3.4 Summnary Semen quality has deteriorated in many Canada - - - - 3 3.6 countries during the last 50 years. The inci- Alberta 3 3 3 3.8 3.6 3.7 dence of testicular cancer has been increas- British Columbia - - 3.1 3.3 3 4.2 ing almost invariably worldwide. Increases Manitoba 3 2.7 3.1 3.4 4 4 in the incidences of cryptorchidism and Maritime Provinces - - 1.9 2.3 2.8 3 hypospadias have been observed in countries in which longitudinal studies have New Brunswick 1.6 1.4 - - 2.7 2.6 been performed. However, there are clear Nova Scotia - - - - 3.2 3.2 regional differences. The prevalence of male Newfoundland 1.5 1.7 1.7 2.1 2.6 2.3 breast cancer has been rising and is higher Ontario - - - 2.7 3.4 4.2 in Denmark than in Finland. Quebec - 1.1 1.4 1.6 2 2.4 Northwest Territory and Yukon - - - 5.5 - 4.8 Changes in Male Saskatchewan 2.2 3.3 3.5 3 3.3 4.5 Reproduction in Wildlife. Estrogenic Effects on United States Developing Animals SEER, White - - - - - 4.9 Changes in male reproduction in wildlife SEER, Black 0.7 involve such issues as feminization, demas- Connecticut 2.4 - 2.8 3.6 - - culinization, reduced fertility, reduced White Black - - - - 4.2 0.6 hatchability, reduced viability of offspring, hormone secretion or and California impaired activity, Alameda altered sexual behavior (79). Since it is not White - 3.1 4.4 3.9 5.4 5.9 possible to review all of the data in detail, Black - 0 0.5 1.2 0.9 0.7 the reader is referred to a recent workshop Bay Area publication titled "Chemically-induced White - - 4.5 4 4.9 5.3 Alterations in Sexual and Functional Black 1 1 1.1 0.8 Development: The Wildlife/Human (continued)

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 747 TOPPARI ETAL.

Table 1. (continued) Connection" (80). Many of the reproductive disorders listed above have been associ- 1960-1963 1964-1968 1969-1972 1973-1977 1978-1982 1983-1987 ated with xenoestrogenic effects on the Chinese - - 3.3 2.2 1.3 - fetus. It has usually not been possible to Japanese - - - 3 0.6 - Los Angeles ascribe lowered reproductive success and White - - - 3.6 4.7 5.8 signs of feminization and demasculiniza- Spanish - - - 3.1 2.5 3.3 tion in wildlife to a single agent; in these Black - - - 0.9 0.7 0.6 cases, chemical analyses of specimens have Japanese - - - 1.7 1.9 2.4 revealed the presence of multiple com- Chinese - - - 0.4 0.6 1.9 pounds, some of which are known to have New York State 2.1 - 2.3 3.2 - - hormonelike activity. There is experimen- New York City - - - - 2.5 2.9 tal evidence that xenoestrogens act cumula- New York State less city - - - - 4.1 4.6 tively, i.e., 10 compounds administered Georgia simultaneously, each one at 1/10 of their Atlanta effective dose, resulted in a potent estro- White - - - 4 3.8 - Black - - - 1 0.5 - genic response (81). Cumulative exposure to nongonadal estrogens is easily docu- Iowa - - 3.7 3.6 4.1 4.4 mented in male birds, amphibians, and fish Louisiana New Orleans by measuring plasma vitellogenin levels. White - - - 3.4 4 3.7 Black - - - 0.6 1.3 0.9 Gastropod Species Michigan Pseudohermaphroditism or imposex- Detroit females developing male characteristics-in White - - 3.2 3.5 4.1 5 marine gastropod species has been reported Black - - 0.9 0.8 1.1 0.9 worldwide: the northeastern United States New Mexico (82), the United Kingdom (83), Alaska Hispanic - - 2.8 2.9 3.1 - (84), and Southeast Asia (85). This phe- Other White - - 4.1 4.2 4.2 nomenon is caused by tributyltin (TBT) American Indian - - 1 1.8 2.1 compounds leached from marine antifoul- Texas ing paints used on ships, boats, and El Paso mariculture pen nets (83). Very low con- Hispanic - 2.8 2.9 - - - centrations (1 ng/liter) of TBT-derived tin Utah - - 3.1 4.7 4 4.4 are effective in induction of imposex (83), Washington and marine areas averaging 6 to 8 ng/liter Seattle - - - 4.3 4.7 5.7 of TBT suffer reproductive failure and local extinction due to female sterility Latin America (86). The use of TBT was restricted in the Brazil 1980s, and recovery of some species has Porto Alegre - - - - 2.5 4.1 been reported after that (87). However, Recife - - 0.8 - 0.4 - imposex still remains a problem for gastro- Sao Paolo - - 1.1 1.4 2 - pod species in several marine areas. The Columbia mechanism of action of TBT is unknown. Cali 2.2 2 1.1 1.5 1.5 1.6 The effect mimics that of an antiestrogen, Cuba - - 0.2 0.3 - 0.6 and therefore presents an example of the Jamaica drastic hormonelike effect of a xenobiotic Kingston 0.8 - 0.1 0.6 - - threatening the existence of several species, Netherlands Antilles - - - 1.1 0.5 - even though higher species seem to be far Costa Rica - - - - 2.1 1.6 less sensitive to TBT. Puerto Rico 0.3 0.7 - 0.5 1 1.2 Reptiles Asia Lake Apopka, Florida, was extensively pol- China luted with dicofol (86), 1,1, l-trichloro- Shanghai - - - 0.9 0.7 0.8 2,2-bis(p-chlorophenyl)ethane (DDT) Tianjin - - - - 0.6 0o5 (and its metabolites 1,1-dichloro-2,2-bis(p- Hong Kong - - - 1.4 1.2 1.1 chlorophenyl)ethane [DDD], 1,1-dichloro- India 2,2-bis(p-chlorophenyl)ethylene [DDE] Bombay - 0.9 0.7 0.9 1 1 and chloro-DDT), and sulfuric acid spilled Bangalore - - - - 0.7 0.7 accidentally from a chemical company in Madras - - - - 0.7 0.6 1980. Further contamination of the lake Poona - -- 1 1 - by agricultural sewage dumping has made (continued) this lake one of Florida's most polluted

748 Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

Table 1. (continued) plasma testosterone concentrations of male juvenile alligators in Lake Apopka were 1960-1963 1964-1968 1969-1972 1973-1977 1978-1982 1983-1987 significantly lower than those of Lake Israel Woodruff males, equaling those of females. All Jews - 1.5 1.6 1.9 2.2 2.2 The plasma estrogen concentrations of Born in Israel - 2.2 1.3 2.1 1.8 2.7 Born Europe/America - 2 2.6 2.7 3.8 3.4 male alligators from these two lakes did not Born Africa/Asia - 0.8 0.8 1 4.9 0.8 differ. However, testes from Lake Apopka Non-Jews - 0.8 0.2 0.8 0.7 1.1 alligators produced significantly more Total population 1.6 - - - - - estradiol in vitro than testes from control Japan alligators; testosterone production was sim- Hiroshima - - - - 1.7 - ilar (93). The discordance in the in vitro Miyagi 0.6 0.4 0.8 0.8 1 1.3 and in vivo findings suggests additional Urban - - - 1 1 - significant differences in steroid metabo- Rural - - - 0.5 0.9 lism and liver function between the alliga- Okayama - 1.2 1.2 - - - Nagasaki - - - 1.1 0.8 1.9 tors from these two lakes, emphasizing the Osaka - - 0.7 0.8 1.4 1.4 complexity of environmental influences. Urban - - - 0.8 - - Reptiles have temperature-sensitive sex Rural - - - 0.5 - determination that can be altered by estro- Kuwait gen treatment (94). This has been demon- Kuwaitis - - - - 1.1 0.8 strated both in alligators (95) and turtles Non-Kuwaitis - - - - 1.3 0.6 (96). The data from Lake Apopka sug- Singapore gested that the gonads of alligators had Indian - - 0.5 0.7 0.6 1.2 been permanently modified, altering Malay - - 0.3 1.2 0.4 1.3 Chinese 0.5 - 0.9 steroidogenesis and inhibiting normal sex- 0.8 0.9 0.9 ual maturation (91,97,98). A recent study (99) has demonstrated Pacific Region that a number of PCB metabolites are Australia capable of acting as synthetic estrogens. As Capital Territory - - - - 4.1 4.2 New South Wales - - - 3 3.5 3.7 with crocodilians, many turtles exhibit Urban - - - 2.9 3.7 - environmental sex determination so that Rural - - - 3.1 2.9 - the temperature at which the egg is incu- South - - - 3.2 3.4 3.4 bated determines the sex of the offspring Tasmania - - - - 3.8 5 (97). Turtle eggs incubated at 26'C pro- Victoria - - - - 4.2 4.2 duce 100% males. However, if eggs incu- Western - - - - 4 3.6 bated at a male-producing temperature are New Zealand 3.5 - - - - - "painted" with either one of two PCB Pacific Polynesian Islands - - - - 1.8 - metabolites (2',4',6'-trichloro-4-bipheny- Maori - 3.4 4.3 4.3 7.9 6 lol or 2',3',4',5'-tetrachloro-4-biphenylol) Non-Maori - - 3.7 4.5 5.3 5.3 sex reversal occurs as if the eggs were European - 3.2 - - - - treated with the natural estrogen, 170- Hawaii estradiol (99). Interestingly, if eggs are All groups 2.2 - - - - - Caucasian 3.1 2.9 3.3 5 4.2 5.4 treated with a dose of both compounds Japanese 1.4 1.2 0.3 1.1 1.4 2.9 (100 pg of each)-a dose that produced a Hawaiian ethnic 3.3 2.9 3 3.9 3.3 3.6 small percentage of sex-reversed turtles (20 Filippino - 0.2 0.3 0.2 0.3 - and 0%, respectively)-they act synergisti- Chinese - 0.7 2.2 0.5 1.2 7.1 cally, producing sex reversal in 80% of the "Based on data from IARC (318-323). eggs treated. Thus, animals that should have developed as males are modified so that their internal and external morphology is that of a normal female. It is wetlands. A number of these pollutants are (91). There are extensive data supporting unknown if this sex reversal produces known to have estrogenic or endocrine-dis- the hypothesis (92,93): female alligators fertile adult females. rupting effects (89). The chemical spill was from Lake Apopka had plasma estradiol followed in the subsequent 3 years by a concentrations 2 times that of normal Fish significant decline in the number of females from the control lake, Lake Modifications in reproductive functioning juvenile alligators, whereas alligator popu- Woodruff. The females exhibited abnormal of male fish living downstream from kraft lations elsewhere were increasing or stable ovarian morphology. Likewise, in males pulp mills have been well documented. For at the same time (90). The population abnormal germ cells were observed in the example, white sucker (Catostomis commer- decline was associated with reproductive testes. Furthermore, Lake Apopka male alli- soni) collected from sites receiving primary disorders that were hypothesized to be gators had abnormally small phalli. Basal effluent exhibited delayed maturation, caused by endocrine-disrupting xenobiotics and luteinizing hormone (LH)-stimulated smaller gonads, and an absence of secondary

Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 749 TOPPARI ET AL.

25- steroidogenesis is disrupted by dietary expo- o England and Wales sure to PCBs (106). Additionally, Freeman * Hungary A et al. (107) observed that cod exposed to 20+- - - - PCBs in vivo exhibited increased metabolism of steroid hormones in vitro by kidney and V) c- liver tissues. Further, dietary exposure of If -M - - - juvenile males to PCBs precluded the rise in C=) 15±- CD CD plasma testosterone concentrations associ- C=. ated with sexual maturity. These data suggest a) - - - that PCBs modify both testicular androgen a) 10-- co synthesis and steroid utilization/degradation cx: in peripheral tissues. Exposure to crude oil also induced a decrease in plasma testos- 5- terone concentrations in the winter flounder, Pseudopleuronectus americanus (108). Organochlorines have also been implicated in a number of developmental and 1960 1970 1980 reproductive abnormalities in fish living in Years the Laurentian Great Lakes of North America. Male coho salmon (Oncorhynchus Figure 6. Prevalence of hypospadias at birth (rate per 10,000 births) in England and Wales, and Hungary. Based on kisutch) living in Lake Erie exhibit a num- data from WHO (67). ber of abnormalities, including decreased fertility, lower plasma concentrations of gonadotropins and steroids (testosterone, 25 A Sweden ------11-ketotestosterone), poor expression of O Norway secondary sex characteristics, and high pre- v Denmarknd cocious sexual maturation (109,110). It is unknown at this time if the above deficiencies are all related or represent different effects. It is hypothesized that some may be due to modifications of the developing embryo (organizational effects), whereas others may be due to disruptive activa- tional events in adults. 15------The observation of an increased prevalence of hermaphroditism in fish in sewage treatment water (STW) lagoons in England and Wales initiated a series of studies examining environmental estrogens using a bioassay involving vitellogenin synthesis in 1960 1970 1980 STW-exposed rainbow trout (Oncorhynchus mykiss) (111,112). Vitellogenin is produced Years in the liver under estrogen control by female fish for the growth of ova (113). Figure 7. Prevalence of hypospadias at birth (rate per 10,000 births) in the Scandinavian countries. Based on data Males produce it only after exogenous from WHO (67). estrogen treatment (114). Vitellogenin production by male fish can therefore be sex characteristics (100). Furthermore, these in Florida rivers downstream from paper used as a biomarker of environmental males had significantly reduced serum and pulp mills (102). Laboratory studies estrogenic activity. STW exposure of caged testosterone concentrations and dysfunc- corroborated the hypothesis that kraft mill rainbow trout induced increases of 500- to tional hypothalamic control of the pitu- effluents containing large amounts of the 100,000-fold in plasma vitellogenin con- itary-gonadal axis. A similar reduction in plant steroids sitosterol and stigmastanol centration, and males were shown to gonad size was reported for both male and caused the changes (103). However, the achieve vitellogenin levels almost as high as female perch (Perca fluviatilis) inhabiting etiologic agents in paper-mill effluents have females, indicating the contamination of sites receiving primary effluent from a kraft not been identified unequivocally. Other water by estrogenic compounds (112). In mill on the coast of the Gulf of Bothnia effects of paper-mill effluents on fish have vitro studies with trout hepatocytes estab- (101). The effects observed in these perch been reviewed by Owens (104) and lished a dose-response relationship with were graded along the pollution gradient Leatherland (105). estrogen exposure and vitellogenesis (111). (2-8 km) from the mill. Masculinization The effect of PCB exposure on testicular Using this culture system, it was demon- and behavioral changes of female mos- function in fish also has been examined. In strated that degradation products of several quito fish (Gambusia affinis) were observed the Atlantic cod (Gadus morhua), testicular alkylphenol-polyethoxylates, a major

750 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

group of surfactants present in sewage, are with morphological abnormalities (123). function of the Leydig cells that secrete estrogenic to fish (111). Their effect on Cryptorchidism (both uni- and bilateral) testosterone to promote differentiation of trout hepatocytes could be blocked with an has increased exponentially in male cubs the embryonic Wolffian ducts into the antiestrogen, Tamoxifen, demonstrating since 1975 so that today > 90% of the male male accessory sex organs, epididymides, that the compounds act through the estro- population exhibits this phenomenon seminal vesicles, and vasa deferentia. gen receptor. Although the estrogenic (eight of nine cubs born since 1985). Male Masculinization of the external genitalia activity of individual xenobiotics was low, sterility may be a problem as well. Female and the rest of the body, except the brain, is their effects may be additive in nature. panthers (n= 3) have high body burdens of also controlled by androgens and occurs various contaminants including p,p'-DDE after conversion of testosterone from the Birds (5.45-57.65 mg/g lipid fresh weight), testis into 5ax-dihydrotestosterone in the tar- Feminization of gulls and terns in several PCBs (7.32-27.06 mg/g lipid fresh get tissue. Female reproductive organs locations along the Pacific coast of the weight), oxychlordane (< 0.0098-2.00 develop in the absence of SRYand thereby United States has been associated with mg/g lipid fresh weight) and trans- in the absence of the testis (128). The DDT and DDE pollution (115,116). Fry nonachlor (< 0.0098-4.82 mg/g lipid fresh female developmental pattern seems to be a and Toone (115) demonstrated the femi- weight) (122). Panthers also have elevated genetic default pathway, and it is largely nizing capacity of some DDT compounds tissue levels of mercury, methoxychlor, and independent of hormonal regulation. by injecting gull eggs. Feminization leads other lipophilic organochlorine com- Disturbances in sexual differentiation occur to a skewed male/female ratio, which is pounds. These contaminants are derived when factors in the male developmental cas- known to increase female-female pairing. primarily from their major food item, the cade go wrong or when the genetic female is Supernormal clutches, i.e., five to six gull raccoon (124). The reproductive abnor- exposed to an elevated plasma-androgen eggs per nest instead of the normal three, malities described above were suggested to concentration. In the first instance, a genetic are often laid after this type of pairing, and be due to the contamination of mothers by male will develop a female phenotype and in the fertility of the eggs is poor (117). endocrine-disrupting environmental xeno- the latter case a female will be virilized. There is some controversy as to whether biotics rather than to problems associated feminization of males or differential mor- with inbreeding (122). Disorders ofGenital Development and tality of males resulted in a skewed sex Testicular Malignancy ratio (118). Fox (86) considered both the Summary The association between disorders of genital feminization of males and their increased Reproductive disorders in gastropod species, development and sexual differentiation and mortality, when compared to females, as reptiles, fish, birds, and mammals have been gonadal malignancy has been observed since possible reasons for the female-female pair- associated with environmental factors. the beginning of this century and is now ing and supranormal clutches in the gulls. Several of the disorders, such as sex reversal well established (129-131). The most fre- in reptiles and vitellogenin production by quent abnormality leading to neoplasia is Mammals male fish, may result from estrogenic action gonadal dysgenesis with the presence of Y Although the mechanism is unknown at ofchemicals in the environment. Fewer data chromosomal material; other disorders this time, elevated PCBs and DDE concen- are available concerning the mammals. include true hermaphroditism and andro- trations are associated with a decrease in However, some endangered species such as gen-insensitivity syndrome. Although the plasma testosterone concentrations in Florida panthers that are exposed to estro- general prevalence of disorders of sex differ- Dall's porpoises, Phocoenoides dalli (119). genic and/or other endocrine-disrupting entiation is low, the high incidence of germ Testicular steroidogenesis in vitro has been contaminants show reproductive disorders cell tumors makes the intersex gonad a good studied in the gray seal (Halichoerus grypus) comparable to those found in the human. model for the study of factors involved in in association with exposure to methyl the pathogenesis of germ-cell neoplasia. mercury (MeHg), cadmium (Cd), arsenic Sexual Differentiation and the Malignant growth frequently appears in the (As), selenium (Se), and a PCB mixture Physiological Role of intersex gonad in early childhood, thus sug- (Aroclor 1254). All contaminants except gesting that the carcinogenic process begins As and Se, stimulated testosterone synthe- Estrogens in utero. The intersex syndrome comprises a sis in vitro from seal testicular tissue (120). Seal Differentiation variety of genetic disorders, as different The mechanism by which this stimulation Sexual differentiation occurs during the from each other as, for example, XY/XO occurs is unknown. Gonadal steroidogene- first trimester of human pregnancy (125). mosaicism and a mutation in the androgen sis is not the only target for PCBs, as a An indifferent gonad develops into a testis receptor gene. The phenomenon of hetero- number of PCB metabolites have been under the influence of the SRYgene on the geneous genetic defects leading to a com- shown to decrease thyroid function in vivo Y chromosome. In addition to SRY, there mon result, malignancy of germ cells, in the common seal, Phoca vitulina (121). are several downstream effectors and auto- indicates that any disruption of early The majority of the remaining (approxi- somal genes (e.g., SOX9 and SF-1) that are gonadal development may render the germ mately 35 individuals) endangered Florida required for normal differentiation of the cells susceptible to neoplastic transforma- panthers (Felis concolor coryi) exhibit a testis (126,127). Sertoli cells in the newly tion by yet unknown mechanisms. There number of developmental abnormalities differentiated testis produce Miillerian are some hypotheses concerning possible and reproductive defects (122). Specifically, inhibiting substance (MIS), which induces mechanisms of neoplastic transformation males (n = 12) showed low ejaculate vol- regression of the Mullerian ducts that -e.g., arrested differentiation of immature ume, low sperm concentrations (3-15 x 106 would otherwise develop into the oviducts, germ cells (132)-or an imbalance in fetal sperm/ml semen), poor sperm motility, and uterus, and upper part of the vagina. Sertoli hormonal environment (133). Androgen- a very high proportion (92.9%) of sperm cells also regulate development and early insensitivity syndrome provides a clue that

Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 751 TOPPARI ET AL.

the lack of the appropriate inductive hor- with a defective estrogen receptor (147) length was significantly prolonged, result- monal environment may arrest fetal gonadal have begun to clarify the possible impor- ing in loss of litters due to difficulties in differentiation and lead to neoplasia later in tance of endogenous estrogens in sexual parturition. No major fertility problems life. It is possible that high levels oftesticular development. Lubahn et al. (146) dis- were reported in the male transgenics. The androgens have a protective function during rupted the estrogen receptor gene by tar- only abnormalities described in the males gonadal development; for example, a relative geted deletion that resulted in complete (as well as in the females) were hernias of excess of maternal testosterone during early estrogen resistance. Both male and female the abdominal wall musculature. This is of pregnancy was shown in black women com- mice survived to adulthood without appar- interest, since estrogen administration has pared to a matched white group, providing ent morphological anomalies. However, been reported to induce inguinal hernias in a possible explanation for the lower inci- females were infertile with hypoplastic male mice (149), and dogs (150,151), and dence of testicular cancer in black men uteri and hyperemic ovaries that contained there is some indication that inguinal her- (134). There is some experimental evidence no corpora lutea. Fertility of the males was nias are a risk factor for testicular cancer that androgens and estrogens may have also decreased. Only 3 of 15 males that (152). It is not known yet whether trans- opposite effects on certain pathways in the paired with normal females produced any genic male mice that overexpress the estro- developing gonad; e.g., free estrogen may offspring, and even those that were initially gen receptor will develop testicular tumors decrease expression of MIS (135,136), fertile lost their ability to sire subsequent later in life. whereas androgens seem to have a stimula- litters. Testicular weights were low and tory effect (137). sperm counts (in the testis and epididymis) Nonmammalian Vertebrates were only 10% of control. Weights of sem- Amphibians and birds differ from mam- The Physiological Role ofEstrogens in inal vesicles and coagulating gland were mals in their sexual differentiation (153). Sexual Differentiation normal. These findings suggest that estro- That is, the female phenotype in birds Estrogens act through a nuclear receptor that gens are necessary but not indispensable for develops under estrogen control, whereas is a ligand-activated transcription factor. In fetal sexual development; i.e., development the male phenotype appears in the absence addition, steroid hormones may effect the is overtly normal, but the sexual organs do of estrogen (154). Exogenous estrogen cell membrane. Estrogens are essential in the not reach their normal size and function. administration causes sex reversal in male development of female secondary sexual There is only one reported patient case birds and frogs. Some reptiles (crocodilians characteristics and in the female reproductive of estrogen resistance-a 28-year-old white and some turtles) have temperature-depen- cycle, fertility, and maintenance of preg- male (147). This man had incomplete epi- dent sex determination; for example, in nancy. In the developing male, the physio- physeal closure and therefore continued turtles, female hatchlings are produced by logical role(s) of estrogens is unclear, though linear growth into adulthood despite other- incubation of the eggs at a higher tempera- by analogy to the situation in adulthood, wise normal pubertal development. He was ture than males, but an excess of estrogen they probably play a role in regulating the normally masculinized and had normal causes feminine differentiation also at low differentiation and function of Leydig cells male genitalia with bilateral descended temperatures typical of males (155,156). (138). The role ofestrogen action in embry- testes (20 and 25 ml) and a normal-sized Estrogen-induced sex reversal can be used onic sexual differentiation is controversial. prostate gland. His sexual functions were as a biomarker of the estrogenicity of an In rats and rabbits (139,140), estrogen syn- normal including morning erections and environmental pollutant, as demonstrated thesis is activated in male and female nocturnal emissions. However, his semen recently for PCBs (99). Estrogen receptors embryos at the time of blastocyst implanta- quality was subnormal: sperm concentra- are expressed both in female and male chick tion in the uterus. Estrogen receptor mRNA tion was 25 million/ml and viability 18% embryos in the Miillerian ducts (154). can be detected in blastocysts and two-cell (normal values: > 20 million; > 50%, Interestingly, the left Miillerian duct that stage embryos (141). Immunohistochemical respectively). His serum testosterone con- develops into an oviduct in females studies by Greco et al. (142) demonstrated centration was normal, whereas estradiol, (regresses in male) has higher estrogen bind- estrogen receptor expression in both male estrone, follicle-stimulating hormone ing capacity than the right, which regresses and female mouse gonads on fetal day 13 (FSH), and LH concentrations were ele- in both sexes (157). Treatment of chick and 15. The gonads lose their estrogen vated. The elevated gonadotropin concen- embryos with diethylstilbestrol (DES) on receptor expression at later ages. These trations suggest that estrogens play a role in day 5 prevented regression of the Miullerian studies suggested a role for estrogens in the regulation of gonadotropin secretion in ducts in both sexes (158). Similar findings development ofthe gonads (143). A putative males, and thereby may have several indi- in the mouse are reviewed in the section molecular target of estrogens could be the rect effects. The patient case and the recep- "Effects of Synthetic Estrogens on the Testis MIS gene that contains a DNA sequence tor gene-deleted mice demonstrate that a in Animal Models." similar to the estrogen response element in normal male phenotype develops in the the upstream regulatory region (144). In absence of estrogen receptor-mediated Estrogens and Sperm contrast, the classical organ ablation studies influences, but semen quality and probably Production Capacity by Jost (145) demonstrated that gonadec- fertility may be compromised as a result. Sperm production is dependent on permis- tomy of the embryo always resulted in a sive actions of FSH and testosterone (and female phenotype. However, maternal and Overexpression ofthe therefore LH). Sperm production capacity placental estrogens might still have con- Estrogen Receptor depends on the number of Sertoli cells in tributed to this developmental pattern. In transgenic mice that overexpress the the seminiferous tubules (which is directly estrogen receptor, normal differentiation of related to the length of the tubules) (159), Defects in the Estrogen Receptor Gene sexual organs was observed (148). However, since each Sertoli cell supports a finite Recent reports on estrogen receptor gene- females in several transgenic lines had number ofgerm cells (160). Sertoli cells pro- deleted mice (146) and a male patient fertility problems and their gestational liferate quickly in rats from embryonic day

752 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

19 to day 15 after birth, then slow down and from the late 1940s to the early 1970s to offspring of mothers who took part in the cease multiplication approximately on post- prevent abortions and pregnancy complica- double-blind study of the effects of DES natal day 20 (161,162). Multiplication is tions (166). Dieckmann and co-workers on pregnancy in 1953 (167), and therefore largely dependent on FSH stimulation performed a double-blind placebo-con- the studies can be considered prospective. (162). In humans, regulation of Sertoli-cell trolled study on the therapeutic value of There were 308 men exposed to DES and proliferation may be very similar. Men DES during pregnancy in the early 1950s 307 men exposed to placebo included in with hypogonadotropic hypogonadism do (167). DES was given to 840 pregnant the study; 31.5% of men exposed to DES not develop normal-sized testes after women and placebo to 806 controls. had an abnormality of their reproductive gonadotropin treatment, which may be a Compliance was verified by a dye indicator tract, whereas only 7.8% of controls had consequence of inadequate Sertoli cell in the urine during the whole study. The an anomaly (174). In the recent follow-up multiplication in early childhood. This women entered the study between weeks 7 study of these males, it was found that the hypothesis is supported by findings in the and 20 of pregnancy (the majority during men who were exposed to DES before monkey (163): Sertoli cells proliferate in the weeks 10-12) and received increasing doses week 11 of gestation had twice as high a neonatal and infantile period but not during of DES until pregnancy week 35 (5-150 frequency of genital anomalies than did or after puberty. Estrogens suppress mg/day). This study clearly indicated that those who were exposed only later (175). gonadotropin production in animals at all the medication was not efficacious in the This finding indicates the importance of ages preceding puberty (164). It is hypothe- indications for which it was used (167). the timing of the exposure (time of organo- sized that this is the case also in humans. Instead, in the reanalysis of the material of genesis). In a small study comprising 17 Decreased gonadotropin stimulation during Dieckmann et al. (167), DES was associ- DES-exposed men, 12 nonexposed volun- the critical developmental phase may result ated with significant increases in abortions, teers, and 11 fertile control men, genital in inadequate Sertoli cell proliferation and neonatal deaths, and premature births anomalies (varicocele, epididymal cysts, small testes (165). Specific FSH gene dele- (168). When Herbst and co-workers absent testes) were reported in 13 of the tion experiments also demonstrated that (169,170) reported the high incidence of a DES-exposed subjects, 4 of the volunteers FSH regulates the size of the testis (TR very rare cancer, clear cell adenocarcinoma and 4 of the fertile normal controls (176). Kumaar, personal communication). At pre- of the vagina, in pubertal girls exposed to Whitehead and Leiter (177) reported geni- sent it is not known whether the number of DES in utero, the U.S. Food and Drug tal abnormalities in 29 of 48 men exposed Sertoli cells in human testes has decreased Administration (FDA) banned the use of to DES. Hypertrophy and squamous meta- and whether this might be a reason for DES during pregnancy. Medical authori- plasia of the prostatic utricle was found decreased sperm counts. ties in Europe that had allowed DES use more frequently in aborted male fetuses for pregnant women soon followed FDA that had been exposed to DES than in Summary regulations. In Europe, approximately nonexposed controls (178), suggesting that Normal masculine differentiation occurs 200,000 French, more than 150,000 DES-exposed males may have an increased under the influence of the SRYgene and Dutch, 63,000 Czechoslovakian, and 7000 risk of prostatic hyperplasia and/or cancer several other autosomal genes, and andro- British women were exposed to DES, when aging. The data connecting DES gens are required for this process. Disorders whereas in the United States 4.8 million exposure to several structural abnormalities of gonadal development are frequently asso- women were prescribed DES during preg- of the male reproductive tract are convinc- ciated with testicular germ cell neoplasia. nancy. In addition, DES was used as an ing and leave little space for speculation on Estrogens act through a specific nuclear anabolic agent in livestock, and the general confounding factors. However, no associa- receptor. Normal masculine differentiation population that used dairy products and tion was found between first-trimester occurs even in the absence of a functioning meat may have been exposed to the hor- exposure to sex hormones, other than estrogen receptor, but the patient with the mone via this route to an unknown, and DES, and external genital abnormalities in receptor defect had poor semen quality. probably variable, extent. Some of the a recent metaanalysis of 14 studies (179). Estrogen receptor-deficient male mice were DES-exposed daughters and sons have In a large cryptorchidism study, no associa- subfertile and few were able to sire one lit- been followed since the 1970s and a tion between the disorder and exposure to ter. Estrogens are involved in the feedback significant number of abnormalities in the estrogens during the pregnancy could be regulation of gonadotropin secretion, and structure and function of reproductive found (180). the suppression of FSH secretion during the organs have been described (171). period of Sertoli cell proliferation (perinatal Semen Quality period) may result in small testes and a low Structural Anomalies Gill et al. (181) studied semen samples sperm production capacity in adult life. Structural anomalies of the reproductive from 88 men exposed to DES and 85 men organs that are significantly more frequent exposed to placebo, who were offspring of Occurrence of Abnormalities in DES-exposed male subjects than in con- the mothers from the 1953 study performed in the Reproductive System trols include meatal stenosis (12.9 vs by Dieckmann and co-workers (167). 1.8%); hypospadias (4.4 vs 0%); epididy- Sperm concentration of men exposed to of the Sons of Women mal cysts (20.8 vs 4.9%); testicular abnor- DES was significantly lower than in the Exposed to Diethylstilbestrol malities, including hypoplastic testis, controls (83 million/ml vs 123 million/ml, during Pregnancy cryptorchidism, and capsular induration p < 0.02). There was no difference in semen Exposure (11.4 vs 2.9%); and microphallus (4 cases volume, whereas the total sperm count, vs 0 cases) (172-174). The data of Bibbo sperm motility grade, the total number of Diethylstilbestrol (DES) was prescribed to and Gill and their co-workers (173,174) motile sperm, the percentage of sperm with more than five million pregnant women are based on the follow-up studies of the normal morphology, and the quality score

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 753 TOPPARI ETAL.

were all statistically lower in men exposed studies have failed to show a statistically New York), 166 hospital controls, and 143 to DES. Azoospermia was found only in significant relationship between DES expo- neighborhood controls. There was no sta- men exposed to DES, and 20.5% com- sure and testicular cancer. There have been a tistically significant association between pared to 3.5% of men who received few case-control studies that evaluated pre- hormone treatment and cancer: 5.8% placebo had a sperm concentration in natal hormonal risk factors for testicular (n = 11) of cases had been exposed to DES semen of less than 20 million/ml. The cancer (186-191). In the first study (186), or other hormones, whereas 2.1% (n =3) groups did not differ in their testosterone, 131 testicular cancer patients, under age of the neighborhood controls and 2.5% FSH, or LH levels (173). In a later study 40, and their matched controls were ana- (n = 4) of the hospital controls had received on the same men (20 controls declined to lyzed. In 6 cases of cancer the mothers had exogenous hormones. Similarly, a participate), sperm concentrations still dif- been treated with hormones during preg- case-control study of 79 testicular cancer fered significantly, whereas other semen nancy, whereas only one mother of the patients from the Connecticut Tumor characteristics were similar between the control cases had received any hormones. Registry failed to show any increased groups (182). Similar results were obtained The difference was not statistically cancer risk in men exposed to DES (191). in another study (176) in which the mean significant, but if another factor, nausea, The studies above have been described sperm concentration of men exposed to was combined with hormone treatment, in detail because they illustrate two major DES was 66.4 million/ml compared to they formed a significant risk factor (rela- problems. First, DES treatment may have 101.7 million/ml in normal volunteers tive risk 4.33). been initiated at various times during preg- (p < 0.05). In this study, the zona-free In the case-control study of Depue et nancy; therefore, the presumed critical hamster egg penetration assay was also per- al. (187), 108 testicular cancer patients, period during which adverse effects of formed: sperm from 14 of 17 men exposed under age 30, were studied. Mothers of 9 estrogens might occur may have been to DES failed to penetrate more than 14% cancer patients had been treated with hor- missed in some of these studies. Second, of the eggs (which is the reference value for mones (2 with DES, 1 with estrogen, 1 the investigated populations of testicular the normal fertility range), suggesting with progestin, and 5 had pregnancy tests cancer patients have been too small to infertility, whereas only 2 of 12 unexposed consisting of a single injection of an estro- determine if a significant difference truly volunteers and none of 11 fertile normal gen-progestin preparation), whereas 2 con- exists between DES-exposed and nonex- controls had an abnormal test result. In the trols had either estrogen treatment or a posed men. When we combined the data study performed by Whitehead and Leiter pregnancy test. The relative risk (8.00) was presented above in a metaanalysis, a mar- (177), only 33% of the men exposed to significantly increased in the men exposed ginally significant increase in testicular DES had normal semen quality. However, to hormones (p = 0.02). However, the cancer incidence for the individuals Andonian and Kessler (183) found no dif- exposures were very heterogeneous, and exposed to hormones (including all hor- ference in semen quality between 24 men single pregnancy tests can hardly be com- mones) was found; Mantel-Haenszel esti- exposed to DES and 24 age-matched pared to long-term DES treatment. mates of the common odds ratio was 2.1 control men. Again, the large 1953 study In a similar study comprising 202 with 95% confidence intervals of 1.3 to population that has been followed prospec- cancer cases and 206 controls, Brown et al. 3.3. Exposure to DES was a significant risk tively appears the most valid for evaluation (189) found no excess risk associated with factor for testicular cancer on the basis of of semen quality. On the basis of that find- the use of hormones during pregnancy: our metaanalysis: odds ratio was 2.6 with ing, DES exposure resulted in a significant mothers of 4 cancer patients and 5 control 95% confidence limits of 1.1 to 6.1. It decrease in semen quality. mothers had received hormone treatment. would be most important to obtain addi- Semen quality and fertility are not in Two mothers in each group had been tional information on the incidence of tes- direct correlation. In the latest follow-up treated with DES, 1 control with estrogen, ticular cancer in men born to mothers who study of the Dieckmann cohort, no differ- 1 case with progesterone, 1 in each group participated in the double-blind, placebo- ence in the fertility between men exposed to had a hormone pregnancy test, and 1 con- controlled DES trial in the 1950s (167). DES and their controls were found (175). trol had an unidentified hormone treat- This is compatible with the earlier findings ment. However, it should be noted that 19 Summary (181) that the majority of the men exposed mothers in this study were medicated for Exposure to DES during pregnancy results to DES had sperm concentrations well bleeding problems, but only 2 (both case in an increased risk for several male repro- above the limit at which fertility is supposed mothers) mentioned a specific hormone ductive disorders, such as cryptorchidism, to be disturbed (20 million/ml), although used; 13 of the treated were case mothers urethral abnormalities, epididymal cysts, the mean sperm concentrations of exposed and 6 were control mothers. and testicular hypoplasia. In addition, the men were lower than those ofcontrols. In a case-control study of273 testicular semen quality of DES sons is worse than cancer patients from northern California that of controls. Incidence of testicular Testicular Cancer (190), no association was found with the cancer is approximately doubled among There is no conclusive evidence to indicate mother's hormone exposure or DES expo- DES sons compared to the general popula- an increased risk of testicular cancer in sure. Mothers of 9 cases and 10 controls tion, but whether this represents a true men exposed to DES (184), although the had been treated with hormones (odds increase of the cancer risk is equivocal. incidence of cryptorchidism is a well- ratio 0.9). Four of the case mothers and 2 known risk factor for testicular cancer and control mothers were exposed to DES. Effects of Synthetic has been observed more frequently in this The case-control study of Schottenfeld Estrogens on the Testis in group (171). Two patient cases with et al. (188) was based on questionnaires Animal Models seminoma in men exposed to DES have received from 190 testicular cancer patients Synthetic estrogens, such as DES, ethinyl been reported (185), but epidemiological (The Sloan Kettering Cancer Hospital, estradiol, and estradiol benzoate have been

754 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

thoroughly studied in several animal models These two studies are cited because they the MIS gene in the Sertoli cells. Some of because of their pharmaceutical applica- indicate that estrogens given neonatally act the structural abnormalities observed after tions. There are comprehensive reviews directly on both the testes and the pituitary birth may arise as a result of incomplete covering this topic (164,192,193). gland. It is noteworthy that the neonatal regression of the Miillerian ducts (198). period in rodents corresponds in many ways Mechanism ofAction to the second and third trimesters of preg- Estrogen Effects on the The effects of estrogens depend on the nancy in the human. Developing Testis dose, time, and probably the duration of Reports on various experimental animals exposure. Estrogens also act at several levels Adverse Effects of (e.g., sheep, rats, mice) describe how expo- in the reproductive system, i.e., they Prenatal Estrogen Treatment sure to exogenous estrogens during the influence specific neuronal areas in the Prenatal (day 11 and 12 postcoitum) expo- neonatal period causes drastic reductions in brain, they modulate gonadotropin secre- sure of mice from the Sv-Sl CP strain (a the secretion of FSH from the pituitary tion from the pituitary gland, and they strain in which the males are susceptible to gland and the presumption is that similar directly affect the reproductive organs. testicular teratomas) to ethinyl estradiol effects would occur before birth (199). As Estrogens probably exert most or all of resulted in an increased incidence of cryp- FSH plays a vital role in controlling multi- their effects through a specific receptor; torchidism (p= 0.0001), and 19 of 224 plication of Sertoli cells at this time (138) such receptors are present in the brain, exposed male animals developed testicular the prediction would be that estrogen- pituitary, gonads, and accessory sex organs teratoma compared to 4 out of 107 controls induced suppression of FSH levels would at one or another time during fetal, prepu- (the odds ratio of 2.4 was not significantly lead to a slower rate of Sertoli cell multipli- bertal, or adult life (143). However, the different) (197). cation. As the number of Sertoli cells precise localization and temporal expres- McLachlan and co-workers have performed in fetal/neonatal life is an important sion of estrogen receptors during differenti- formed a large series of studies on the factor influencing the maximum level of ation and development of the testis and effects of prenatal exposure of mice to DES sperm production in adult life, the conse- male reproductive tract are poorly (192,193). In most of the studies, preg- quences of such a change in terms of sperm described and further, more definitive, nant mice were treated with 0.01, 1, 10, or counts is obvious; moreover, such an effect studies are needed. The effects of DES are 100 pg/kg/day DES or corn oil on days 9 is irreversible once Sertoli cell multiplica- not unique to this compound but are prob- to 16 of pregnancy (time of sexual differen- tion stops in early postnatal life. There is ably shared by all estrogens (164). Many of tiation). The high doses are closely equiva- abundant evidence from man (hypogo- the synthetic estrogens are more effective lent to those used for pregnant women nadotropic hypogonadism) and from animal in lower doses than endogenous estradiol (173). Male offspring from these pregnan- species that suppression of FSH levels in because they are not bound by sex hor- cies suffered from the same structural and early postnatal life results in just such mone-binding globulin (SHBG), which functional anomalies reported in men changes [reviewed by Sharpe (138)]. Recent normally binds approximately 95% of cir- exposed to DES, i.e., epididymal cysts, cel- evidence from the fetal sheep (165) also culating estradiol, rendering it biologically lular atypia in the prostate, cryptorchidism, shows that suppression of FSH secretion in inactive. Estrogens are metabolized rapidly testicular hypoplasia, poor semen quality, the fetal male during the second half of in the testes, e.g., by specific sulfotrans- and subfertility (192). In addition, Sertoli- gestation results at birth in testes that con- ferases, after which they cannot bind to cell hyperplasia, interstitial testicular tain 40% fewer Sertoli cells than occurs in their receptor (194). If the active center of tumors, squamous metaplasia of the semi- control animals. the enzyme is occupied by a xenobiotic, nal vesicles, and rete testis adenocarcinoma It is therefore hypothesized that pro- metabolism of endogenous estrogens may were found frequently in the male offspring longed exposure of the developing male, be disturbed and high levels of active hor- of mice exposed to DES during pregnancy during both fetal and postnatal life, to mone may be available. This is particularly (198). The analogy between the findings exogenous estrogens (perhaps even at low important during fetal development when in the human and the mouse illustrates levels) could reduce Sertoli cell number the levels ofambient estrogens are high. how informative and relevant the results and thus reduce sperm output (and sperm from animal studies are. counts) in adult life. Experiments involving Adverse Effects of exposure of rats to various xenoestrogens Neonatal Estrogen Treatment Estrogen Effects on the during the period of Sertoli cell multiplica- Long-lasting suppression of spermatogene- Mullerian Ducts tion have shown that in adult life such sis and atrophy of reproductive organs in Miillerian inhibiting substance secreted by exposure results in small (8-12%) but neonatally estrogen-treated rats and mice Sertoli cells is responsible for regression of highly significant reductions in testis size have been described by many workers since the Mullerian ducts. Analysis of male and a corresponding decrease in daily the 1950s (164). A single injection of mouse embryos exposed to DES revealed sperm production (200). These effects estradiol benzoate (250 pg on day 5) delayed and incomplete regression of the have been achieved after exposure to rela- resulted in a marked delay of the onset of Mullerian ducts (192). In vitro organ cul- tively low levels of the chemicals (alkylphe- puberty in rats (195). When mice were ture experiments verified the inhibitory nols, phthalates; 1 mg/liter in drinking treated with repeated doses of estradiol on action of DES on Mullerian duct regres- water of pregnant rats) under test. For days 1 to 5 after birth, the testes were irre- sion (136). Estrogen receptors were found example, butylbenzyl phthalate has been versibly damaged, and subsequent treat- both in the Miillerian ducts and Sertoli found to occur in butter and margarine at ment with testosterone and gonadotropins cells at the time of regression (143). concentrations as high as 47.8 mg/kg (201). failed to maintain spermatogenesis in the Estrogens could either affect the Mullerian Such findings suggest that there is the theo- majority of these estrogenized mice (196). ducts directly or influence the expression of retical possibility that human exposure to

Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 755 TOPPARI ET AL.

such chemicals might have contributed to from the early 1980s in the European DDE, dicofol, perthane, methoxychlor), the decline in sperm counts in men Union (EU). The current legislation cyclodienes (chlordane, oxychlordane, trans- described earlier. demands extensive documentation for nonachlor, heptachlor, heptachlorepoxide, safety by regulatory agencies before a aldrin, and dieldrin), hexachlorobenzene, Summary chemical can be used in foods or commer- and hexachlorocyclohexanes (206). Many Diethylstilbestrol treatment of experimental cial products. However, many chemicals of these, most notably DDT, were used in animals in utero results in increased inci- were introduced before these strict regula- large quantities until the 1960s when the dence of cryptorchidism; urethral abnor- tions were enforced. Thus, the present situ- use of DDT was banned or restricted in malities; testicular hypoplasia; poor semen ation is that man and wildlife are exposed Western countries. Hexachlorobenzene, quality; and infertility, abnormalities in to a very wide range of chemicals. For the however, was used in the United States accessory sex organs, rete testis adenocarci- majority of these we do not know whether until 1985. DDT products are still used noma, interstitial cell hyperplasia, and they are estrogenic, whether their effects widely in many developing countries. tumors. Thus, the outcome of DES expo- are additive, or even what the true expo- Despite restrictions on their use, these sure ofexperimental animals is highly analo- sure to these chemicals is. compounds are still circulating in the envi- gous to the findings in humans. Recent data A xenoestrogen can induce its estro- ronment because many of them bioaccu- in the rat suggest that perinatal exposure to genic effect in multiple ways: it may act mulate and become concentrated in body xenoestrogens, such as butylbenzyl phtha- directly through estrogen receptors, or it lipids (biomagnify). The breakdown and late, results in decreased size of the testes may disturb estrogen metabolism, thus elimination of these compounds is very and daily sperm production in adult life. increasing the levels of the endogenously slow; therefore, their effects can be persis- produced ligand. Different estrogenic and tent, lasting for generations (DDT has a Environmental Chemicals antiestrogenic ligands form functionally half-life of > 60 years in the environment). with Known Estrogenic different complexes with the estrogen Long-term exposure to small amounts of Effects receptor, and their transcriptional effects organochlorine contaminants leads to the Estrogenic effects are not restricted to a depend on the cell type and promoter accumulation of considerable burdens in small group of therapeutic agents but (205). Thus, the same compound may animal and human tissues (207,208). It is appear in several groups of compounds potentially have an estrogenic effect in one therefore not the amount of DDT to which that are used daily in industry, agriculture, system or at one concentration, and an a mother is exposed during pregnancy that or in the home (79,80,202-204). The antiestrogenic effect in another system or at is critical but rather her lifetime exposure major groups of environmental chemicals, another concentration. Furthermore, effects that will determine the level of exposure of such as organochlorine pesticides, PCBs, of many compounds influencing other hor- the fetus and the breast-fed infant. dioxins, alkylphenol polyethoxylates, phy- mone systems (e.g., antiandrogens) may Commercial DDT contains several iso- toestrogens, and other xenoestrogens, cur- mimic those of estrogens. mers ofwhich p,p'-DDT is the most preva- rently known to have estrogenic effects in A number of chemicals, mainly pesti- lent (75-80%), whereas the proportion of vertebrates or in assays in vitro are dis- cides and many of which are currently the most estrogenic isomer o,p'-DDT is 10 cussed here. A major problem is the deter- being used, have been implicated as envi- to 25% (89,209). The o,p'-isomers are less mination of those chemicals that are ronmental hormones possessing endocrine- stable than the p,p' configurations and are estrogenic (or otherwise endocrine-disrupt- disrupting properties (80). In the public therefore found only in low concentrations ing, i.e., disturbing normal endocrine debate on male reproductive disorders, this in nature (210). However, p,p'-DDT was homeostasis). At present, tens of thousands has misled many to suppose that all of also reported to have estrogenic actions of man-made chemicals are being used, yet these chemicals are estrogenic. In fact, both on the rat uterus (211) and in the the effects on the endocrine system have many of these compounds have not been MCF-7 breast cancer cell line (212). The been studied for only a few of these. The adequately tested for estrogenic activity. estrogenic activity of DDT isomers estrogenic activity of the majority of chem- However, for many others, a large toxico- compared to that of estradiol is very weak icals (e.g., alkylphenols, phthalate esters, logical database exists, including data on (103_106 times less potent). However, the bisphenol-A) has been detected by acci- reproductive toxicity, effects on steroid- long half-life and bioaccumulative proper- dent, not by intent, although recently some metabolizing enzymes, and effects on hor- ties of DDT indicate that levels of human screening of chemicals used in large vol- mone-producing tissues. A short summary exposure may be sufficient to induce umes has been attempted (204). Hence, it of the most relevant toxicologic effects estrogenic effects in certain circumstances. is highly possible that other estrogenic known for a number of xenoestrogens and This is particularly true for the period chemicals remain unidentified. However, it other environmental chemicals that have from the 1940s to 1960s when DDT was should be remembered that many of the been implicated as environmental hormones used widely including in direct application chemicals to which man is exposed have is given in "Appendix A." It also outlines the to humans. been tested (often in two- or three-genera- safety assessment procedures and principles Antiandrogenic (demasculinizing) and tion studies) before being approved for use; applied world-wide by regulatory agencies. estrogenic (feminizing) effects often mani- and hence, if any of these chemicals were a Below is a short examination of each of fest themselves in the same way, although strong estrogen, this would probably have the groups of chemicals that are known to through distinct receptors (213). Therefore been discovered. This is especially the case be estrogenic. the recent discovery that p,p'-DDE, the for chemicals that are currently approved for main metabolite of DDT in the body, is a use in food production, such as food addi- Organochlorine Pesticides potent antiandrogen (214) may explain tives and pesticides, and for new chemicals Organochlorine pesticides include some of the estrogenic effects observed in that have been produced in large amounts dichlorodiphenylethanes (DDT, DDD, the environment; many of these effects

756 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

may occur due to an antiandrogen activity accounts for 90% of the total HCH found quite active as estrogenic compounds [i.e., ofa xenobiotic. in human milk (225). Lindane was reported more than 1/100 of estradiol activity Fry and Toone (115) induced femi- to have both estrogenic and antiestrogenic (235)]. Antiestrogenic effects have been nization in male California gulls (Larus effects in female rats (226). In male wean- found in MCF-7 breast cancer cells with californicus) by injecting eggs with DDT in ling rats fed with ,B-HCH (0, 2, 10, 50, or 3,3',4,4'-tetrachloro-biphenyl (a dioxinlike amounts that were comparable to those 250 mg/kg) for 13 weeks, liver enzyme PCB), a form known to bind to the Ah found in seabird eggs in southern California induction occurred at doses > 2 mg/kg; receptor, mediating the effect (233). in the late 1960s. A skewed sex ratio in favor testis weights decreased at doses > 50 Reproductive failure of seals in the Wadden of females in large gull populations mg/kg; and testicular atrophy resulted Sea has been attributed to PCBs (236), and suggested the possibility of a causal relation- from a dose of 250 mg/kg (227). has been supported by laboratory studies ship with the estrogenic action of DDT In hamsters, a single injection of the (121). However, this relationship may not (86). The effects of DDT metabolites and weakly estrogenic chlordecone (Kepone) in necessarily have been a consequence of the dicofol in reptiles have been discussed in the neonatal males reduced testicular and epi- estrogenicity ofthe PCBs. section "Changes in Male Reproduction in didymal weight (228). Estrogenicity of Dioxins and Furans. Polychlorinated Wildlife. Estrogenic Effects on Developing chlordecone was also demonstrated in rats dibenzo-p-dioxins (PCDDs) and polychlori- Animals." In mammals, the effects of DDT (229) and birds (230). nated dibenzofurans (PCDFs) consist of 75 compounds on male reproductive function and 135 different congeners, respectively are less apparent (215). Polychiorinated Biphenyls (237). The most toxic congener is 2,3,7,8- Methoxychlor is estrogenic in the E- Polychlorinated biphenyls are industrial tetrachlorodibenzo-p-dioxin (TCDD), com- SCREEN assay (204). It was also found to chemicals used since 1929 as heat transfer monly referred to as dioxin. These be estrogenic in vivo in rats (216). and hydraulic fluids, adhesives, flame retar- compounds are formed as unwanted by- Methoxychlor or DDT exposure of neona- dants, dielectric fluids for capacitors and products in the manufacture of chlorinated tal rats did not affect male reproductive transformers, and waxes (231). PCBs con- hydrocarbons. Other sources include incin- organ weights in adulthood (217), and sist of 209 congeners, which are found in eration processes, paper and pulp bleach- neither induced epididymal cysts (218), different mixtures in commercial products. ing, and emissions from steel foundries and which were found frequently in mice Before the production of PCBs was banned from motor vehicles (238). exposed to DES (192). However, exposure in the United States in 1977, hundreds of Most of the animal studies on dioxins throughout gestation and lactation in millions of kilograms were produced, and a have been performed with TCDD rodents resulted in slightly smaller testes and large proportion of the synthesized product [reviewed by Peterson et al. (239)]. epididymides and in lower sperm counts in is still in the environment because of bioac- Dioxins act through an Ah receptor that is male offspring than in controls (219,220). cumulation and slow biotransformation. also involved in mediating the antiestro- It was suggested that the inability of the The biological effects caused by the var- genic effects of TCDD (240,241). neonatal rodent to metabolize methoxychlor ious congeners differ, not only in potency However, the role of the Ah receptor has to its active estrogenic form might explain but also qualitatively. Several non-ortho- not been established for several of the toxic the discrepancy between these studies (220). and mono-ortho-substituted PCB con- effects that are found in males exposed to Chlorinated cyclodienes induce liver geners induce effects similar to those TCDD (242). There is considerable litera- enzymes that hydroxylate testosterone caused by chlorinated dioxins and dibenzo- ture documenting the toxic effects of diox- (221). Chlordane disturbed spermatogene- furans; i.e., the toxicity is probably medi- ins on the male reproductive system sis and caused dose-related damage to the ated through interaction with the aryl (239,242). Prenatal and lactational expo- testes of mice fed for 30 days with 0.08 mg hydrocarbon (Ah) receptor. Other PCB sure of male rats to TCDD profoundly dis- or 0.25 mg of the active ingredient (222). congeners presumably act by different turbed the developing male reproductive Mating studies of dieldrin-exposed rats sug- mechanisms. In addition, there are PCB organs: anogenital distance was shortened, gest male-dependent disturbances in fertility congeners that are intermediate in this testicular descent was delayed, and the (223). In the E-SCREEN assay, chlordane respect; i.e., they elicit a mixed spectrum of weights of all sex organs were reduced and heptachlor were not estrogenic, but the effects. Our knowledge of possible interac- (243). Furthermore, spermatogenesis was heptachlor derivative 1-hydroxy chlordane tions between the various groups of PCBs inhibited, sexual behavior was feminized was (212). In addition, dieldrin was shown is still very limited (232). and demasculinized, and the regulation of to be estrogenic (81). Both estrogenic and antiestrogenic LH secretion was feminized (244-246). Hexachlorobenzene was also reported effects have been reported for different Perinatal suppression of plasma testos- to induce liver enzymes hydroxylating PCB congeners (233). The estrogenic terone concentrations appeared to be androgens (221). Long-term studies have potency appears to depend on the percent- involved in the changes described. The demonstrated liver and kidney anomalies age of chlorine: less-chlorinated PCBs effects were elicited by a single maternal in exposed animals but indicate no effect (Aroclors 1221, 1232, 1242, and 1248) oral dose of TCDD on day 15 of on fertility (224). have estrogenic activity whereas more chlo- pregnancy [ED50 approximately 0.16 Hexachlorocyclohexanes (HCHs) com- rinated congeners do not (209). The stabil- pg/kg; at this dose, TCDD had no dis- prise several isomeric forms; these com- ity of the compounds increases with higher cernible effect on the mother (242)]. Most pounds are also called benzene hexachloride chlorination. Less-chlorinated compounds of the effects were found at the lowest dose (BHC). y-HCH has the common name lin- were shown to transfer more readily across level tested (0.064 pg/kg), whereas in other dane and is the most acutely toxic of the the placenta than were the highly chlori- studies, reproductive toxicity had been isomers (215). The most persistent and nated PCBs (234). PCBs are hydroxylated observed with doses of < 0.001 Ipg/kg/day bioaccumulating isomer is ,-HCH, which in animals, and these hydroxybiphenyls are (247,248). The mechanism of action of

Environmental Health Perspectives - Vol 104, Supplement 4 - August 1996 757 TOPPARI ET AL.

dioxins and furans may not be primarily Phytoestrogens have been shown to Exposure of Humans to estrogenic or antiestrogenic (e.g., anti- have estrogenic effects both in vitro and in Environmental Chemicals estrogenic effects of TCDD cannot be vivo (212,258,261). Feeding of rams with with Estrogenic Activity and reversed by high estrogen concentrations), clover that is rich in isoflavone resulted in Their Effects on Male but it is certainly a hormonelike effect. decreased sperm counts (262,263). The This necessitates the close surveillance of effects are receptor mediated and depending Reproductive Health humans exposed to TCDD. on the dose are either estrogenic or antie- The major routes of exposure to man-made strogenic in an adult animal (264,265). chemicals are thought to be Alkylphenol Polyethoxylates Because of these potential antiestrogenic * Dietary: pesticides, (including chemicals Alkylphenols and related compounds are effects, high doses of isoflavones in the diet used in the formulation of commercial present in surface waters and aquatic sedi- have been proposed as being beneficial for products), food additives (including ments (249-251). They are products of reducing the risk of hormone-dependent synthetic flavoring substances), contam- the microbial breakdown of alkylphenol cancers (266). However, little considera- inants (such as PCBs, dioxins, metals, polyethoxylates (APEs) that are widely tion has been given to neonatal and industrial chemicals, especially those used in industrial surfactants (250,251). childhood life when exposure to phytoe- that are biomagnified in food chains), These effective, nonionic surfactants are strogens would be presumed to have non- packaging and wrapping materials (e.g., used in detergents, paints, herbicides, pesti- beneficial effects. plastics, food wraps) cides, and cosmetics, to name a few major * Environmental: from the pollution of groups of products. Over 300 million kilo- Other Xenoestrogens air and water grams of APEs are produced annually. After Bisphenol-A, a plastic monomer that was * Domestic: from contact with house- sewage treatment, approximately 60% of the released from polycarbonate flasks during hold products, cosmetics, clothing, and APEs are released into the aquatic environ- autoclaving, was shown to have an estro- probably many others ment as short-chain APEs (e.g., nonylphe- genic effect on MCF-7 breast cancer cells * Occupational: inhalation, dermal con- noldiethoxylate [NP2EO]), alkylphenol (267). Notably, bisphenol-A is used exten- tact and ingestion of active compounds, carboxylic acids (e.g., nonylphenoxycar- sively as a plasticizer, e.g., it is used in the depending on the occupation. boxylic acid [NPlEC]), and alkylphenols lacquer coating of food cans that are then Occupational exposure has not been (e.g., nonylphenol [NP]; and octylphenol heated for sterilization purposes (268). considered in any detail in this report, [OP]) (252-254). Alkylphenols are rela- Other common chemicals used in the plas- which is aimed at describing the situation tively persistent and bioaccumulate in the tic industry include the phthalate esters, for the general population. However, valu- lipids of living organisms (255,256). butylbenzyl phthalate and di-n-butylphtha- able information on possible association NP and OP were shown to be estro- late, which were shown to act as weak between exposure to chemicals and effects genic both in vivo and in vitro in mam- estrogens on breast cancer cells (203). In on humans may originate from studies in malian systems (202,257). The effects the E-SCREEN assay, di-n-butylphthalate occupational settings where high exposures were estrogen receptor mediated. OP was was not estrogenic (204). Neither of the have taken place. Because of better docu- more potent than NP, reaching approxi- phthalate esters had antagonist effects, sug- mentation and higher exposure, such mately 1/1000 of the potency of estradiol. gesting that their action may be cumulative studies are more likely to reveal adverse In the trout hepatocyte assay, many other (203). Phthalates are the most abundant effects of chemicals on humans than are the APEs were also shown to be weakly man-made environmental pollutants; and studies on the general population (269). estrogenic (111). human intake per day by various routes, The diet is usually regarded as the most especially through the diet, is measured in important source of foreign chemicals. Phytoestrogens tens of milligrams (201). A food antioxi- Very preliminary estimates of some expo- Phytoestrogens are natural compounds pre- dant, butylated hydroxyanisole (BHA), was sures in Denmark are included in sent in plants and are ingested daily in mil- also estrogenic in breast cancer cell assays "Appendix A." It should not be overlooked ligram quantities. The active substances are (203). However, the estrogenic potency of that the exposure via routes other than isoflavones (genistein and daidzen) and BHA and phthalate esters was lower than food, such as air, drinking water, and par- coumestans (coumesterol) (258). Fungal that of octylphenol or o,p'-DDT. ticularly the skin (e.g., detergents) may be metabolites, such as zearalenone, found in highly significant. Current knowledge on foodstuff (e.g., grain) also belong to the Summary actual exposures is rather limited, in partic- phytoestrogens. Reproductive disorders in Numerous environmental chemicals, such ular with regard to the domestic exposure sheep are well documented after the eating as many organochlorine pesticides, PCBs, to chemical products. A close examination of red clover containing high amounts of alkylphenol polyethoxylates, phthalates, of all possible exposures to chemicals sus- genistein (259). Some of the food prod- and phytoestrogens are known to have pected to be environmental hormones has ucts rich in phytoestrogens include rye, estrogenic effects in vertebrates or in assays not been possible within the constraints of wheat, cabbage, sprouts, spinach, and soy- in vitro. However, only a few of the tens of this report. Humans may be exposed to bean. Soybean is by far the richest source thousands of man-made chemicals have xenoestrogens in multiple ways. Direct of plant estrogens and is used ubiquitously been tested for estrogenic or other administration of synthetic hormones, such in the food industry as a protein source, endocrine activity, and therefore, it is as DES or ethinyl estradiol is obvious, but including infant milk formula substitutes. highly possible that other estrogenic chem- the same hormones may be found in meat Phytoestrogens may not bioaccumulate or icals remain unidentified. A major problem and dairy products in some countries biomagnify but they are readily metabo- at present is how to fill this gap in our (270). Cow's milk has a high concentra- lized and excreted (260). knowledge rapidly and cost effectively. tion of endogenous estrogens during late

758 Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

gestation, and this milk is also collected for large population ate rice oil contaminated day, from certain food sources alone human consumption. The ratio of estro- with PCBs, PCDFs, and polychlorinated ter- (201). This level of intake would be gens in this milk to that in plasma is phenyls. Exposed pregnant women had likely to result in estrogenic effects based generally greater than one (271), and increased fetal loss and the surviving infants on data from in vitro (203) and in vivo therefore much higher than in human had low birth weight (280). The exposed (200) studies, though not all phthalates breast milk. Occupational exposure in the infants were delayed in their developmental may be estrogenic. pharmaceutical industry is a possibility for a milestones. The behavioral effects were prob- small minority. Estrogens occur in measur- ably mainly due to exposure in utero, since Phytoestrogens able amounts in sewage effluent used for the effects were also found in bottle-fed Large amounts (hundreds of milligrams) of irrigation (272). Phytoestrogens are ingested infants (232). The boys at pubertal age (11 phytoestrogens such as isoflavones can be in large amounts, and weakly estrogenic and 14 years of age; n = 23) exposed prena- ingested daily by humans, especially in a alkylphenolic compounds are applied to the tally to very high levels of contaminants vegetarian diet. Antiestrogenic action of skin and ingested daily. Important issues to had significantly smaller penis lengths isoflavones was demonstrated in women consider are the quantity of estrogenic com- when compared to controls, whereas 31 who consumed a soy protein-enriched diet pounds present, their potency, their capabil- boys at prepubertal age (8 and 10 years of containing 45 mg isoflavones daily for 1 ity to bioaccumulate and biomagnify, and age) did not differ from controls (281). month: the follicular phase was prolonged, their additive, synergistic, or antagonistic The plasma levels of PCBs in Yusho were and the midcycle surge of LH and FSH effects. Concentrations of organochlorine originally 0.01 to 0.1 ppm while body fat was suppressed (266,287). This may have contaminants in human reproductive tissue, levels ranged up to 76 ppm. In the normal occurred because of phytoestrogen- adipose tissue, and blood from the general population the corresponding figures are induced production of SHBG, which population (Table Al) and in human breast <0.005 ppm and 0.5 to 10 ppm (282). would reduce bioavailability of endoge- milk (Table A2) worldwide are included in The reproductive ability of the exposed nously produced estrogens. In post- "Appendix 1"[adapted from Thomas and males has yet to be ascertained. menopausal women, dietary soy induced Colburn (206)]. Dioxins and Furans. Humans are nor- no significant estrogenic or antiestrogenic mally exposed to dioxin and furan levels far effects (288). SHBG and gonadotropin Organochlorine Pesticides and below those that induce reproductive disor- levels remained unchanged in these Polychlorinated Biphenyls ders in animal experiments (206). After women, whereas the vaginal epithelium The daily intake of DDT is now small in the Seveso accident in Italy in 1976 where showed a tendency for a higher percentage Europe and North America, and it may a factory exploded during the production of superficial cells (indicative of estrogenic- not have a significant influence alone. of 2,4,5-trichlorophenol, TCDD serum ity). A study of postmenopausal women in Chlordecone (Kepone), which is estrogenic, levels on a lipid basis ranged from 1770 to Australia suggested an estrogenic influence caused an occupational risk to workers 56,660 ppt (ng/kg) in exposed children by phytoestrogens on the vaginal epithe- exposed to high levels of the compound: (283), while the background concentration lium (289). Many infants are fed with soy- exposed men had oligozoospermia, in human milk fat is approximately 2 ppt based milk-substitute formulas rich in decreased sperm motility and abnormal (225). No reproductive disorders in adults phytoestrogens. There are no data avail- sperm morphology (273). have been described after the accident able about the possible endocrine effects in Exposure to several organochlorine pes- (284,285), though the exposed male chil- children, but it is presumed that the ticides and PCBs at the same time may dren have not yet been evaluated. situation would be radically different from lead to untoward effects, as exemplified by the adult because of the negligible endoge- the suggested association between slight Recendy Found Xenoestrogens nous production of estrogens in infancy disturbances of development in children Measurement of human exposure to (especially in male infants). Female rats exposed in utero and contaminants present alkylphenols, phthalate esters and bisphe- exposed neonatally to phytoestrogens show in Lake Michigan fish (274-278). The risk nol-A is difficult because the data are still an increased incidence of premature was related to lifetime intake of contami- sparse. There is, however, considerable anovulatory syndrome in adult life. This nated fish by the mothers and to PCB levels concern because these compounds are so syndrome is recognized as a classical conse- in maternal serum and milk rather than to ubiquitous in the modern environment. quence of inappropriate exposure to intake during pregnancy. Reproductive dis- Some nonylphenol ethoxylates and one estrogens in neonatal life (290). Many orders have not been described, but the octylphenol ethoxylate were found at con- phytoestrogens, such as lignans and affected male children have not yet reached centrations of 15 to 29 ng/liter in drinking isoflavonoids, are metabolized and excreted reproductive age. These studies have been water in New Jersey (286). In different in urine similarly to endogenous estrogens extensively debated, and more investiga- fish species, concentrations of 0.13 to 3.1 (260). Thus, they may not bioaccumulate tions in populations with well-known high mg/kg dry weight of nonylphenol and its in the body. exposures are needed to resolve this issue. two ethoxylates have been found (256). Levels of some PCB congeners were Since these compounds have very weak Summary inversely correlated to sperm motility in estrogenicity, it is too early to estimate Humans are exposed to environmental semen samples in which sperm concentra- whether environmental exposure to them estrogens in multiple ways. Diet, drink- tion was < 20 million/ml (279). The small has any influence on humans, as the level of ing water, air, and the skin are the routes number of subjects in the study leaves the exposure of man is not known. However, through which xenoestrogens enter the significance of the finding open. Severe for phthalates it is established that human body. Several of the known xenoestrogens poisoning accidents occurred in Yusho, intake per day is likely to be several have a weak estrogenic activity but are Japan, and Yu-Cheng, Taiwan, where a hundreds of micrograms per kilogram per highly persistent and accumulate in fat.

Environmental Health Perspectives - Vol 104, Supplement 4 - August 1996 759 TOPPARI ET AL.

Moreover, many of them have additive treatment during pregnancy" will be the consequences for both logistics and inter- effects. Some of these compounds, such as only exposure data that can be obtained). pretation (e.g., male breast cancer cases phthalates, are present in high concentra- In this situation the so-called ecological may have to be related to pregnancies tions in certain food items, whereas approach may be necessary. Rather than 50-70 years ago). concentrations of pesticides and PCBs are relating adverse outcomes to exposure for Epidemiological studies are needed to very low. The level of exposure to pesticides individuals, this is done for groups. Two monitor the trends in incidence of tes- and other contaminants is rather well mon- general dimensions are used: temporal and ticular cancer and birth defects (cryp- itored, whereas very little is known about spatial. Indeed the temporal trends in torchidism, hypospadias), and changes in exposure to other xenoestrogens. selected outcomes have been instrumental semen quality and infertility. Follow-up of in raising the suspicions and concerns semen quality is very important, since the Methods for Evaluation of described in this report. It is hypothesized sperm concentration has decreased drasti- Estrogenlike Effects, Human that putative exposures do vary consider- cally during the last two generations (5) ably along the temporal gradients. Spatial and the declining trend appears to be con- Exposure to Estrogenic trends have so far been exploited in only a tinuing (11). Risk factors for testicular Compounds, and Trends in few cases, e.g., Danish-Finnish sperm con- cancer have been analyzed in several Male Reproductive Health centration and testicular cancer incidence case-control studies. Estrogen treatment of The complexity of our chemical environ- (16,21,22), which also illustrates that mothers whose sons have developed testic- ment is increasing continuously, and it covariation of outcomes can sometimes be ular cancer has remained an equivocal risk may not be possible to forecast all of the informative when few if any exposure mea- factor. It is unfortunate that there are no effects of compounds released in nature. surements are available. In general we are prospective studies yet concerning the tes- However, we have to use all the available not aware of much hard evidence to ticular cancer risk among males exposed to methods and develop new ones to monitor directly support an explanation in terms of DES. The studies by Bibbo and Gill, and possible adverse effects of chemicals and specific exposures of the spatial gradients. their co-workers (174,291) preceded the natural compounds on human and Sources of variation in the spatial gradient time when testicular cancer would have wildlife. Reproduction is a major concern are, for example, ethnic differences, with become manifest. Epidemiological studies because disturbances of this process rapidly the associated difficulties in separating of the risk factors for poor semen quality threaten populations as a whole. The male genetic, cultural, and socioeconomic gradi- and infertility are needed. Furthermore, reproductive system is very sensitive to the ents from the exposures of our prime inter- epidemiological studies should be com- influence of an excess of estrogen; there- est. In particular, the considerable success bined with studies on hormone metabo- fore, estrogenlike effects in the environ- from nutritional epidemiology of migrant lism when a link between disorder and ment are a primary suspect for causing the studies (such as Japanese immigrants to hormone effect is sought. In women, increased reproductive disorders of men Hawaii or California) might form an inspi- dietary influence on estrogen metabolism and wildlife animals. ration for epidemiological designs for the and the risk of breast cancer was analyzed Test methods are needed to screen problem of environmental estrogens. in large epidemiological studies in which chemicals for estrogenic effects. Further Turning to outcomes, the original both exogenous and endogenous hormones methods are needed to assess estrogenic focus was on defects of the male reproduc- and their metabolites were analyzed metic- exposure of humans and other species. tive system. To this we have added specific ulously in blood, urine, and feces (292); Methods to analyze the mechanism of estrogen responses (gynecomastia, male this provided the basis for the link between action of estrogenic compounds are neces- breast cancer). Schematically, there are high levels of free endogenous estrogen sary. And finally, it is important to assess advantages and problems (Table 2). (found in women with Western-type diet) the reproductive toxicity of chemicals and Responses occurring with delay require and increased breast cancer risk. There are to predict their effects in the environment, a long time span for these kinds of repro- no comparable studies among men concern- including the effects on organisms, popula- ductive studies. For both prospective and ing reproductive disorders, although if these tions, communities, and ecosystems. retrospective studies this has negative disorders are induced in fetal/childhood life Epidemiology The general research question is whether Table 2. Assessment of outcomes. exposure to environmental estrogens, in par- Outcome Advantages Problems ticular during fetal or neonatal life, has Hypospadias Inborn: fast ascertainment Not uniformly recorded; no generally adverse effects on the male reproductive sys- Available in some registries agreed diagnostic criteria tem. When designing epidemiological Cryptorchidism Inborn: fast ascertainment Spontaneous descent; no generally studies to evaluate this question, one has to Available in some registries agreed diagnostic criteria consider both exposure and reproductive Semen quality Widely available and reasonably uniform Functional importance of sperm health outcome. Exposure measurements sperm concentration measurements; concentration unknown, other may sometimes be directly available, e.g., sperm output probably related to parameters not often uniformly controlled DES treatment during pregnancy, the Sertoli cell number measured but in most cases such information has to be Fecundity Direct importance Essentially impossible to measure obtained retrospectively. This, however, may separately from confounding factors raise difficulties of recall bias and recall Testicular cancer Well registered, important Rather rare uncertainty (often an answer to a general Male breast cancer Well registered, important Very rare, occurs late in life question such as "have you had hormone

760 Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

but are not manifest until adulthood, then be screened. Biological material used here already proved that such effects can be there are large obstacles to deducing any is either a total extract or a nuclear extract detected (200). The importance of these cause-and-effect relationships. "Time to of cells or a tissue containing a high con- particular studies is that they may provide pregnancy" analysis and exposure assess- centration of estrogen receptor (e.g., the reference values of harmful exposure (i.e., ment allow the study of fecundity of rabbit or rat uterus, MCF-7 cells). Binding pg/kg/day) that can then be related to mea- exposed males (293). experiments are very useful, but it should sured levels of exposure in man. be remembered that binding of a com- The other main uses of mammalian Ecoepidemiology pound does not necessarily imply its animal models will be in determining the Causal relationships between environmen- biologic activity. Moreover, the concentra- mechanisms of xenoestrogen-induced distal contaminants and specific disease states tion of a chemical needed to displace 50% ruption of reproductive development. Mice in wildlife are difficult to obtain because bound [3H]estradiol is orders of magnitude and rats have been used already for some these animals are not living in controlled higher than that needed to elicit a biologi- such studies, and probes or antibodies for conditions. An ecoepidemiologal approach cal response. some of the key genes that are expressed to circumvent these problems is to evaluate There are a number of bioassays avail- during normal reproductive development systematically the relationship between able for assessment of the estrogenicity of a are available. Evaluation of the effects of proposed causal agents and specific out- chemical. These can be divided into two exposure to xenoestrogens at specific times comes. The criteria used to develop causal groups: animal models and in vitro models. during development on expression of these inference include a) probability, b) time Animal Models. A range of different key genes (or on the effects of the gene order, c) strength of association, d) species such as chicken, rat, mouse, trout, products) will obviously be extremely valu- specificity, e) consistency on replication, f) and reptiles has been used in estrogenicity able, especially as any changes identified predictive performance, and g) coherence testing. The most widely used rodent will be able to be related to consequences [reviewed by Fox (294)]. This approach is bioassay measures the increase in uterine in adult life. often more useful in rejecting a causal rela- weight in the rat (295). However, it is There is also a need for exploitation of tionship than in providing definitive sup- well known that this bioassay gives a crude animal models other than rodents, as these port. However, the use of this systematic estimate of an estrogen effect, that it is have limitations in terms of direct access to approach, well-formed sampling designs, insensitive, costly, laborious, and cannot the developing male fetus (i.e., exposure is and rigorous statistical analysis can provide be adopted for large-scale screening. always through the mother). In this respect us with data useful in risk assessment. Moreover, there is no standard procedure there is a place for the use of the fetal sheep Continuous monitoring of the wildlife for the uterotropic assays. Instead, different or pig and the developing opossum. In the populations is needed to detect any distur- laboratories use different protocols (multi- former, it is possible to cannulate blood ves- bances in reproduction. Furthermore, tim- ple doses vs single dose, 24 vs 72 hr, etc.). sels in the fetus at mid-term and to then ing and magnitude of chemical exposure Vitellogenesis in male fish, e.g., trout, can directly administer a compound to the fetus should be considered when estrogenic be used to test for the presence ofestrogenic and to monitor consequences in terms of effects are examined. compounds in the aquatic environment. changes in blood hormone levels (165); for Mammalian animal models. Various example, this system could be used to Experimental Models for Testing laboratory or domestic animal species can determine what blood level of xenoestrogen Estrogenicity and Estrogenic Molecules be used to address specific aspects of the in the fetus results in reduced secretion of Different approaches can be used to assay possible effects in vivo of estrogenic chemi- FSH from the pituitary gland and is thus estrogenic activity, depending on whether cals. These should be particularly useful for likely to reduce Sertoli cell multiplication or not the compound is known to have studies of delayed effects (i.e., where expo- (see "Effects of Synthetic Estrogens on the such an activity. In the first case, direct sure to the chemical(s)occurs in fetal/neona- Testis in Animal Models"). In contrast, the physicochemical analyses, such as gas chro- tal life and the reproductive consequences opossum provides fairly unique direct access matography-mass spectroscopy or high- are manifest in adult life) and for studies of to a fetus that is born before sexual differen- performance liquid chromatography the effects of chronic low-level exposure to tiation has occurred. Thus administration of (HPLC), can be used to assess the pres- xenoestrogens. For example, possible xenoestrogens to the young at this stage ence, and if necessary, the concentration of effects of these chemicals on Sertoli cell (either through the skin or orally) could a xenoestrogen(s). Many banned chemicals multiplication can be studied in developing provide a direct and controlled way of pin- (e.g., PCBs) can be detected by this rats or mice by exposing them during ges- pointing when and how such exposure method. Specific immunoassays can be tation and/or during the first 3 postnatal might result in disruption of the normal cas- used when appropriate antibodies are avail- weeks of life (the period of Sertoli cell mul- cade of male reproductive development; this able. Different approaches must be used tiplication extends from day 16 of gesta- would enable identification of critical win- when putative estrogenicity of a com- tion to postnatal days 15-18 in the rat). dows of developmental susceptibility to pound(s) is screened or the cause of a sus- Sperm output, fertility, and testis size and estrogens. It is established already that a) the pected estrogenic contamination is searched morphology can then be evaluated in adult pathway of development in these animals is for. Binding assays and bioassays are com- life and Sertoli cell number determined in comparable to that in eutherian mammals, monly used for this purpose. other animals at around day 18 of postna- and b) exposure of the young to estrogens is The principle of binding assays is to try tal life. In this way, possible chronic effects able to disrupt completely the normal differ- to displace radiolabeled estradiol from its of low-level exposure to xenoestrogens dur- entiation of the testis and seminiferous cord receptor by increasing molar concentrations ing development on adult reproductive formation (296). of either unlabeled estradiol (reference ability can be evaluated in a fairly simple Vitellogenesis. Vitellogenin is a yolk curve) or different purified xenobiotics to way. Such studies are ongoing and have protein produced normally in the liver of

Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 761 TOPPARI ET AL.

female fish under estrogen control (297). in vitro assays are the MCF-7 cell tests and Estrogen effects on MCF-7 cells can Very little, if any, vitellogenin can be the recombinant yeast cell assays. Serum also be monitored after transient transfec- detected in male fish (298). However, that is required in the culture of breast- tion of cells with reporter plasmids exposure to estrogens activates the vitel- cancer cell lines must be charcoal-stripped pTKLUC and pERE-TKLUC by assaying logenin gene, resulting in increased vitel- to remove endogenous estrogens. for luciferase activity (203). The reporter logenin levels in the blood of male fish, MCF-7 cell line. Human breast cancer plasmid pTKLUC contains the herpes sim- and clear dose-response effects have been cell lines (MCF-7 cells) that are sensitive to plex virus thymidine kinase (TK) promoter documented (299). Vitellogenesis can be estrogen have been used to screen chemi- inserted in the Bgl II site of the luciferase used both in field studies and in laboratory cals for their estrogenic effects (81). The reporter plasmid pGL2-Basic (Promega). in vitro studies (300). It is measured by a test -also called the E-SCREEN -is pERE-TKLUC contains a single copy of specific radioimmunoassay (301). Sex based on the dose-response relationship the vitellogenin A2 estrogen response ele- reversal could also be used as an end point between the proliferation of MCF-7 cells ment inserted upstream of the TK pro- for estrogenic effects in an aquatic milieu. and the amount of estrogen to which the moter in pTKLUC. In field studies, caged male rainbow cells are exposed during 6 days of culture. ZR-75 cell line. Several human estro- trout are exposed to water in areas of inter- Estradiol is used as a standard. By compar- gen-responsive cell lines, such as ZR-75, in est, and their vitellogenin levels are fol- ing the effects of a xenoestrogen and estra- addition to MCF-7 can be used in cell pro- lowed over time (112). This is a good diol, we can present the relative estrogenic liferation assays (203). screening method for the presence of estro- potency of a compound. Soto et al. (81) Recombinant yeast cell lines. The genic compounds in different aquatic habi- use the following concepts to describe most sensitive estrogen measurement at the tats. Vitellogenin production by superfused estrogenicity: "Concentration" denotes the moment is a recombinant cell bioassay com- trout hepatocytes can also be used as a sen- dose at which an estrogenic effect is prising a yeast cell line (Saccharomyces cere- sitive in vitro bioassay for analysis of estro- detected; proliferative efficiency (PE) mea- visiae) transformed with plasmids encoding genic compounds in the water (111). sures the ratio between the highest cell the human estrogen receptor and an estro- Fish reproductive tests. The Office of number in the presence and in the absence gen-responsive promoter fused to the struc- Pollution Prevention and Toxics (OPPT) of estrogen; relative proliferative efficiency tural gene for ,B-galactosidase (303). This in the U.S. Environmental Protection (RPE) measures the ratio between the max- assay detects 100-fold lower estradiol levels Agency (U.S. EPA) is using two fish repro- imal cell yield achieved by a xenobiotic and than the widely used traditional radioim- ductive assays: a) fish partial chronic toxic- that of estradiol; relative proliferative munoassays. However, the sensitivity of the ity test that measures effects from adult to potency (RPP) measures the ratio between assay to other estrogenic hormones than the early life stage, and b) fish whole life the dose of xenobiotic and that of estradiol estradiol is lower. This bioassay revealed an toxicity test that measures effects from egg needed to achieve a proliferative effect. 8-fold difference in the estrogen levels of to early life stage to adult and then to early Alkylphenols, pesticides, phytoestrogens, prepubertal boys and girls (303), who were life stage (OPPTS 850.1500). and synthetic estrogens have been analyzed previously believed to have similar hor- Reptile egg assays. Reptile sex differen- using this test (81,212). Some endocrinol- mone concentrations (304). This has tiation is temperature- and estrogen- ogists consider cell proliferation a hallmark important implications for future studies dependent (155). Exposure of turtle eggs of estrogen action; and if a compound does since it is now possible to measure quanti- to a putative xenoestrogen by spotting or not have this property, regardless of what tatively increased estrogen levels in prepu- painting them with the compound results other end points it influences, it should not bertal boys. The assay is simple to perform in a high proportion of feminization or be called an estrogen. The E-SCREEN is once the cell line is established since only intersex conditions of hatchlings in a tem- technically an easy method and can be used the measurement of P-galactosidase activity perature that would normally result in for screening and assessment of approxi- is needed to observe the effect. 100% males (99). This assay is suitable for mate estrogenic potency of a compound. Sensitivity ofthe Assays. The sensitiv- studies ofestrogen exposure in wildlife. Other hormones and growth factors tested ity of the assays for measurement of 171- In Vitro Assays. Two complementary so far have not influenced this assay (204). estradiol is given in the Table 3. aspects can be studied by in vitro assays: However, the possibility remains that some The recombinant yeast cell assay proliferation of an estrogen-dependent cell compounds may influence these cells appears to be the most sensitive of the line, and the induction of an estrogen-con- through other than estrogenic pathways assays, although the E-SCREEN may not trolled function. In the first case, cell num- and thereby confound the results. be far behind. Both assays detect estradiol ber and the incorporation of radiolabeled MCF-7 cells were also used to detect the levels that are less than 0.1 pg/ml. thymidine are the parameters measured. In estrogenic effect of bisphenol-A that was Traditional radioimmunoassays with an the second case (induction of function), released from polycarbonate flasks during antibody strictly specific for estradiol had a there are several possibilities. Prolactin pro- autoclaving (267). Progesterone receptor maximum sensitivity of 2 pg/ml (305). duction by pituitary cells in response to an induction and [3H]thymidine incorporation However, sensitivity can be improved to estrogenic compound was an early attempt combined with DNA measurement were less than 0.1 pg/ml (306). In the E- to measure estrogenic activity in vitro, but used to measure estrogen activity (81). This SCREEN assay, the lowest estradiol level to the assay was abandoned because of its approach covers the proliferative effect produce a maximal proliferative effect is 10 nonspecificity. For assessment of the estro- (thymidine incorporation) and the change- pg/ml, but detectable proliferation occurs genic activity in the aquatic environment, of-function effect (progesterone receptor already at 0.3 pg/ml (307). Activity, the in vitro assay for detecting vitellogenin induction) of the compound studied. This significandy different from the unstimulated in fish is extremely useful (302). At pre- adds to the technical difficulty and the control, has been detected at 0.03 pg/ml (C sent, the most commonly used mammalian expense of the assay. Sonnenschein and AM Soto, personal

762 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

Testicular Cancer ModeLs Table 3. Sensitivities of estrogen assays. There are no animal models for testicular Assay Sensitivity References germ cell tumors. This necessitates more Radioimmunoassay (RIA) Traditional RIA: 2 pg/ml Korenman et al. (305) intensive search for such a model. Improved RIA: 0.3 pg/ml Legan et al. (306) Testicular stromal tumors can be produced MCF-7 cell assay (E-SCREEN) 0.03-0.3 pg/mI Soto and Sonnenschein (307) in mice both by gene deletion (309) and AM Soto and C Sonnenschein, personal communica- by overexpression of an oncogene under tion the control of a testis-specific promoter Recombinant yeast cell assay 0.02-0.2 pg/ml Klein et al. (303); JP Sumpter, personal communication (e.g., inhibin alpha promoter connected to SV 40 large-T antigen). The transgene approach has been used to produce testicu- communication). When the ultrasensitive Although some proestrogens are sup- lar somatic cell tumors from which new tes- estrogen assays are used, extreme care posed to be metabolized into an active form ticular cell lines have been developed (K should be taken in the way the samples (to in the liver, and thus they are expected to Kananen, personal communication). Direct be analyzed) are handled, stored (quality of be inactive in in vitro assays, such com- transformation of germ cells with SV 40 plastic vials), and processed, because major pounds (e.g., alkylphenol monoethoxylates, large-T antigen has been used to immortal- problems may be encountered from low some PCB congeners, methoxychlor) have ize spermatogonia (310). When a tempera- levels of contaminating estrogens, e.g., been found to be estrogenic for MCF-7 ture-sensitive p53 mutant is co-transfected plasticizers leaching from plastic storage cells, suggesting that these cells may also into these cells they were reported to differ- tubes, syringes, cell culture plates, etc. metabolize some chemicals (204,257). entiate through meiosis (311). These Complementary to these techniques, recombinant cell models may give possibil- the binding of the compounds to proteins, General Reproductive Toxicity Testing ities for detailed analyses of many features such as SHBG and a-fetoprotein (AFP), Only a few chemicals have been tested ofgene activity and its regulation in specific should always be studied in order to address specifically for estrogenic or other testicular cell types. Although the cells are the difficult problem ofbioavailability. endocrine activity. However, many chemi- transformed, most of their characteristic In conclusion, although a wide variety cals have been tested for their reproductive functions should remain unchanged. Such of different assays are used (and probably toxicity, and thereby indirectly for estro- approaches may be extremely useful in always will be!), in general they have pro- genicity, although weak estrogenic activity improving our understanding of primordial duced very consistent results; that is, a most probably does not.appear in these germ cell development and of the factors chemical that is estrogenic in one assay is tests. Current toxicity tests are designed pri- that are involved in arrest of development estrogenic in all other assays. For example, marily to detect overt teratogenic and resulting in precancerous carcinoma in situ alkylphenolic compounds are weakly reproductive effects or toxicity in the (CIS) cells. estrogenic in molecular assays (257), cell adult-they have not been designed to seek Brinster and co-workers (312,313) proliferation assays (81), and in vitro estro- out effects where there is a long lag-phase demonstrated that stem cell spermatogonia gen-binding assays (203). There appears to between induction and manifestation of the survived a transit to mouse seminiferous be no species specificity; i.e., a chemical effect. Moreover, unless such changes were tubules that were either partially or totally that is estrogenic in one animal appears to gross and pathological, they would proba- missing their own germ cells. The trans- be estrogenic in all other species. Again, bly be dismissed as insignificant, e.g., small ferred spermatogonia produced normal alkylphenolic compounds provide a good reduction in testis size and sperm count. spermatozoa. Spermatogenic cells survive example. These chemicals are weakly estro- Toxicity testing used by U.S. regulatory poorly in vitro, but spermatogonia are the genic in fish, birds, and mammals (257). agencies and by the National Toxicology most viable (314-316). This raises the pos- This is not at all surprising since the struc- Program includes three test segments sibility of in vitro experimentation, which ture of the estrogen receptor is highly con- (Segments I-III), Fertility Assessment by can be continued in vivo by injecting cul- served between species. Continuous Breeding (FACB), and either tured spermatogonia back into the seminif- Limitations of in Vitro Testing. two-generation or multigeneration tests erous tubules. Human CIS cells could be Absorption and metabolism of a compound (308). Segment I covers fertility and repro- transferred to mouse seminiferous tubules are important for its effects. Because a chem- ductive function in male and female; (which are an immunologically privileged ical is, or is not, estrogenic in vitro does not Segment II deals with developmental toxi- site) to follow possible carcinogenesis. necessarily mean that it will exhibit this cology and teratology; and Segment III Endocrine manipulation of recipient ani- activity in vivo. It may not be absorbed. It tests perinatal and postnatal toxicity. mals would enable study of the role of hor- may be metabolized in the gut to an inactive Thus, reproductive toxicity testing mones in the promotion of tumor growth. or nonabsorbable metabolite. Alternatively, a belongs to the routine toxicology that has chemical that is nonestrogenic in vitro may to be undertaken before a new compound Summary be converted to an estrogenic metabolite in can be released to the environment. The Multidisciplinary approaches are needed either the gut or the liver. This possibility is problem is that although testing appears into studies of environmental effects on important as it points out the limitation of extensive, it may miss important late effects male reproductive health. Epidemiological in vitro screens (false negatives). Screening of and actions that occur only when additive methods should be used to follow up the classes of compounds and identification of factors are present. The reproductive toxic- trends in male reproductive health and to the chemical structure(s) associated with ity tests used in the current safety assess- identify putative association to environ- estrogenicity are obvious ways by which ment of various chemicals are discussed in mental factors. Similarly, ecoepidemiology this limitation could be minimized. more detail in "Appendix B". gives valuable information, on concurrent

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 763 TOPPARI ET AL.

changes in wildlife. Experimental studies in their reproductive toxicity; even where reproductive system (breast, prostate). laboratory animals and cell lines provide such studies have been performed, their The etiology of these differences should direct evidence of the effects of xenobi- design and intent were such that they be explored to ascertain to what extent otics. Both physicochemical and biological may have failed to exclude delayed they reflect ethnic/genetic, lifestyle, or methods should be used for the effects of the sort that can be induced environmental influences. identification of estrogenic-/endocrine-dis- by estrogen (or other hormone) expo- * There is an urgent need for a clearer rupting compounds in the environment sure during development. understanding of the hormonal envi- and for the determination of human expo- * If the above changes in male reproduc- ronment ofthe fetus during the period of sure to these agents. tive health do result from fetal and/or male sexual differentiation and develop- neonatal exposure to environmental ment, especially of what mechanisms are General Discussion on the estrogens, then we will not know what present to protect the fetus from mater- Association between the current prevalence in male children nal estrogens, e.g., ox-fetoprotein, steroid- Chemicals in the Environment is for another 20 to 40 years. This is metabolizing enzymes, etc. Because of due to the delay between induction and the inaccessibility of the human fetus at and Male Reproductive Health manifestation of many of the effects, this stage, these studies will probably The possibility that exposure of humans e.g., lowered sperm counts or the devel- have to be undertaken in animals. and wildlife to environmental estrogenic opment of testicular cancer in adult * Investment of resources in obtaining the chemicals might result in adverse changes life. The indications are that these information necessary to assess whether in reproductive development, function problems are still increasing in inci- environmental estrogens pose a health and/or behavior is not particularly new; dence in the general population. risk to man will have major spin-off in concern was first expressed two decades ago * Based on what is known about the terms of our understanding of fetal and in relation to DDT. Indeed, the first widespread imprinting effects of steroid neonatal determinants of disease, the demonstration that a range of man-made hormones (androgens and estrogens as routes and importance of human expo- chemicals could be estrogenic when admin- well as antagonists of these hormones) sure to nonpesticide environmental istered to animals stems from 1938 (317). when there is exposure during chemicals and of the etiology of geo- The issue has resurfaced because of two fetal/neonatal development, it is possi- graphic and ethnic differences in disease. new developments. The first is the increas- ble that there are other adverse health ing appearance in the human population of consequences in man of such exposures Suggestion for a Strategy to several adverse changes of the sort that ofwhich we are currently unaware, e.g., Strengthen the Evaluation of researchers predicted might occur if human permanent changes to the immune sys- exposure to environmental estrogens was tem, growth, etc. Hazards of Estrogens or widespread. The second has been the dis- * There is enough evidence from a variety Xenoestrogens covery of many new environmental estro- of sources to suggest that environmental Several lines of evidence have indicated gens to which we are exposed daily. In light estrogens have the capacity to induce adverse trends in male reproductive health of these developments, the foregoing report adverse minor and major reproductive over the last few decades. Clinical and has reevaluated the strength of the available defects in man, but there are inadequate experimental research has demonstrated information to ascertain whether or not data on the level of actual human expo- that the reproductive disorders reported there is real concern for human health sure to these chemicals, which means may be interrelated and may have a com- stemming from exposure to environmental that no accurate risk assessment can be mon origin in fetal life or childhood. estrogens, or whether the data are more made. There are also insufficient animal Studies of wildlife and human epidemiol- consistent with a theoretical, but unlikely, data to permit such an assessment. ogy suggest that environmental factors are effect on the human. The following step- * Determination ofwhether environmen- involved in the origin of reproductive disor- wise conclusions have been reached: tal estrogens do exert harmful effects in ders. Common environmental contami- * All of the best evidence available points man will probably have to rely on nants such as alkylphenols and phthalate with some certainty to a rising tide in weight of evidence rather than estab- esters, and natural factors such as phytoe- Europe and many other countries of lishment of precise cause and effect. To strogens, have been shown to be endocrine- human male reproductive disorders provide this evidence as accurately as disrupting agents, many of them being involving sperm counts (and probably possible, future work should address estrogenic. Our hypothesis is that the sperm quality), testicular cancer, mal- the problem at several levels, including adverse trends in human male reproductive formation of the external genitalia, and epidemiology, laboratory animal health is, at least in part, associated with possibly testicular maldescent. studies, wildlife studies, more complete exposure to environmental estrogenic com- * There are insufficient data to prove or identification of estrogenic chemicals pounds during early development. Testing disprove that these adverse changes in (i.e., screening), and accurate assess- of this hypothesis necessitates a multidisci- male reproductive health are the result, ment of the routes and levels of human plinary research approach, including epi- wholly or partially, of exposure to envi- exposure to these chemicals. demiological and experimental studies, ronmental estrogens. In addressing this * There are substantial differences in the examination of wildlife, and analyses of issue, we have identified alarming gaps incidence of the reproductive defects, environmental contaminants and human in our knowledge of the route and lev- referred to previously, in different coun- exposure to them. Extensive international els of human exposure to nonpesticide tries/races and these also appear to show collaboration that combines the existing environmental chemicals. Many ofthese possible relationships to the incidence of strengths of various institutions and inves- chemicals have not been evaluated for other hormone-dependent diseases of the tigators is probably required if this complex

764 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

Male reproductive other aspects of male reproductive function. health disorders Semen quality, testis size, levels of sex hormones and gonadotropins as well as presence of cryptorchidism and gynecomastia Human epidemiology Animal epidemiology may all be indicators of abnormal male Incidence of male reproductive disorders Incidence of male reproductive disorders reproductive function. The aim of the Follow-up on humans exposed to in domestic and wildlife species estrogens/xenoestrogens Follow-up on wildlife exposed to study is to examine geographic and racial xenoestrogens differences and to establish reference values for semen quality, blood levels of sex hormones and gonadotropins, testis size, and prevalence of gynecomastia and cryp- Steering committees ------Task forces ------Workshops torchidism, since these conditions may provide us with information on environmental 4 4 and genetic influences on male reproductive development and function. Furthermore, Human risk such studies may be starting points for future prospective studies on secular trends assessment Monitoring of wildlife populations (vitellogenin) in male reproductive function. aStudies in laboratory In order to obtain comparable groups animals I_ of individuals, well-defined types of men __ tA should be investigated, such as army recruits and semen donor candidates. The Biological and biochemical consequences of estrogen/ study should involve men from the United xenoestrogen exposure States, France, Finland, Denmark, United (in vivo screening) Kingdom, and possibly Taiwan or Japan, with the final selection of countries to be determined by the study groups. Environmental The following parameters should be profile assessed: * semen quality: volume, density, morphology, computerized motility assessment; to ensure a standardization of the methods, the laboratories involved in the project should circulate relevant samples Figure 8. Suggestion for a research strategy to strenghten the evaluation of hazards of estrogens or xenoestrogens. or videotapes * serum levels of testosterone, estradiol, FSH, and LH health problem is to be addressed in the results in biochemical and biological * morphometric measurements: weight, most accurate and effective way. effects (in vivo screening). This approach height, testis size would be complemented by studies on the * cryptorchidism Research Strategy environmental distribution of these chemi- * gynecomastia The proposed research strategy is illus- cals and by assessment of the routes and * time to pregnancy if the men have trated schematically in Figure 8 and is level of human exposure. This informa- reached an age when they have wished expanded upon in the relevant sections tion, in conjunction with the relevant to become fathers below. There are three fundamental com- epidemiological evidence, should enable The rest of serum and semen samples ponents to this strategy (boxed in Figure 8). some form of risk assessment to be made should be stored in the freezer for possible The first component centers on the col- for man and for wildlife. future analysis for, e.g., environmental pol- lection and evaluation of epidemiological lutants or other relevant parameters. evidence for male reproductive disorders in Epidemiological Studies in Man Denmark and Finland provide an inter- man and animals. The intention of this In epidemiological studies, trends in male esting comparison. The incidence of testic- approach is to define the scale of male repro- reproductive health must be followed up ular cancer in Denmark is remarkably ductive disorders, to establish geographical further in different geographic areas and in higher than in Finland (22). Conversely, differences in their distribution and to assess populations with different ethnic and racial semen quality seems to be markedly better whether there is an increase in -incidence of backgrounds. In addition, every attempt in Finland compared to Denmark (16). such disorders in groups with known expo- should be made to use some of the existing The incidence of testicular cancer is sure to estrogens or xenoestrogens. data or cohort studies, if these are appro- increasing in both countries, whereas less is The second and third components of priate, although in some instances collec- known about semen quality. The incidence the strategy are designed to identify xenoe- tion of new data is essential. of cryptorchidism in Denmark in 1960 is strogens using a variety of in vitro screen- Incidence of Male Reproductive known from a careful study of 2701 new- ing methods and then to evaluate whether Disorders in Man. There are geographic borns (60). Additional information is their administration to laboratory animals and racial variation in semen quality and available from the doctors' and midwives'

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 765 TOPPARI ET AL.

notes of 17,767 newborn males during the breast cancer. This information could among domestic animals. Such animals period 1957 to 1960. No information on be obtained from many, if not all, of might serve as good models for evaluation possible increase in the incidence of cryp- the countries in the EU. of the environmental impact on human torchidism in Denmark is available and * Information from the newborn cohort male reproduction, although it should be neither is there information on the inci- established in the 1960s in Denmark kept in mind that, for breeding of many of dence of other minor malformations of the could also be used to analyze the trends these species, individuals with the best male external genitalia. There are no large for genital malformations through a new reproductive function have been highly studies on the incidence of cryptorchidism study. The Bristol-coordinated European selected. Nevertheless, if any adverse trends in Finland. According to the hypothesis cohort study which involves large num- are still evident, this would provide power- that testicular cancer, poor semen quality bers of children in several Western and ful supporting data for the human studies and developmental anomalies in the male Eastern European countries will give and might open up new possibilities for reproductive tract are interrelated, we pos- additional data on these trends. identifying causal agents. tulate that the incidence of cryptorchidism Risk factors may be shared among The species that might be of interest are is higher in Denmark than in Finland and many disorders. Mothers of men with tes- the bull, pig, stallion, and dog. The indica- that the prevalence of genital malforma- ticular cancer were reported to have an tors of male reproductive studies are, as in tions is increasing in newborns. increased risk for breast cancer (190), sug- humans, cryptorchidism, hypospadias ,and Follow-up Data on Humans Exposed gesting common risk factors, such as a semen quality. It is well known that dogs to EstrogensiXenoestrogens. There are sit- high estrogen exposure. In national cancer develop spermatocytic seminoma and not uations in which exposure of pregnant registries, it is possible to analyze the asso- the classical, premeiotic germ cell-derived women to estrogens or estrogenic com- ciation between testicular cancer and seminomas and nonseminomas. It remains pounds have been clearly identified. DES mother's breast cancer in large patient to be seen whether such tumors occur treatment in the United States and some groups, and such a study is ongoing for among the other species. Available retro- countries of Europe or contamination of example in Denmark. spective data (e.g., databases for artificial foodstuffs with PCBs and PCDFs in Japan Occupational Studies. Studies on the insemination stations) on reproductive and Taiwan are such examples. offspring of women occupationally exposed function in these animals should be ana- These situations are true experimental to xenoestrogens. Comparison between lyzed. New, prospective studies should be models and maximum effort should be population groups in so-called ecological initiated, if suitable studies can be designed made to collect existing data on the off- studies may suffer from various weaknesses, and the archival data warrant such investiga- spring of these women. The following are including confounding bias. Also, if the tion. The overall importance of these analy- the possibilities: main effect on male reproductive function ses is that they may indicate whether adverse * Obtain additional information from is due to prenatal exposure, it may be very changes in male reproductive function are men born to mothers who participated difficult to determine the causative expo- unique to man or more prevalent in man. in the well-designed DES trial made in sure in any detail because it happened in the United States in the 1950s (167). the distant past. These problems do not Wildlife Studies There have been more than 300 men indicate that such studies should not be If environmental estrogens are currently in the follow-up studies of these carried out; but it would be important to present in biologically significant back- cohorts (174). supplement this evidence with studies of ground concentrations, a survey of wildlife * Obtain information from other men more specific exposures, such as those that would provide important supportive infor- born to exposed mothers. occur in certain occupations. mation concerning possible human risk. A possibility would be to study brothers Among the industrial chemicals so far One extremely useful biomarker of estro- of women known to be exposed to DES, as identified as xenoestrogens are certain addi- gen exposure is the synthesis of the hepatic there is probably an increased possibility tives to plastic materials (e.g., alkylphenols, protein vitellogenin. This is normally syn- that they also were exposed to DES. The phthalates, bisphenol-A) and some pesti- thesized in the liver of egg-laying females size of the study group and a paired control cides. It would therefore be of interest to after estrogen stimulation. Synthesis of this group would have to be defined carefully. find female occupational groups with protein is usually correlated with seasonal * New case-control studies should be known exposures to such compounds. One reproductive activity in females but it is done especially for testicular cancer, group of particular interest is greenhouse not found in males. Numerous laboratory since, e.g., the DES-exposed males have workers. Because of the enclosed space of studies, with a variety of wildlife species, now reached the age at which the inci- greenhouses, pesticide exposure invariably have demonstrated that vitellogenin syn- dence oftesticular cancer peaks. occurs, partly because of percutaneous thesis is stimulated in males after exoge- * Taiwanese boys prenatally exposed to absorption. Prospective and cross-sectional nous estrogen treatment. Recent evidence PCBs are now in puberty (281). It is studies of the male offspring of the female (112) suggests that both caged male trout proposed to obtain all the available data greenhouse workers could then be carried in English rivers receiving sewage effluent on the reproductive tract disorders out with emphasis on cryptorchidism and and males of native fish species from vari- from these men, and later semen analy- other indicators of adverse effects on repro- ous English rivers and from the river Seine ses when available. ductive organs. in Paris exhibit elevated plasma vitellogenin More information is needed on the levels. These data suggest that environmen- incidence and trends for genital defects in Epidemiological Studies in tal estrogens exist in these rivers at biologi- various countries. Domestic Animals cally significant levels. Thus, a survey of * Attention should be mainly accorded to There is a lack of information on possible male fish, amphibians, reptiles, and birds cryptorchidism, hypospadias, and male secular trends in male reproductive function living in wetland areas would provide a

766 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

definitive measure of estrogen exposure in possible antiandrogens or antiestrogens in and binding components. As an example, this environment. the environment. Additional assays for an early step would be to study the relative Fish would provide a very important other biomarkers may be required, binding in vitro of selected chemicals to model system as they transport large quan- although the plasma steroid analyses per- SHBG, AFP, and the estrogen receptor. tities of water through gills that are highly formed simultaneously with vitellogenin To understand the etiology of estrogen- vascular. Second, various species feed at dif- assays may provide clues to reproductive associated developmental abnormalities fering trophic levels of the food chain, mak- abnormalities due to other mechanisms of that may be related to later disease or dys- ing a comparison among levels possible. endocrine disruption. function, it is crucially important to study Additionally, other species such as frogs, Surveys need to be performed on a the ontogeny of estrogen responsiveness in salamanders, turtles, and alligators or croco- series of wetlands that represent a contin- target tissues associated with reproduction diles could be used, depending on the eco- uum in water quality. Sites with known in appropriate animal models. In addition logical knowledge base for such species and contamination will serve as positive con- to whole-animal studies, experiments the specific questions to be addressed. For trols, whereas other localities with reduced should be attempted in organ and cell cul- example, many fish and amphibian species input of manmade chemical contaminants tures of fetal genital and gonadal tissues; are short lived and are limited to specific can serve as negative controls. One point of additionally, studies with transgenic mice water bodies whereas other fish, turtles, and caution should be made concerning the would be illustrative. These studies should alligators are long lived and may migrate negative control localities-areas perceived certainly include or emphasize the between various wetland areas. as being pristine, such as arctic and antarc- ontogeny and regulation of the estrogen We would propose that a series of tic regions, are known to have animal and receptor and related signaling molecules. rivers, lakes, and marshlands be examined human populations with elevated body This information would help to predict in Europe and North America. If possible, burdens of various persistent environmen- which chemicals would most likely alter a few common species from similar genera tal contaminants. Thus, it is important sexual development and provide strategies should be examined to minimize inter- that representative species be examined for prevention. species variation. However, as the relative from each test locality for persistent conta- Estrogenic chemicals can alter male sensitivities ofvarious species are unknown, minants (e.g., DDE, total PCBs, dioxin). reproduction through chronic or acute this survey should examine species that These measurements of contaminant load mechanisms. Long-term exposure to these occur in large numbers. Detailed sampling should not be used as a measure of possible chemicals during fetal/neonatal life can should be designed and similar protocols estrogenic contamination, but rather as a alter the expression of various other hor- used. In addition to vitellogenin, plasma general measure of possible environmental mones, leading to impaired development androgens and estrogens should be exam- pollution. Initial surveys should examine as of the testis such that the capacity for ined to ensure that natural estrogens are many male fish as possible (n=-100 males/ sperm production in adult life is irre- not elevated in those males that exhibit a wetland). Blood samples would be col- versibly reduced. Direct effects on the positive response for plasma vitellogenin. lected rapidly for analysis of vitellogenin, adult testis that interfere with sperm pro- These data would also provide supporting androgens, and estrogens. Specific body duction are also possible, although less evidence of abnormal reproductive activity characteristics should be noted such as size, likely. On the other hand, hormonally as previous studies have demonstrated body weight, and other descriptive mor- active xenobiotics can alter a process or abnormal steroidogenesis following expo- phometric characteristics. For alligators or expression of a gene during a short window sure to various contaminants. other wildlife having phallic structures, of susceptibility that has long-lasting con- The limitation of this study, as pre- measurements of size and developmental sequences. Characterization in detail of sented, is the need for antibodies that rec- abnormalities should be noted. these later consequences is recommended. ognize vitellogenin from the various species There are a few key genes involved in the of interest. Vitellogenin is a large, globular Experimental Studies -in VWo and development of the reproductive tract and protein, approximately 400 to 500 kD, in Vivo Screening gonads, and these provide important targets with a complex, three-dimensional struc- The action of estrogenic chemicals, espe- for study. For example, the effect of estro- ture. Currently, a universal antibody that cially during fetal development in most gen on the expression, regulation, and would cross-react with the vitellogenin of species, is, in part, determined by their rel- structure of SRY, Wilm's tumor gene (WT- all oviparous species is not available but a ative binding with intracellular targets and 1), MIS, MIS receptor, growth factor number of laboratories are presently devel- extracellular binding proteins. Estrogenic receptors and their ligands, estrogen recep- oping and screening possible candidates. chemicals with relatively poor binding to tor, genes for steroidogenic and steroid- Collaboration with these laboratories extracellular proteins, such as SHBG and metabolizing enzymes, and homeotic genes should be fostered. However, if a universal AFP, would be expected to have a dispro- and related genes should be determined. assay is not possible, other methods such as portionate effect on developing estrogen The class of chemicals called environ- polyacrylamide gel electrophoresis (SDS- target tissues. This is the case for DES. mental estrogens comprises a diverse group PAGE) can be used to detect plasma vitel- Experiments are proposed to evaluate the of molecules. Studies to determine the logenin, although this technique would micropharmacokinetics of a selected group metabolism, distribution, bioaccumulation, provide poor quantification of plasma con- of priority environmental agents with estro- and excretion of representative members of centrations as compared to a validated genic activity with regard to distribution this functional class in adult, pregnant, and radioimmunoassay. An additional limita- within both the maternal and fetal com- fetal systems (both in vivo and in vitro) are tion of this survey involves the specificity of partments. Care must be given as to selec- proposed. This also includes identification vitellogenin as an estrogen-induced prod- tion of the appropriate animal model with of their active estrogenic forms. These data uct. A survey for vitellogenin will not detect special regard to endogenous estrogen levels are important to make accurate exposure

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 767 TOPPARI ET AL.

estimates for susceptible target tissues in the * How are these chemicals distributed in male reproductive health have changed human fetus. the body? Are they mobilized (redis- dramatically for the worse over the past 30 The incidence of testicular cancer in tributed) in certain states, such as to 50 years. The most fundamental change humans is still increasing while its etiology is pregnancy? has been the striking decline in sperm still unknown. A plausible hypothesis is that * How are they metabolized? Is it the counts in the ejaculate of normal men; estrogenic exposure early in testicular devel- metabolites, rather than the parent recent evidence from Paris indicates that opment is associated with later germ cell compounds, that are of concern? For this decrease amounts to about 2% per year neoplasia in humans. Studies to elucidate example, alkylphenol polyethoxylates (a over the last two decades. The result is that molecular and morphological markers for group of surfactants) are not estrogenic, many otherwise normal men now have testicular development are needed. The cell but their degradation products are. sperm counts so low that their fertility is type at risk for neoplastic transformation * If humans (and wildlife) are exposed likely to be impaired. Over the last half- should also be studied and the process deter- simultaneously to a mixture of estro- century, the incidence of testicular cancer mined. Creative application ofmodern mol- genic, and/or antiestrogenic, anti- has increased progressively in many coun- ecular biological and cell biological androgenic, or otherwise endocrine tries to become now the most common techniques should be considered to establish disrupting compounds, rather than to cancer in young men. Other disorders of the malignant cell lines. Targeting germ an individual chemical, would the over- the male reproductive tract may also be cell-specific gene promoters for transgenic all effect be diminished or enhanced increasing in incidence, with several approaches to specific cell transformation (additivity, synergism, antagonism)? European countries reporting a progressive and the development of new animal models Presently we do not have adequate rise in hypospadias (a malformation of the to study should be encouraged. Studies on answers (or even preliminary answers in external genitalia) and an apparently emerg- the physiology and pathophysiology of some cases) to any of these questions, ing trend toward an increasing incidence of estrogens in developing and adult male mainly because we do not know what testicular maldescent. germ cells are needed to evaluate the role of chemicals are of the greatest concern. These observations suggest that male estrogens in testicular neoplasia. Attention A major emphasis should be placed on reproductive health has declined progres- should be given to new experimental mod- how to identify priority compounds that sively since the Second World War as a els, such as the pig and nonhuman primates. justify examination (in addition to those result of changes in environmental or In order to improve the safety assessment already identified as estrogens). The aim lifestyle factors. While the etiologies under- of environmental chemicals, it should be would be to test for estrogenic activity of lying these apparent changes are currently assured that future toxicological testing used the chemicals to which man (and probably unclear, both clinical and laboratory for the generation of safety data adequately wildlife) is exposed the most. research suggests that all of the described takes into account the possible detrimental The usefulness of various sources of changes in male reproductive health appear effects on male reproduction of chemicals information should be explored, such as interrelated and may have a common origin possessing estrogenic and/or otherwise existing product registers and OECD data in fetal life or childhood. This means that endocrine-disrupting activity. This probably on high-volume chemicals. In addition, the increase in some of the disorders seen necessitates the development of new test advice and help may be obtained from today originated 20 to 40 years ago, and strategies. It is anticipated that the goal can industry, which may be able to assist in the prevalence of such defects in male be achieved by combining modifications of identifying the more important chemicals. babies born today will not become mani- existing protocols for reproductive toxicity Collectively, these approaches should fest for another 20 to 40 years or more. testing with the introduction of new in vitro enable identification of the major man- Trends in the reproductive health of and in vivo assays. The strategy should be made chemicals. These chemicals could species other than man also raise the possi- based on insight into the mechanisms of then be screened to determine whether bility of environmental factors as partial action as currently known and emerging they possess any endocrine-disrupting etiologic contributions in a decline noted from future research activities. The develop- activity. If no effect is observed, the in male reproductive health of wildlife. For ment of improved methods should be an significance of the compound (as far as this example, wild panthers in the United States international collaborative effort to obtain context is concerned) is likely to be trivial. have been reported to have an increase in worldwide acceptance oftheir possible use in If effects are observed, more detailed undescended testes and a decrease in semen regulatory safety assessments. studies, including dose-response studies on quality, whereas male alligators in some The exposure to possible xenoestrogens animals, should be undertaken. If the lakes in Florida have been shown to have should be considered in far more detail results ofthese studies raise concern, further abnormalities in their sex hormone levels than currently practiced in such studies. metabolic studies, such as accumulation in (tending toward femaleness) and to have Methods should be elaborated that address the body and access to the fetus, would pro- smaller than normal genitalia. Male fish in the following questions: vide further information and enable a better some parts of the United Kingdom have * What chemicals are man (and wildlife) extrapolation to the human situation. been shown to express a femalelike exposed to, by which routes, and to response when studied in a relatively nat- what degree? Conclusions ural setting. Earlier studies of fish-eating * Do these chemicals actually get absorbed Male reproductive health has received birds in the United States demonstrated (they will not cause effects if they are not remarkably little attention considering nests containing male hatchlings that were absorbed)? that subfertility affects 5% or more of apparently feminized. A recent report of * What are the concentrations of these men and that prostatic hypertrophy or lactating male fruit bats suggested that the chemicals in man (and wildlife)? Do cancer is a major problem for older men. males were, in some way, exposed to a any of them accumulate? It is now evident that several aspects of female sex hormone. Recent laboratory

768 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

studies showed that when estrogenic forms not at all well known for humans, the large and comprising field and clinical studies, of PCBs were painted on turtle eggs, the number of chemicals in numerous environ- epidemiological research, and basic and male hatchlings were sex-reversed to mental categories suggests adequate avail- applied laboratory research with a compre- females. Taken together, this growing body ability. For example, environmental hensive view to the relationships that will of evidence suggests that environmental chemicals reported to be estrogenic evolve. We consider that this approach will factors that resemble female sex hormones include, but are not limited to, some ubiq- optimize the existing strengths of various may be having an adverse effect on the uitous chlorinated hydrocarbons, such as institutions and investigators on a world- reproductive capacity and well being of PCBs and DDT; some products of deter- wide basis and will be the most cost-effec- diverse species. gent and surfactant manufacture, such as tive use of fiscal and intellectual resources. It has been well established that expo- the alkylphenols; and some products There are many research needs in this sure of the male fetus to supranormal levels released from plastics such as bisphenol-A area; a detailed description of these needs of estrogens can result in many, if not all, and some phthalates. Many other com- and strategies to fill them are considered in of the reproductive defects referred to ear- pounds in our natural and synthetic envi- the full report. In addition to key basic, lier. Experiments in which potent synthetic ronment demonstrate estrogenic activities clinical, and epidemiological studies, the estrogens, such as DES, were given to preg- and more are being discovered as the highest priorities go to investigations to fill nant laboratory animals have demonstrated search continues. Although not the subject crucial data gaps necessary to make that prenatal exposure to synthetic estro- of this report, in considering and evaluat- informed decisions about the risk these gens results in a spectrum of adverse effects ing the possible role of estrogenic chemi- chemicals may pose to human health. The on the male offspring including unde- cals in male reproductive disorders, it most pressing areas of need are reliable esti- scended testes, testicular cancer, decreased should not be forgotten that many chemi- mates of exposure of humans to estrogenic semen quality, epididymal cysts, hypospa- cals may have a detrimental effect on male chemicals at different ages including fetal dias, and poor fertility. Men similarly reproductive health through mechanisms life, and of how the exposures relate to exposed in utero to DES have been other than an estrogenic effect. adverse effects on the reproductive system; reported to display related abnormalities The reproductive health trends for use of animal models to identify what lev- such as cryptorchidism, lower sperm men, the emerging data from wildlife, the els of exposure to estrogenic chemicals do, counts, cysts of the epididymis, and ini- well-controlled experimental data with and what levels do not, impair reproduc- tially decreased fertility. The wealth of developmental exposure to estrogens, and tive development; more penetrating studies experimental results and associated clinical the less well-studied, but consistent, data of the molecular events that lead to reports suggests strongly that prenatal on human cohorts, as well as the growing impaired spermatogenesis and testicular exposure to exogenous estrogens may play knowledge concerning hormonally active cancer; improved studies on the structural an etiologic role in the trends observed in environmental chemicals, all point in the and analytical chemistry of chemicals with human male reproductive health. same direction. We conclude that these special emphasis on the prediction of estro- The growing number of reports issues, taken in toto, indicate the need for a genic activity from the molecular structure demonstrating that common environmen- vigorous research effort to understand the of a chemical; and the development of tal contaminants and natural factors pos- extent of the problem, its underlying etiol- rapid in vitro and in vivo test systems for sess estrogenic activity presents the ogy, and the development of a strategy for the detection of estrogenic potency of envi- working hypothesis that the adverse trends prevention and intervention. Since the ronmental chemicals and the harmoniza- in human male reproductive health may health research issues involved are complex tion of these screening tests with improved be, at least in part, associated with expo- and multifactorial and since these issues techniques for quick, reliable analysis of this sure to estrogenic environmental chemicals apparently involve research needs in many chemical class. The proposed high-priority during fetal and childhood development. disciplines, countries and, indeed, species, research will help provide the information The reproductive health trends in men are this report outlines a multinational, inter- necessary for appropriate risk assessments. consistent with this hypothesis. While disciplinary research effort involving acade- exposure levels to estrogenic chemicals are mic, government, and industrial scientists

REFERENCES 1. DEPA. Miloprojekt Nr. 290. Copenhagen:Danish 309:1-2 (1994). Environmental Protection Agency, 1995. 7. Bromwich P, Cohen J, Stewart I, Walker A. Decline in sperm 2. Nelson CMK, Bunge RG. Semen analysis: evidence for chang- counts: an artefact of changed reference range of "normal"? Br ing parameters of male fertility potential. Fertil Steril Med J 309:19-22 (1994). 25:503-507 (1974). 8. Keiding N, Giwercman A, Carlsen E, Skakkebaek NE. Falling 3. Leto S, Frensilli FJ. Changing parameters of donor semen. sperm quality [letter]. Br Med J 309:131 (1994). Fertil Steril 36:766-770 (1981). 9. Skakkeb6k NE, Keiding N. Changes in semen and the testis. 4. Bostofte E, Serup J, Rebbe H. Has the fertility of Danish men Br Med J 309:1316-1317 (1994). declined through the years in terms of semen quality? A com- 10. MacLeod J, Wang Y. Male fertility potential in terms of semen parison of semen qualities between 1952 and 1972. Int J Fertil quality: a review of the past, a study of the present. Fertil Steril 28:91-95 (1983). 31:103-116 (1979). 5. Carlsen E, Giwercman A, Keiding N, Skakkebxk NE. Evidence 11. Auger J, Kunstmann JM, Czyglik F, Jouannet P. Decline in for decreasing quality of semen Turing past 50 years. Br Med J semen quality among fertile men in Paris during the past 20 305:609-613 (1992). years. N Engl J Med 332:281-285 (1995). 6. Farrow S. Falling sperm quality: fact or fiction? Br Med J 12. Irvine S, Cawood E, Richardson D, MacDonald E, Aitken J.

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 769 TOPPARI ET AL.

Evidence of deteriorating semen quality in the United 40. Seddon JM, Savory L, Scott-Conner C. Cryptorchidism: the Kingdom: birth cohort study in 577 men in Scotland over 11 role of medical education in diagnosis. South Med J years. Br Med J 312:467-471 (1996). 78:1201-1204 (1985). 13. Van Waeleghem K, De Clercq N, Vermeulen L, Schoonjans F, 41. Campbell DM, Webb JA, Hargreave TB. Cryptorchidism in Comhaire F. Deterioration of sperm quality in young healthy Scotland. Br Med J 295:1237-1238 (1987). Belgian men. Hum Reprod 11:325-329 (1996). 42. Morley R, Lucas A. Undescended testes in low birthweight 14. Ginsburg J, Hardiman P. Decreasing quality of semen. Br Med infants. Br Med J 295:753 (1987). J 305:1229 (1992). 43. Swerdlow AJ, Melzer D. The value of England and Wales con- 15. Bujan L, Mansar A, Pontonnier F, Mieusset R. Time series genital malformation notification scheme data for epidemiol- analysis of sperm concentration in fertile men in Toulouse, ogy: male genital tract malformations. J Epidemiol Community France between 1977 and 1992. Br Med J 312:471-473 (1996). Health 42:8-13 (1988). 16. Suominen J, Vierula M. Semen quality of Finnish men. Br 44. Choi H, Kim KM, Koh SK, Kim KS, Woo YN, Yoon JB, Choi Med J 306:1579 (1993). SK, Kim SW. A survey of externally recognizable genitourinary 17. Forman D, M0ller H. Testicular cancer. Cancer Surv anomalies in Korean newborns. J Kor Med Sci 4:13-21 (1989). 19/20:323-341 (1994). 45. Benson RCJ, Beard CM, Kelalis PP, Kurland LT. Malignant 18. Nethersell ABW, Drake LK, Sikora K. The increasing inci- potential of the cryptorchid testis. Mayo Clin Proc 66:372-378 dence of testicular cancer in East Anglia. Br J Cancer (1991). 50:377-380 (1984). 46. Correy JF, Newman NM, Collins JA, Burrows EA, Burrows 19. Pike MC, Chilvers CED, Bobrow LG. Classification of testicu- RF, Curran JT. Use of prescription drugs in the first trimester lar cancer in incidence and mortality statistics. Br J Cancer and congenital malformations. Aust N Z J Obstet Gynaec 56:83-85 (1987). 31:340-344 (1991). 20. Boyle P, Kaye SB, Robertson AG. Changes in testicular cancer 47. Ansell PE, Bennett V, Bull D, Jackson MB, Pike LA, Pike MC, in Scotland. Eur J Cancer Clin Oncol 23:827-830 (1987). Chilvers CED, Dudley NE, Gough MH, Griffiths DM, 21. Hakulinen T, Andersen A, Malker B, Pukkala E, Schou G, Redman C, Wilkinson AR, Macfarlane A, Coupland CAC. Tulinius H. Trends in cancer incidence in the Nordic coun- Cryptorchidism: a prospective study of 7500 consecutive male tries. Acta Pathol Microbiol Immunol Scand (Suppl) birt s: 1984-1988. Arch Dis Child 67:892-899 (1992). 288:1-151 (1986). 48. Berkowitz GS, Lapinski RH, Dolgin SE, Gazella JG, Bodian 22. Adami H, Bergstrom R, Mbhner M, Zatonski W, Storm H, CA, Holzman IR. Prevalence and natural history of cryp- Ekbom A, Tretli S, Teppo L, Ziegler H, Rahu M, Gurevicius torchidism. Pediatrics 92:44-49 (1993). R, Stengrevics A. Testicular cancer in nine northern European 49. Williams P. The imperfectly migrated testis. Lancet i:426-427 countries. Int J Cancer 59:33-38 (1994). (1936). 23. Stone JM, Cruickshank DG, Sandeman TF, Matthews JP. 50. Johnson WW. Cryptorchidism. JAMA 113:25-27 (1939). Trebling of the incidence of testicular cancer in Victoria, 51. Baumrucker GO. Incidence of testicular pathology. Bull U S Australia: 1950-1985. Cancer 68:211-219 (1991). Army Med Depart 5:312-314 (1946). 24. Pearce N, Sheppard RA, Howard JK, Fraser J, Lilley BM. Time 52. Ward B, Hunter WM. The absent testicle. Br Med J trends and occupational differences in cancer of the testis in 5179:110-111 (1960). New Zealand. Cancer 59:1677-1682 (1987). 53. Panayotou PC. The incidence of undescended testis in boys 25. Wilkinson TJ, Colls BM, Schluter PJ. Increased incidence of attending elementary schools in Greece. Br J Clin Pract germ cell testicular cancer in New Zealand Maoris. Br J Cancer 19:501-507 (1965). 65:769-771 (1992). 54. Cour-Palais IJ. Spontaneous descent of the testicle. Lancet 26. Sptiz MR, Sider JG, Pollack ES, Lynch HK, Newell GR. i:1403-1405 (1966). Incidence and descriptive features of testicular cancer among 55. Blom K. Undescended testis and the time ofspontanous descent United States whites, blacks, and hispanics: 1973-1982. in 2516 schoolboys. Ugeskr Laeger 146:616-617 (1984). Cancer 58:1785-1790 (1986). 56. Onuora VC, Evbuomwan I. Abnormal findings associated with 27. Harris LE, Steinberg AG. Abnormalities observed during the undescended testis in Nigerian children. Indian J Pediatr first six days of life in 8,716 live-born infants. Pediatrics 56:509-511 (1989). 14:314-326 (1954). 57. Yiicesan S, Dindar H, Olcay I, Okur H, Kilicaslan S, Ergoren 28. McIntosh R, Merritt KK, Richards MR, Samuels MH, Bellows Y, Tiiysiiz C, Koca M, Civilo B, En IS. Prevalence of congeni- MT. The incidence of congenital malformations: a study of tal abnormalities in Turkish school children. Eur J Epidemiol 5,964 pregnancies. Pediatrics 14:505-521 (1954). 9:373-380 (1993). 29. McDonald AD. Maternal health and congenital defect. N Engl 58. Chilvers C, Pike MC, Forman D, Fogelman K, Wadsworth J Med 258:767-773 (1958). MEJ. Apparent doubling of frequency of undescended testis in 30. Pitt DB. A study of congenital malformations. Aust N Z J England and Wales in 1962-1981. Lancet ii:330-332 (1984). Obstet Gynaecol 2:23-30 (1962). 59. Thorup J, Cortes D. The incidence of maldescended testes in 31. Scorer CG. The descent of the testis. Arch Dis Child Denmark. Pediatr Surg Int 5:2-5 (1990). 39:605-609 (1964). 60. Buemann B, Henriksen H, Villumsen AL, Westh A, Zachau- 32. Halevi HS. Congenital malformations in Israel. Br J Prev Soc Christiansen B. Incidence of undescended testis in the new- Med 21:66-77 (1967). born. Acta Chir Scand (Suppl) 283:289-293 (1961). 33. Mital VK, Garg BK. Undescended testicle. Indian J Pediatr 61. Hohlbein R. Migbildungsfrequenz in Dresden. Zentralbl 39:171-174 (1972). Gyniikol 18:719-731 (1959). 34. Heinonen OP, Slone D, Shapiro S. Birth Defects and Drugs in 62. Sweet RA, Schrott HG, Kurland R, Culp OS. Study of the Pregnancy. Littleton:Publishing Sciences Group, 1977. incidence of hypospadias in Rochester, Minnesota, 1940-1970, 35. Mau G, Schnakenburg Kv. Maldescent of the testes-an epi- and a case-control comparison of possible etiologic factors. demiological study. Eur J Pediatr 126:77-84 (1977). Mayo Clin Proc 49:52-58 (1974). 36. Czeizel A, Erodi E, Toth J. An epidemiological study on unde- 63. Czeizel A. Increasing trends in congenital malformations of scended testis. J Urol 126:524-527 (1981). male external genitalia. Lancet i:462-463 (1985). 37. Hirasing RA, Grimberg R, Hirasing HD. De frequentie van 64. Czeizel A, Toth J, Czvenits E. Increased birth prevalence of iso- niet normaal ingedaalde testes bij jonge kinderen. Ned T lated hypospadias in Hungary. Acta Paediatr Hung Geneesk 126:2294-2296 (1982). 27:329-337 (1986). 38. Hsieh J-T, Huang T-S. A study on cryptorchidism. J Formos 65. Kallen B, Winberg J. An epidemiological study of hypospadias Med Assoc 84:953-959 (1985). in Sweden. Acta Paediatr Scand (Suppl) 293:1-21 (1982). 39. Matlai P, Beral V. Trends in congenital malformations of exter- 66. Kallen B, Bertollini R, Castilla E, Czeizel A, Knudsen LB, nal genitalia. Lancet i:108 (1985). Martinez-Frias ML, Mastroiacovo P, Mutchinick 0. A joint

770 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

international study on the epidemiology of hypospadias. Acta o,p'-DDT in the mammalian uterus and avian oviduct. Science Paediatr Scand (Suppl) 324:5-52 (1986). 162:371-372 (1968). 67. World Health Organization. Congenital Malformations 90. Jennings ML, Percival HF, Woodward AR. Evaluation of alli- Worldwide: A Report from the International Clearinghouse for gator hatchlings and egg removal from three Florida lakes. Proc Birth Defects Monitoring Systems. Oxford:Elsevier, 1991. Ann Conf Southeast Assoc Fish Wildl Agencies 42:283-294 68. Bjerkedal T, Bakketeig LS. Surveillance of congenital malfor- (1988). mations and other conditions of the newborn. Int J Epidemiol 91. Guillette LJJ. Endocrine-disrupting environmental contami- 4:31-36 (1975). nants and reproduction: lessons from the study of wildlife. In: 69. Avellan L. The incidence of hypospadias in Sweden. Scand J Women's Health Today: Perspectives on Current Research and Plast Reconstr Surg 9:129-139 (1975). Clinical Practice (Popkin DR, Peddle LJ, eds). New 70. Wallace HM, Baumgartner L, Rich H. Congenital malforma- York:Parthenon, 1994;201-207. tions and birth injuries in New York City. Pediatrics 92. Guillette LJJ, Gross TS, Masson GR, Matter JM, Percival HF, 12:525-535 (1953). Woodward AR. Developmental abnormalities of the gonad and 71. Chung CS, Myrianthopoulos NC, Yoshizaki H. Racial and abnormal sex hormone concentrations in juvenile alligators prenatal factors in major congenital malformations. Am J Hum from contaminated and control lakes in Florida. Environ Genet 20:44-60 (1968). Health Perspect 102:680-688 (1994). 72. Shapiro RN, Eddy W, Fitzgibbon J, O'Brien G. The incidence 93. Guillette LJJ, Gross TS, Gross DA, Rooney AA, Percival HF. of congenital anomalies discovered in the neonatal period. Am Gonadal steroidogenesis in vitro from juvenile alligators J Surg 96:396-400 (1958). obtained from contaminated or control lakes. Environ Health 73. Finley WH, Gustavson K-H, Hall TM, Hurst DC, Bargainer Perspect 103(Suppl 4): 31-36 (1995). CM, Wiedmeyer JA. Birth defects surveillance: Jefferson 94. Bull JJ. Sex determination in reptiles. Q Rev Biol 55:3-21 County, Alabama and Uppsala County, Sweden. South Med J (1980). 87:440-445 (1994). 95. Bull JJ, Gutzke WHN, Crews D. Sex reversal by estradiol in 74. Leung TJ, Baird PA, McGillivray B. Hypospadias in British three reptilian orders. Gen Comp Endocrinol 70:425-428 Columbia. Am J Med Genet 21:39-48 (1985). (1988). 75. Lowry RB, Thunem NY, Silver M. Congenital anomalies in 96. Gutzke WHN, Bull JJ. Steroid hormones reverse sex in turtles. American Indians of British Colombia. Gen Epidemiol Gen Comp Endocrinol 64:368-372 (1986). 3:455-467 (1986). 97. Guillette LJJ, Crain DA, Rooney AA, Pickford DB. 76. Hautau ER. Congenital malformations in infants born to Organization versus activation: the role of endocrine-disrupting Michigan residents in 1958. J Mich State Med Soc contaminants (EDCs) during embryonic development in 59:1833-1836 (1960). wildlife. Environ Health Perspect 103(Suppl 7): 157-164 (1995). 77. Davis DL, Bradlov HL, Wolff M, Woodruff T, Hoel DG, 98. Guillette LJJ, Pickford DB, Crain DA, Rooney AA, Percival Anton-Culver H. Medical hypothesis: xenoestrogens as pre- HF. Reduction in penis size and plasma testosterone concentra- ventable causes of breast cancer. Environ Health Perspect tions in juvenile alligators living in a contaminated environ- 101:372-377 (1993). ment. Gen Comp Endocrinol 101:32-42 (1996). 78. Ewertz M, Holmberg L, Karialainen S, Tretli S, Adami H-O. 99. Bergeron JM, Crews D, McLachlan JA. PCBs as environmental Incidence of male breast cancer in Scandinavia: 1943-1982. estrogens: turtle sex determination as a biomarker ofenvironmen- Int J Cancer 43:27-31 (1989). tal contamination. Environ Health Perspect 102:780-781 (1994). 79. Colborn T, vom Saal FS, Soto AM. Developmental effects of 100. Munkittrick KR, Port CB, Van Der Kraak GJ, Smith IR, endocrine-disrupting chemicals in wildlife and humans. Rokosh DA. Impact of bleached kraft mill effluent on popula- Environ Health Perspect 101:378-384 (1993). tion characteristics, liver MFO activity, and serum steroid levels 80. Colborn T, Clement C, Colborn T, Clement C, eds. of a Lake Superior white sucker (Catostomus commersoni) Chemically-induced Alterations in Sexual and Functional population. Can J Fish Aquat Sci 48:1371-1380 (1991). Development: The Wildlife/Human Connection. Princeton: 101. Andersson T, Forlin L, Harg I, Larsson A. Physiological distur- Princeton Scientific, 1992. bances in fish living in coasta water polluted with bleached kraft 81. Soto AM, Chung KL, Sonnenschein C. The pesticides endosulfan, mill effluents. Can J Fish Aquat Sci 45:1525-1536 (1988). toxaphene, and dieldrin have estrogenic effects on human estro- 102. Howell WM, Black DA, Bortone SA. Abnormal expression of gen-sensitive cells. Environ Health Perspect 102:380-383 (1994). secondary sex characters in a population of mosquitofish, 82. Smith BS. Reproductive anomalies in stenoglossan snails caused Gambusia affinis holbrooki: evidence for environmental-induced by antifouling bottom paints. J Appl Toxicol 1: 15-21 (1981). masculinization. Copeia 4:676-681 (1980). 83. Bryan GW, Gibbs PE, Hummerstone LG, Burt GR. The 103. Howell WM, Denton TE. Gonopodial morphogenesis in decline of the gastropod Nucella lapillus around south-west female mosquitofish, Gambusia affinis affinis, masculinized by England: evidence for the effect of tributyltin from antifouling exposure to degraded products from plant sterols. Environ Biol paints. J Mar Biol Assoc UK 66:611-640 (1986). Fish 24:43-51 (1989). 84. Short JW, Rice SD, Brodersen CC, Stickle WB. Occurrence of 104. Owens JW. The hazard assessment of pulp and paper effluents tri-N-butyltin caused imposex in the North Pacific marine snail in the aquatic environment: a review. Environ Toxicol Chem Nucella lima in Auke Bay, Alaska. Mar Biol 102:291-297 10:1511-1540 (1991). (1989). 105. Leatherland JF. Endocrine and reproductive function in Great 85. Ellis DV, Pattisina LA. Widespread neogastropod imposex: a Lakes salmon. In: Chemically-induced Alterations in Sexual biological indicator of global TBT contamination. Mar Pollut and Functional Development: The Wildlife/Human Bull 21:248-253 (1990). Connection. (Colborn T, Clement C, eds). Adv Mod Environ 86. Fox GA. Epidemiological and pathobiological evidence of con- Toxicol 21:129-145 (1992). taminant-induced alterations in sexual development in free-liv- 106. Freeman HC, Sangalang GB, Flemming B. The sublethal ing wildlife. In: Chemically-induced Alterations in Sexual and effects of polychlorinated biphenyl (Aroclor 1254) diet on the Functional Development: The Wildlife/Human Connection, Atlantic cod Gadus morhua. Sci Total Environ 24:1-11 (1982). Vol 23 (Colborn T, Clement C, eds). Adv Mod Environ 107. Freeman HC, Sangalang GB, Uthe JF. The effects of pollutants Toxicol 21:147-158 (1992) and contaminants on steroidogenesis in fish and marine mam- 87. Evans SM, Hutton A, Kendall MA, Samosir AM. Recovery in mals. Adv Environ Sci Technol 16:197-211 (1984). populations of dogwhelks Nucella lapillus (L) suffering from 108. Turscott B, Walsh J, Burton M, Payne J, Idler D. Effect of imposex. Mar Pollut Bull 22:409-413 (1991). acute exposure to crude oil petroleum on some reproductive 88. Clark DR. Difocol (Kelthane) as an Environmental hormones in salmon and flounder. Comp Biochem Physiol Contaminant. Washington:U.S. Fish and Wildlife Service, 1990. 75C:121-130 (1983). 89. Bitman J, Cecil HC, Harris SJ, Fries GF. Estrogenic activity of 109. Leatherland JF, Sonstegard RA. Bioaccumulation of

Environmental Health Perspectives - Vol 104, Supplement 4 - August 1996 771 TOPPARI ET AL.

organochlorines by yearling coho salmon (Oncorhynchus 131. Berkowitz GD. Abnormalities of gonadal determination and kisutch Walbaum) fed diets containing Great Lakes coho differentiation. Semin Perinatol 16:289-298 (1992). salmon, and the pathophysiological responses of the recipients. 132. Sell S, Pierce B. Maturation arrest of stem cell differentiation is Comp Biochem Physiol 72C:91-99 (1982). a common pathway for the cellular origin of teratocarcinomas 110. Morrison PA, Leatherland JF, Sonstegard RA. Plasma cortisol and epithelial cancers. Lab Invest 70:6-22 (1994). and sex steroid levels in Great Lakes coho salmon 133. Rajpert-De Meyts E, Skakkebxk NE. The possible role of sex (Oncorhynchus kisutch Walbum) in relation to fecundity and egg hormones in the development of testicular cancer. Eur Urol survival. Comp Biochem Physiol 80A:61-68 (1985). 23:54-61 (1993). 111. Jobling S, Sumpter JP. Detergent components in sewage 134. Henderson BE, Bernstein L, Ross RK, Depue RH, Judd HL. effluent are weakly oestrogenic to fish: an in vitro study using The early in utero oestrogen and testosterone environment of rainbow trout (Oncorhynchus mykiss) hepatocytes. Aquatic blacks and whites: potential effects on male offspring. Br J Toxicol 27:361-372 (1993). Cancer 57:216-218 (1988). 112. Purdom CE, Hardiman PA, Bye VJ, Eno NC, Tyler CR, 135. Hutson JM, Ikawa H, Donahoe PK. Estrogen inhibition of Sumpter JP. Estrogenic effects of effluents from sewage treat- Mullerian inhibiting substance in the chick embryo. J Pediatr ment works. Chem Ecol 8:275-285 (1994). Surg 17:953-959 (1982). 113. Bromage NR, Cumaranatunga R. Egg production in the rain- 136. Newbold RR, Suzuki Y, McLachlan JA. Mullerian duct main- bow trout. Rec Adv Aquacul 3:65-138 (1988). tenance in heterotypic organ culture after in vivo exposure to 114. Clemens MJ. Regulation of egg protein synthesis by steroid diethylstilbestrol. Endocrinology 115:1863-1868 (1984). hormones. Prog Biophys Molec Biol 28:71-108 (1978). 137. Kuroda T, Lee MM, Ragin RC, Hirobe S, Donahoe PK. 115. Fry DM, Toone CK. DDT-induced feminization of gull Miillerian inhibiting substance production and cleavage is embryos. Science 213:922-924 (1981). modulated by gonadotropins and steroids. Endocrinology 116. Fry DM, Toone CK, Speich SM, Peard RJ. Sex ratio skew and 129:2985-2992 (1991). breeding patterns of gulls: demographic and toxicological con- 138. Sharpe RM. Regulation of spermatogenesis. In: The Physiology siderations. Stud Avian Biol 10:26-43 (1987). of Reproduction (Knobil E, Neill JD, eds). New York:Raven 117. Hunt GLJ, Hunt MW. Female-female pairing in western gulls Press, 1994; 1-72. (Larus occidentalis) in southern California. Science 139. George FW, Wilson JD. Estrogen formation in the early rabbit 196:1466-1467 (1977). embryos. Science 199:200-201 (1978). 118. Conover MR, Hunt GLJ. Female-female pairing and sex ratios in 140. Dickman Z, Day SK. Proceedings: Two theories: the preim- gulls: a historical perspective. Wilson Bull 96:619-625 (1984). plantation embryo is a source of steroid hormones controlling 119. Subramanian A, Tanabe S, Tatsukawa R, Saito S, Miyazaki N. (1) morula-blastocyst transformation and (2) implantation. J Reduction in the testosterone levels by PCBs and DDE in Reprod Fertil 35:615-617 (1973). Dall's porpoises of the northwestern North Pacific. Mar Pollut 141. Hou Q, Gorski J. Estrogen receptor and progesterone receptor Bull 18:643-646 (1987). genes are expressed differentially in mouse embryos during 120. Freeman HC, Sangalang GB. A study of the effects of methyl preimplantation development. Proc Natl Acad Sci USA mercury, cadmium, arsenic, selenium, and a PCB (Aroclor 90:9460-9464 (1993). 1254) on adrenal and testicular steroidogenesis in vitro, by the 142. Greco TL, Furlow JD, Duello TM, Gorski J. Immunodetection grey seal Halachoerus grypus. Arch Environ Contam Toxicol of estrogen receptors in fetal and neonatal male mouse repro- 5:369-383 (1977). ductive tracts. Endocrinology 130:421-429 (1992). 121. Brouwer A, Reijnders PJH, Koeman JH. Polychlorinated 143. Greco TL, Duello TM, Gorski J. Estrogen receptors, estradiol, biphenyl (PCB)-contaminated fish induces vitamin A and thy- and diethylstilbestrol in early development: the mouse as a roid hormone deficiency in the common seal (Phoca vitulina). model for the study of estrogen receptors and estrogen sensitiv- Aquatic Toxicol 15:99-105 (1989). ity in embryonic development of male and female reproductive 122. Facemire CF, Gross TS, Guillette LJJ. Reproductive impair- tracts. Endocr Rev 14:59-71 (1993). ment in the Florida panther: nature or nurture. Environ Health 144. Guerrier D, Boussin L, Mader S, Josso N, Kahn A, Picard J-Y. Perspect 103(Suppl 4):79-86 (1995). Expression of the gene for anti-Miullerian hormone. J Reprod 123. Roelke ME. Florida Panther Biomedical Investigations: Health Fertil 88:695-706 (1990). and Reproduction. Final Report. Endangered Species Project 145. Jost A, Vigier B, Prepin J, Perchellet JP. Studies on sex differen- E-1II-E-6 7506. Gainsville, FL:Florida Game and Fresh Water tiation in mammals. Recent Prog Horm Res 29:1-41 (1973). Fish Commission, 1990. 146. Lubahn DB, Moyer JS, Golding TS, Couse JF, Korach KS, 124. Bigler WJ, Jenkins JH, Cumbie PM, Hoff GL, Prather EC. Smithies 0. Alteration of reproductive function but not prena- Wildlife and environmental health: raccoons as indicators of tal sexual development after insertional disruption of the mouse zoonoses and pollutants in the Southeastern United States. J estrogen receptor gene. Proc Natl Acad Sci USA Am Vet Med Assoc 167:592-597 (1975). 90:11162-11166 (1993). 125. Pelliniemi LJ, Dym M. The fetal gonad and sexual differentia- 147. Smith EP, Boyd J, Frank GR, Takahashi H, Cohen RM, tion. In: Maternal-Fetal Endocrinology (Tulchinsky D, Little Specker B, Williams TC, Lubahn DB, Korach KS. Estrogen AB, eds). London:W.B.Saunders, 1994;297-320. resistance caused by a mutation in the estrogen-receptor gene in 126. Foster JW, Dominguez-Steglich MA, Guioli S, Kwok C, a man. N EnglJ Med 331:1056-1061 (1994). Weller PA, Stevanovic M, Weissenbach J, Mansour S, Young 148. Davis VL, Couse JF, Goulding EH, Power SGA, Eddy EM, ID, Goodfellow PN, Brook JD, Schafer AJ. Campomelic dys- Korach KS. Aberrant reproductive phenotypes evident in trans- plasia and autosomal sex reversal caused by mutations in an genic mice expressing the wild-type mouse estrogen receptor. SRY-related gene. Nature 372:525-530 (1994). Endocrinology 135:379-386 (1994). 127. Ikeda Y, Shen W-H, Ingraham HA, Parker KL. Developmental 149. Hertz R. The estrogen problem-retrospect and prospect. In: expression of mouse steroidogenic factor-1, an essential regulator Estrogens in the Environment. II: Influences on Development ofthe steroid hydroxylases. Mol Endocrinol 8:654-662 (1994). (McLachlan JA, ed). New York:Elsevier, 1985;1-1 1. 128. Haqq CM, King C-Y, Ukiyama E, Falsafi S, Haqq TN, 150. Jabara AG. Some tissue changes in the dog following stilboe- Donaoe PK, Weiss MA. Molecular basis of mammalian sexual strol administration. Aust J Exp Biol 40:293-308 (1962). determination: activation of Mullerian inhibiting substance 151. Schwartz E, Tornaben JA, Boxill GC. Effects of chronic oral gene expression by SRY. Science 266:1494-1500 (1994). administration of a long-acting estrogen, quinestrol, to dogs. 129. Scully RE. Neoplasia associated with anomalous sexual devel- Toxicol Appl Pharmacol 14:487-494 (1969). opment and abnormal sex chromosomes. Pediatr Adolesc 152. Morrison AS. Cryptorchidism, hernia, and cancer of the testis. Endocrinol 8:203-217 (1981). J Natl Cancer Inst 56:731-733 (1976). 130. Savage MO, Lowe DG. Gonadal neoplasia and abnormal sex- 153. Burns RK. Role of hormones in the differentiation of sex. In: ual differentiation. Clin Endocrinol 32:519-533 (1990). Sex and Internal Secretions (Young WC, Corner GW, eds).

772 Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

Baltimore:Williams and Wilkins, 1961 ;76-160. HW. Association of diethylstilbestrol exposure in utero with 154. Teng CS, Teng CT. Studies on sex-organ development: cryptorchidism, testicular hypoplasia and semen abnormalities. ontogeny of cytoplasmic oestrogen receptor in chick mullerian J Urol 122:36-39 (1979). duct. Biochem J 150:183-190 (1975). 175. Wilcox AJ, Baird DD, Weinberg CR, Hornsby PP, Herbst AL. 155. Wibbels T, Bull JJ, Crews D. Synergism between temperature Fertility in men exposed prenatally to diethylstilbestrol. N Engl and estradiol: a common pathway. J Exp Zool 260:130-134 J Med 332:1411-1416 (1995). (1991). 176. Stenchever MA, Williamson RA, Leonard J, Karp LE, Ley B, 156. Wibbels T, Bull JJ, Crews D. Chronology and morphology of Shy K, Smith D. Possible relationship between in utero diethyl- temperature-dependent sex determination. J Exp Zool stilbestrol exposure and male fertility. Am J Obstet Gynecol 260:371-381 (1991). 140:186-193 (1981). 157. MacLaughlin DT, Hutson JM, Donahoe PK. Specific estradiol 177. Whitehead ED, Leiter E. Genital abnormalities and abnormal binding in embryonic mullerian ducts: a potential modulator of semen analyses in male patients exposed to diethylstilbestrol in regression in the male and female chick. Endocrinology utero. J Urol 125:47-50 (1981). 113:141-145 (1983). 178. Driscoll SG, Taylor SH. Effects of prenatal maternal estrogen on 158. Teng CS. Quantitative change in fibronectin in cultured mul- the male urogenital system. Obstet Gynecol 56:537-542 (1980). lerian mesenchymal cells in response to diethylstilbestrol and 179. Raman-Wilms L, Tseng AL-i, Wighardt S, Einarson TR, mullerian-inhibiting substance. Dev Biol 140:1-7 (1990). Koren G. Fetal genital effects of first-trimester sex hormone 159. Orth JM, Gunsalus GL, Lamperti AA. Evidence from Sertoli exposure: a metaanalysis. Obstet Gynecol 85:141-149 (1995). cell-depleted rats indicates that spermatid number in adults 180. Beard M, Melton LJ, O'Fallon WM, Noller KL, Benson RC. depends on numbers of Sertoli cells produced during perinatal Cryptorchism and maternal estrogen exposure. Am J Epidemiol development. Endocrinology 122:787-794 (1988). 120:707-716 (1984). 160. Russell LD, Peterson RN. Determination of the elongate sper- 181. Gill WB, Schumacher GFB, Bibbo M. Genital and semen matid-Sertoli cell ratio in various mammals. J Reprod Fertil abnormalities in adults males two and one-half decades after in 70:635-641 (1984). utero exposure to diethylstilbestrol. In: Intrauterine Exposure to 161. Orth J. Proliferation of Sertoli cells in fetal and postnatal rats: a Diethytstilbestrol in the Human (Herbst AL, ed). Chicago: quantitative autoradiographic study. Anat Rec 203:485-492 American College of Obstetricians and Gynecologists, 1978. (1982). 182. Schumacher GFB, Gill WB, Hubby MM, Bluogh RR. Semen 162. Orth JM. The role of follicle-stimulating hormone in control- analysis in males exposed in utero to diethylstilbestrol (DES) or ling Sertoli cell proliferation in testes of fetal rats. placebo. Reprod Obstet Gynecol 9:100-101 (1981). Endocrinology 115:1248-1255 (1984). 183. Andonian RW, Kessler R. Transplacental exposure to diethyl- 163. Rey RA, Campo SM, Bedecarras P, Nagle CA, Chemes HE. Is stilbestrol in men. Urology 13:276-279 (1976). infancy a quiescent period of testicular development? 184. Vessey MP. Epidemiological studies of the effects of diethyl- Histological, morphometric, and functional study of the semi- stilbestrol. IARC Sci Publ 96:335-348 (1989). niferous tubules of the Cebus monkey from birth to the end of 185. Conley GR, Sant GR, Ucci AA, Mitcheson HD. Seminoma puberty. J Clin Endocrinol Metab 76:1325-1331 (1993). and epididymal cysts in a young man with known diethylstilbe- 164. Arai Y, Mori T, Suzuki Y, Bern HA. Long-term effects of peri- strol exposure in utero. JAMA 249:1325-1326 (1983). natal exposure to sex steroids and diethylstilbestrol on the 186. Henderson BE, Benton B, Jing J, Yu MC, Pike MC. Risk fac- reproductive system of male mammals. Int Rev Cytol tors for cancer of the testis in young men. Int J Cancer 84:235-268 (1983). 23:598-602 (1979). 165. Thomas GB, McNeilly AS, Gibson F, Brooks AN. Effect of 187. Depue RH, Pike MC, Henderson BE. Estrogen exposure dur- pituitary-gonadal suppression with a gonadotrophin-releasing ing gestation and risk of testicular cancer. J Natl Cancer Inst hormone agonist on fetal gonadotrophin secretion, fetal 71:1151-1155 (1983). gonadal development and maternal steroid secretion in the 188. Schottenfeld D, Warshauer ME, Sherlock S, Zauber AG, Leder sheep. J Endocrinol 141:317-324 (1994). M, Payne R. The epidemiology of testicular cancer in young 166. Palmlund I, Apfel R, Buitendijk S, Cabau A, Forsberg J-G. adults. Am J Epidemiol 112:232-246 (1980). Effects of diethylstilbestrol (DES) medication during pregnancy: 189. Brown LM, Pottern LM, Hoover RN. Prenatal and perinatal risk report from a symposium at the 10th International Congress of factors for testicular cancer. Cancer Res 46:4812-4816 (1986). ISPOG. J Psychosom Obstet Gynaecol 14:71-89 (1993). 190. Moss AR, Osmond D, Bacchetti P, Torti FM, Gurgin V. 167. Dieckmann WJ, Davis ME, Rynkiewicz LM, Pottinger RE. Hormonal risk factors in testicular cancer. A case-control study. Does administration of diethylstilbestrol during pregnancy have Am J Epidemiol 124:39-52 (1986). therapeutic value? Am J Obstet Gynecol 66:1062-1081 (1953). 191. Gershman ST, Stolley PD. A case-control study of testicular 168. Brackbill Y, Berendes HW. Dangers of diethylstilbestrol: cancer using Connecticut tumour registry data. Int J Epidemiol review of a 1953 paper. Lancet 2:520 (1978). 17:738-742 (1988). 169. Herbst AL, Scully RE. Adenocarcinoma of the vagina in adoles- 192. McLachlan JA. Rodent models for perinatal exposure to cence: a report of seven cases including six clear cell carcinomas diethylstilbestrol and their relation to human disease in the (so-called mesonephromas). Cancer 25:745-757 (1970). male. In: Developmental Effects of Diethylstilbestrol (DES) in 170. Herbst AL, Ulfelder H, Poskanzer DC. Adenocarcinoma of the Pregnancy (Herbst AL, Bern HA, eds). New York: vagina. Association of maternal stilbestrol therapy with tumour Thieme-Stratton, 1981; 148-157. appearance in young women. N Engl J Med 284:878-881 193. Newbold RR, McLachlan JA. Diethylstilbestrol associated (1971). defects in murine genital tract development. In: Estrogens in 171. Stillman RJ. In utero exposure to diethylstilbestrol: adverse the Environment. II: Influences on Development (McLachlan effects on the reproductive tract and reproductive performance JA, ed). New York:Elsevier, 1985;288-318. in male and female offspring. Am J Obstet Gynecol 194. Song W, Moore R, McLachlan JA, Negishi M. Molecular charac- 142:905-921 (1982). terization of a testis-specific estrogen sulfotransferase and aberrant 172. Henderson BE, Benton B, Cosgrove M, Baptista J, Aldrich J, liver expression in obese and diabetogenic C57BL/KsJ-db/db Townsend D, Hart W, Mack TM. Urogenital tract abnormali- mice. Endocrinology 136:2477-2484 (1995). ties in sons of women treated with diethylstilbestrol. Pediatrics 195. Brown-Grant K, Fink G, Greig F, Murray MAF. Altered sexual 58:505-507 (1976). development in male rats after oestrogen administration during 173. Gill WB, Schumacher GFB, Bibbo M. Pathological semen and the neonatal period. J Reprod Fertil 44:25-42 (1975). anatomical abnormalities of the genital tract in human male 196. Ohta Y. Response of testis to androgen and gonadotropins in subjects exposed to diethylstilbestrol in utero. J Urol neonatally estrogenized and androgenized mice. Endocrinol 117:477-480 (1977). Jpn 24:287-294 (1977). 174. Gill WB, Schumacher GFB, Bibbo M, Straus FHI, Schoenberg 197. Walker AH, Bernstein L, Warren DW, Warner NE, Zheng X,

Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 773 TOPPARI ET AL.

Henderson BE. The effect of in utero ethinyl oestradiol expo- 218. Gellert RJ, Wilson C. Reproductive function in rats exposed sure on the risk of cryptorchid testis and testicular teratoma in prenatally to pesticides and polychlorinated biphenyls (PCB). mice. Br J Cancer 62:599-602 (1990). Environ Res 18:437-443 (1979). 198. Newbold RR, Bullock BC, McLachlan JA. Adenocarcinoma of 219. Gray LEJ, Ostby J, Ferrell J, Rehnberg G, Linder R, Cooper R, the rete testis. Diethylstilbestrol-induced lesions of the mouse Goldman J, Slott V, Laskey J. A dose-response analysis of rete testis. Am J Pathol 125:625-628 (1986). methoxychlor-induced alterations of reproductive development 199. Sharpe RM. Falling sperm counts in men-is there an and function in the rat. Fundam Appl Toxicol 12:92-108 endocrine cause? J Endocrinol 137:357-360 (1993). (1989). 200. Sharpe RM, Fisher JS, Millar MM, Jobling S, Sumpter JP. 220. Gray LEJ. Chemical-induced alterations of sexual differentia- Gestational and lactational exposure of rats to xenoestrogens tion: a review of effects in humans and rodents. In: Chemically results in reduced testicular size and sperm production. Environ induced Alterations in Sexual and Functional Development: Health Perspect 103:1136-1143 (1995). The Wildlife/Human Connection (Colborn T, Clement C, 201. Sharman M, Read WA, Castle L, Gilbert J. Levels of di-(2-eth- eds). Adv Mod Environ Toxicol 21:203-230 (1992). ylhexyl)phthalate and total phthalate esters in milk, cream, but- 221. Haake J, Kelley M, Keys B, Safe S. The effects of organochlo- ter and cheese. Food Add Contam 11:375-385 (1994). rine pesticides as inducers of testosterone and benzo[a]pyrene 202. Soto AM, Justicia H, Wray JW, Sonnenschein C. p- hydroxylases. Gen Pharmac 18:165-169 (1987). Nonylphenol an estrogenic xenobiotic released from "modified" 222. Balash KJ, Al-Omar MA, Latif BMA. Effect of chlordane on polystyrene. Environ Health Perspect 92:167-173 (1991). testicular tissues of Swiss mice. Bull Environ Contam Toxicol 203. Jobling S, Reynolds T, White R, Parker MG, Sumpter JP. A 39:434-442 (1987). variety of environmentally persistent chemicals, including some 223. ATSDR. Toxicological Profile for Aldrin/Dieldrin. phthalate plasticizers, are weakly estrogenic. Environ Health Atlanta:Association for Toxic Substances and Drug Research, Perspect 103:582-587 (1995). 1987. 204. Soto AM, Sonnenschein C, Chung KL, Fernandez MF, Olea 224. Arnold DL, Moodie CA, Charbonneau SM, Grice HC, N, Olea Serrano F. The E-SCREEN assay as a tool to identify McGuire PF, Bryce FR, Collins BT, Zawidzka ZZ, Krewski estrogens: an update on estrogenic environmental pollutants. DR, Nera EA, Munro IC. Long-term toxicity of hexa- Environ Health Perspect 103(Suppl 3):113-122 (1995). chlorobenzene in the rat and the effect of dietary vitamin A. 205. McDonnell DP, Clemm DL, Hermann T, Goldman ME, Pike Food Chem Toxicol 23:779-793 (1985). JW. Analysis of Estrogen receptor function in vitro reveals three 225. Jensen AA, Slorach SA. Chemical Contaminants in Human distinct classes of antiestrogens. Mol Endocrinol 9:659-669 Milk. Boston:CRC Press, 1991. (1995). 226. Cooper RL, Chadwick RW, Rehnberg GL, Goldman JM, 206. Thomas KB, Colborn T. Organochlorine endocrine disrupters Booth KC, Hein JF, McElroy WK. Effect of lindane on hor- in human tissue. In: Chemically-induced Alterations in Sexual monal control of reproductive function in the female rat. and Functional Development: The Wildlife/Human Toxicol Appl Pharmacol 99:384-394 (1989). Connection (Colborn T, Clement C, eds). Adv Mod Environ 227. Van Velsen FL, Danse LHJC, Van Leeuwen FXR, Dormans Toxicol 21:365-394 (1992). JAMA, Van Logten MJ. The subchronic oral toxicity of the B- 207. Calabrese EJ. Human breast milk contamination in the United isomer of hexachlorocyclohexane in rats. Fundam Appl Toxicol States and Canada by chlorinated hydrocarbon insecticides and 6:697-712 (1986). industrial pollutants: current status. J Am College Toxicol 228. Gray LEJ. Neonatal chlordecone exposure alters behavioral sex 1:91-98 (1982). differentiation in female hamsters. Neuroendocrinol 3:67-80 208. Norstrom RJ, Simon M, Muir DCG, Schweinsburg RE. (1982). Organochlorine contaminants in arctic marine food chains: 229. Hammond B, Katzenellenbogen B, Krauthammer N, identification, geographical distribution, and temporal trends McConnel J. Estrogenic activity of the insecticide chlordecone in polar bears. Environ Sci Technol 22:1063-1071 (1988). (Kepone) and interaction with uterine estrogen receptors. Proc 209. Bitman J, Cecil HC. Estrogenic activity of DDT analogs and Natl Acad Sci USA 76:6641-6645 (1979). polychlorinated biphenyls. Agr Food Chem 18:1108-1112 230. Palmiter RD, Mulvihill ER. Estrogenic activity of the insecti- (1970). cide kepone on the chicken oviduct. Science 201:356-358 210. Lamont TG, Bagley GE, Reichel WL. Residues of o,p'-DDD (1978). and o,p'-DDT in brown pelican eggs and mallard ducks. Bull 231. Safe SH. Polychlorinated biphenyls (PCBs), dibenzo-p-dioxins Environ Contam Toxicol 5:231-236 (1970). (PCDDs), dibenzofurans (PCDFs), and related compounds: 211. Bustos S, Denegri JC, Diaz F, Tchernitchin AN. p,p'-DDT is environmental and mechanistic considerations which support an estrogenic compound. Bull Environ Contam Toxicol the development of toxic equivalency factors (TEFs). Crit Rev 41:496-501 (1988). Toxicol 21:51-88 (1990). 212. Soto AM, Lin T-M, Justicia H, Silvia RM, Sonnenschein C. 232. Ahlborg UG, Hanberg A, Kenne K. Risk assessment ofpolychlo- An "in culture" bioassay to assess the estrogenicity of xenobi- rinated biphenyls (PCBs). Nord 1992:26. Copenhagen:Nordic otics (E-SCREEN). In: Chemically-induced Alterations in Council ofMinisters, 1992. Sexual and Functional Development: The Wildlife/Human 233. Krishnan V, Safe S. Polychlorinated biphenyls (PCBs), Connection (Colborn T, Clement C, eds). Adv Mod Environ dibenzo-p-dioxins (PCDDs), and dibenzofurans (PCDFs) as Toxicol 21:295-309 (1992). antiestrogens in MCF-7 human breast cancer cells: quantitative 213. Sharpe RM. Another DDT connection. Nature 375:538-539 structure-activity relationships. Toxicol Appl Pharmacol (1995). 120:55-61 (1993). 214. Kelce WR, Stone CR, Laws SC, Gray LEJ, Kemppainen JA, 234. Lilienthal H, Winneke G. Sensitive periods for behavioral toxi- Wilson EM. Persistent DDT metabolite, p,p'-DDE is a potent city of polychlorinated biphenyls: determination by cross-fos- androgen receptor antagonist. Nature 375:581-585 (1995). tering in rats. Fundam Appl Toxicol 17:368-375 (1991). 215. Smith AG. Chlorinated hydrocarbon insecticides. In: 235. Korach KS, Sarver P, Chae K, McLachlan JA, McKinney JD. Handbook of Pesticide Toxicology (Hayes WJ, Laws ER, eds). Estrogen receptor-binding activity of polychlorinated hydro- New York:Academic Press, 1991;731-915. biphenyls conformationally restricted structural probes. Mol 216. Bulger WH, Muccitelli RM, Kupfer K. Studies on the in vivo Pharmacol 33:130-126 (1988). and in vitro estrogenic activities of methoxychlor and its 236. Reijnders PJH. Reproductive failure in common seals feeding on metabolites. Role of hepatic mono-oxygenase in methoxychlor fish from polluted coastal waters. Nature 324:456-457 (1986). activation. Biochem Pharmacol 27:2417-2423 (1978). 237. Safe SH. Comparative toxicology and mechanism of action of 217. Gellert RJ, Heinrichs WL, Swerdloff R. Effects of neonatally polychlorinated dibenzo-p-dioxins and dibenzofurans. Annu administered DDT homologs on reproductive function in male Rev Pharmacol Toxicol 26:371-399 (1986). and female rats. Neuroendocrinology 16:84-94 (1974). 238. Skene SA, Dewhurst IC, Greenberg M. Polychlorinated

774 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

dibenzo-p-dioxins and polychlorinated dibenzofurans: the risk lipophilic metabolites of nonionic surfactants in freshwater to human health. A review. Hum Toxicol 8:173-203 (1989). organisms. Environ Pollut 79:243-248 (1993). 239. Peterson RE, Theobald HM, Kimmel GL. Developmental and 257. White R, Jobling S, Hoare SA, Sumpter JP, Parker MG. reproductive toxicity of dioxins and related compounds: cross- Environmentally persistent alkylphenolic compounds are estro- species comparisons. Crit Rev Toxicol 23:283-335 (1993). genic. Endocrinology 135:175-182 (1994). 240. Safe S, Astroff B, Harris M, Zacharewski T, Romkes M, Biegel 258. Verdeal K, Ryan S. Naturally occurring estrogens in plant L. 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and related foodstuff-a review. J Food Prot 42:577-583 (1979). compounds as antioestrogens: characterization and mechanism 259. Bradbury RB, White DE. Oestrogens and related substances in of action. Pharmacol Toxicol 69:400-409 (1991). plants. Vitamins, Hormones 12:207-233 (1954). 241. Lucier GW, Portier CJ, Gallo MA. Receptor mechanisms and 260. Adlercreutz H, van der Wildt J, Kinzel J, Attalla H, Wahala K, dose-response models for the effects of dioxins. Environ Health Makela T, Hase T, Fotsis T. Lignan and isoflavonoid conju- Perspect 101:36-44 (1993). gates in human urine. J Steroid Biochem 52:97-103 (1995). 242. Peterson RE, Moore RW, Mably TA, Bjerke DL, Goy RW. 261. Verdeal K, Ryan DS. Naturally-occurring oestrogens in plant Male reproductive system ontogeny: effect of perinatal exposure foodstuffs-a review. J Food Protect 7:577-583 (1979). to 2,3,7,8-tetrachlorodibenzo-p-dioxin. In: Chemically-induced 262. Price KR, Fenwick GR. Naturally occurring oestrogens in Alterations in Sexual and Functional Development: The foods-a review. Food Add Contam 2:73-106 (1985). Wildlife/Human Connection (Colborn T, Clement C, eds). 263. Kaldas RS, Hughes GL. Reproductive and general metabolic Adv Mod Environ Toxicol 21:175-94 (1992). effects of phytoestrogens in mammals. Reprod Toxicol Rev 243. Mably TA, Moore RW, Bjerke DL, Peterson RE. The male 3:81-89 (1989). reproductive system is highly sensitive to in utero and lacta- 264. Shutt DA. The effects of plant estrogens on animal reproduc- tional TCDD exposure. In: Biological Basis for Risk tion. Endeavour 35:110-113 (1976). Assessment of Dioxins and Related Compounds (Gallo MA, 265. Setchell KDR, Borriello SP, Hulme P, Krik DN, Axelson M. Scheuplein RJ, van der Heijden CA, eds). New York:Cold Nonsteroidal estrogens of dietary origin: possible roles in hor- Spring Harbor Laboratory Press, 199 1;69-78. mone-dependent disease. Am J Clin Nutr 40:569-578 (1992). 244. Mably TA, Bjerke DL, Moore RW, Gendron-Fitzpatrick A, 266. Cassidy A, Bingham S, Setchell K. Biological effects of a diet of Peterson RE. In utero and lactational exposure of male rats to soy protein rich in isoflavones on the menstrual cycle of pre- 2,3,7,8-tetrachlorodibenzo-p-dioxin. 3: Effects on spermatoge- menopausal women. Am J Clin Nutr 60:333-340 (1994). nesis and reproductive capability. Toxicol Appl Pharmacol 267. Krishnan AV, Stathis P, Permuth SF, Tokes L, Feldman D. 114:118-126 (1992). Bisphenol-A: an estrogenic substance is released from polycar- 245. Mably TA, Moore RW, Goy RW, Peterson RE. In utero and bonate flasks during autoclaving. Endocrinology lactational exposure of male rats to 2,3,7,8-tetrachlorodibenzo- 132:2279-2286 (1993). pdioxin. 2: Effects on sexual behaviour and the regulation of 268. Brotons JA, Olea-Serrano MF, Villalobos M, Pedraza V, Olea uteinizing hormone secretion in adulthood. Toxicol Appl N. Xenoestrogens released from lacquer coatings in food cans. Pharmacol 114:108-117 (1992). Environ Health Perspect 103:608-612 (1995). 246. Mably TA, Moore RW, Peterson RE. In utero and lactational 269. Hansen E, Meyer 0. Animal models in reproductive toxicol- exposure of male rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin. 1. ogy. In: Handbook of Laboratory Animals Science. Vol II: Effects on androgenic status. Toxicol Appl Pharmacol Animal Models (Svendsen P, Hau J, eds). Boca Raton, 114:97-107 (1992). FL:CRC Press, 1994. 247. Bowman RE, Schantz SL, Weerasinghe NCA, Gross ML, 270. McLachlan JA, Korach KS, Newbold RR, Degen GH. Barsotti DA. Chronic dietary intake of 2,3,7,8-tetra- Diethylstilbestrol and other estrogens in the environment. chlorodibenzo-p-dioxin (TCDD) at 5 or 25 parts per trillion in Fundam Appl Toxicol 4:686-691 (1984). the monkey: TCDD kinetics and dose-effect estimate of repro- 271. Grosvenor CE, Picciano MF, Baumrucker CR. Hormones and ductive toxicity. Chemosphere 18:243-252 (1989). growth factors in milk. Endocr Rev 14:710-728 (1992). 248. ECETOC. Exposure of Man to Dioxins: A Perspective on 272. Shore LS, Gurevitz M, Shemesh M. Estrogen as an environmen- Industrial Waste Incineration. Technical Report No 49. tal pollutant. Bull Environ Contam Toxicol 51:361-366 (1993). Brussels:European Centre for Ecotoxicology and Toxicology of 273. Guzelian PS. Comparative toxicology of chlordecone (Kepone) Chemicals, 1992. in humans and experimental animals. Annu Rev Pharmacol 249. Ahel M, Conrad J, Giger W. Persistent organic chemicals in Toxicol 22:89-113 (1982). sewage effluents. 3: Determination of nonylphenoxy carboxylic 274. Fein GG, Jacobson JL, Jacobson SW, Schwartz PM, Dowler acids by high resolution gas chromatography. Environ Sci JK. Prenatal exposure to polychlorinated biphenyls. Effects on Technol 21:697, (1987). birth size and gestational age. J Pediatr 105:315-320 (1984). 250. Ahel M, Giger W, Koch M. Behaviour of alkylphenol poly- 275. Jacobson SW, Fein GG, Jacobson JL, Schwarz PM, Dowler JK. ethoxylate surfactants in the aquatic environment. I: The effects of PCB exposure on visual recognition memory. Occurrence and transformation in sewage treatment. Water Res Child Dev 56:853-860 (1985). 28:1131-1142 (1994). 276. Jacobson JL, Jacobson SW. New methodologies for assessing 251. Ahel M, Giger W, Schaffner C. Behaviour of alkylphenol poly- the effects of prenatal toxic exposure on cognitive functioning ethoxylate surfactants in the aquatic environment. II: in humans. In: Toxic Contaminants and Ecosystem Health: A Occurrence and transformation in rivers. Water Res Great Lakes Focus (Evans M, ed). New York:John Wiley and 28:1143-1152 (1994). Sons, 1988;374-388. 252. Giger W, Brunner PH, Schaffner C. 4-Nonylphenol in sewage 277. Jacobson JL, Jacobson SW, Humphrey HEB. Effects of expo- sludge: accumulation of toxic from nonionic surfactants. sure to PCBs and related compounds on growth and activity in Science 225:623-625 (1984). children. Neurotox Terat 12:319-326 (1990). 253. Holt MS, Mitchell GC, Watkinson RJ. The environmental 278. Jacobson JL, Jacobson SW, Humprey HEB. Effects of in utero chemistry, fate, and effects of nonionic surfactants. In: The exposure to polychlorinated biphenyls and other contaminants Handbook of Environmental Chemistry. Vol 3, Part F on cognitive functioning in young children. J Pediatr (Hutzinger 0, ed). Berlin:Springer-Verlag, 1992; 89-144. 116:38-45 (1990). 254. Naylor GC, Mierure JP, Weeks JA, Castaldi RJ, Romano RR. 279. Bush B, Bennett A, Snow J. Polychlorinated biphenyl con- Alkylphenol ethoxylates in the environment. J Am Oil Chem geners, p,p'-DDE, and sperm function in humans. Arch Soc 69:695-703 (1992). Environ Contam Toxicol 15:333-341 (1986). 255. Ekelund R, Bergman A, Granmo A, Berggren M. 280. Lione A. Polychlorinated biphenyls and reproduction. Reprod Bioaccumulation of 4-nonylphenol in marine animals. A re- Toxicol Rev 2:83-89 (1988). evaluation. Environ Pollut 64:107-120 (1990). 281. Guo YL, Lai TJ, Ju SH, Chen YC, Hsu CC. Sexual develop- 256. Ahel M, McEvoy J, Giger W. Bioaccumulation of the ments in biological findings in Yucheng children. DIOXIN

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 775 TOPPARI ET AL.

'93: 13th International Symposium on Chlorinated Dioxins Estrogen levels in childhood determined by an ultrasensitive and Related Compounds 14:235-238 (1993). recombinant cell bioassay. J Clin Invest 94:2475-2480 (1994). 282. Jensen AA. PCBs, PDDs, PCDFs in human milk and blood 304. De Peretti E, Forest MG. Unconjugated dehydroepiandtros- and adipose tissue. Sci Tot Environ 64:259-293 (1987). terone plasma levels in normal subjects from birth to adoles- 283. Kimbrough RD. How toxic is 2,3,7,8-tetrachlorodibenzodioxin cence in human: the use of a sensitive radioimmunoassay. J to humans? J Toxicol Environ Health 30:261-271 (1990). Clin Endocrinol Metab 43:982 (1976). 284. Bertazzi PA, Zochetti C, Pesatori AC, Guercilena S, Sanarico 305. Korenman SG, Stevens RH, Carpenter LA, Robb M, M, Radice L. Ten-year mortality study of the population Niswender GD, Sherman BM. Estradiol radioimmunoassay involved in the Seveso incident in 1976. Am J Epidemiol without chromatography: procedure, validation and normal 129:1187-1200 (1989). values. J Clin Endocrinol Metab 38:718-720 (1974). 285. WHO. Polychlorinated Dibenzo-para-dioxins and 306. Legan SJ, Karsch FJ, Foster DL. The endocrine control of sea- Dibenzofurans. Environ Health Criteria 88. Geneva:World sonal reproductive function in ewe: a marked change in Health Organization, 1989. response to the negative feedback action of estradiol on luteiniz- 286. Clark LB, Rosen RT, Hartman TG, Louis JB, Suffet IH, ing hormone secretion. Endocrinology 101:818-824 (1977). Lippincott RL, Rosen JD. Determination of alkylphenol 307. Soto AM, Sonnenschein C. The role of estrogens on the prolif- ethoxylates and their acetic acid derivatives in drinking water eration of human breast cancer cells (MCF-7). J Steroid by particle beam liquid chromatography/mass spectrometry. Int Biochem 23:87-94 (1985). J Environ Anal Chem 47:167-180 (1992). 308. Thomas JA. Toxic responses of the reproductive system. In: 287. Cassidy A, Bingham S, Carlson J, Setchell KDR. Biological Casarett and Doull's Toxicology: The Basic Science of Poisons effects of plant estrogens in premenopausal women. FASEB J (Amdur MO, Doull J, Klaassen CD, eds). New York: 7:A866 (1993). Pergamon Press, 1991;484-520. 288. Baird DD, Umbach DM, Lansdell L, Hughes CL, Setchell 309. Matzuk MM, Finegold MJ, Su J-GJ, Hsueh AJW, Bradley A. KDR, Weinberg CR, Haney AF, Wilcox AJ, McLachlan JA. a-Inhibin is a tumour-suppressor gene with gonadal specificity Dietary intervention study to assess estrogenicity of dietary soy in mice. Nature 360:313-319 (1992). among postmenopausal women. J Clin Endocrinol Metab 310. Hofmann MC, Narisawa S, Hess RA, Millan JL. 80:1685-1690 (1995). Immortalization of germ cells and somatic testicular cells using 289. Wilcox G, Wahlqvist ML, Burger HG, Medley G. Oestrogenic the SV40 large antigen. Exp Cell Res 201:417-435 (1992). effects of plant foods in postmenopausal women. Br Med J 311. Hofmann MC, Hess RA, Goldberg E, Millan JL. Immortalized 301:905-906 (1990). germ cells undergo meiosis in vitro. Proc Natl Acad Sci USA 290. Clarkson TB, Anthony MS, Hughes CLJ. Estrogenic soybean 91:5533-5537 (1994). isoflavones and chronic disease: risks and benefits. Trends 312. Brinster RL, Avarbock MR. Germline transmission of donor Endocrinol Metab 6:11-19 (1995). haplotype following spermatogonial transplantation. Proc Natl 291. Bibbo M, Gill WB, Azizi F, Bluogh R, Fang VS, Rosenfield Acad Sci USA 91:11303-11307 (1994). RL, Schumacher GFB, Sleeper K, Sonek MG, Wied GL. 313. Brinster RL, Zimmermann JW. Spermatogenesis following Follow-up study of male and female offspring of DES-exposed male germ-cell transplantation. Proc Natl Acad Sci USA mothers. Obstet Gynecol 49:1-8 (1977). 91:11298-11302 (1994). 292. Adlercreutz H. Diet, breast cancer and sex hormone metabo- 314. Toppari J, Eerola E, Parvinen M. Flow cytometric DNA analy- lism. Ann N YAcad Sci 595:281-290 (1990). sis of defined stages of rat seminiferous epithelial cycle during 293. de Cock J, Westveer K, Heederik D, te Velde E, van Kooij R. in vitro differentiation. J Androl 6:325-333 (1985). Time to pregnancy and occupational exposure to pesticides in 315. Toppari J, Parvinen M. In vitro differentiation of rat seminifer- fruit growers in The Netherlands. Occup Environ Med ous tubular segments from defined stages of the epithelial cycle: 51:693-699 (1994). morphologic and immunolocalization analysis. J Androl 294. Fox GA. Practical causal inference for ecoepidemiologists. J 6:334-343 (1985). Toxicol Environ Health 33:359-373 (1991). 316. Toppari, J. Rat spermatogenesis in vitro. Studies on differentia- 295. Lan NC, Katzenellenbogen BS. Temporal relationships tion of seminiferous tubule segments at defined stages of the between hormone receptor binding and biological responses in epithelial cycle. PhD dissertation:University of Turku, Turku, the uterus: studies with short- and long-acting derivatives of Finland 1986. estradiol. Endocrinology 98:220-227 (1976). 317. Dodds EC, Lawson W. Molecular structure in relation to oestro- 296. Moore HDM, Thurstan SM. Sexual differentiation in the gray- genic activity. Compounds without phenanthrene nucleus. Proc tailed opossum, Monodelphis domestica, and the effects of oestra- Royal Soc London 118(Series B):222-232 (1938). diol on testis development. J Zool 221:639-658 (1990). 318. UICC. Cancer Incidence in Five Continents, Vol 1. A 297. Sumpter JP, Jobling S. Vitellogenesis as a biomarker for estro- Technical Report (Doll R, Payne P, Waterhouse J, eds). genic contamination of the aquatic environment. Environ Geneva:International Union Against Cancer (Distributed by Health Perspect 103(Suppl 7):173-178 (1995). Springer Verlag, Berlin), 1966. 298. Copeland PA, Sumpter JP, Walker JP, Croft M. Vitellogenin 319. UICC. Cancer Incidence in Five Continents, Vol 2 (Doll R, levels in male and female rainbow trout (Salmo gairdneri) at Muir CS, Waterhouse JAH, eds). Geneva:International Union various stages of the reproductive cycle. Comp Biochem Against Cancer, 1970. Physiol 83B:487-493 (1986). 320. IARC. Cancer Incidence in Five Continents, Vol 3 299. Bulger WH, Kupfer D. Estrogenic action of DDT analogs. Am (Waterhouse J, Muir C, Correa P, Powell J, eds). IARC J Ind Med 4:163-173 (1983). Scientific Publ No 15. Lyon:International Agency for Research 300. Maitre JI, Mercier L, Dolo L, Valotaire Y. Characterization of on Cancer, 1976. estradiol specific receptors and induction of vitellogenin and 321. IARC. Cancer Incidence in Five Continents, Vol 4 vitellogenin mRNA in the rainbow trout liver (Salmo gairneri). (Waterhouse J, Muir C, Shanmugaratnam K, Powell J, eds). Biochimie 67:215-225 (1985). IARC Scientific Publ No 42. Lyon:International Agency for 301. Sumpter JP. The purification, radioimmunoassay and plasma Research on Cancer, 1982. levels of vitellogenin from the rainbow trout (Salmo gairdnerz). 322. IARC. Cancer Incidence in Five Continents, Vol 5 (Muir C, In: Trends in Comparative Endocrinology (Lofts B, Holmes Waterhouse J, Mack T, Powell J, Whelan S, eds). IARC WH, eds). Hong Kong:Hong Kong University Press, Scientific Publ No 88. Lyon: International Agency for Research 1985;355-357. on Cancer, 1987. 302. Vaillant C, Le Guellec C, Pakdel F, Valotaire Y. Vitellogenin 323. IARC. Cancer Incidence in Five Continents, Vol 6 (Parkin gene expression in primary culture of male rainbow trout hepa- DM, Muir CS, Whelan SL, Gao YT, Ferlay J, Powell J, eds). tocytes. Gen Comp Endocrinol 70:284-290 (1988). IARC Scientific Publ No 120. Lyon: International Agency for 303. Klein KO, Baron J, Colli MJ, McDonnell DP, Cutler GBJ. Research on Cancer, 1992.

776 Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

Appendix A: Toxicological Evaluations of embryo/fetotoxicity, and on effects on the genetic materials. In addition, studies on Pesticides and Environmental Chemicals absorption, biotransformation, bioaccumulation, distribution, and excretion are also required. In these studies, effects on Regulatory Requirements for * An evaluation of effects and residues of macromolecules, such as DNA, enzymes, Safety Data on Pesticides and the pesticide in vegetable and animal and other biochemical parameters are Other Chemicals food commodities after controlled often included. application trials using the principles of The toxicological data that are used in There is a principal difference between Good Agricultural Practice (GAP) the risk assessment are usually generated pesticides and other environmental chemi- * An evaluation of the possible intake of from animal experiments and in vitro cals. Pesticides are approved by the authori- pesticide residues from treated food investigations. When human data are avail- ties for use in the production of food crops. crops. In establishing the MRL for each able, for example from occupational or This means that in a number of cases food commodity, exceeding the ADI is accidental exposure, these of course are residues are accepted in food. However, the not accepted. considered highly valuable. pesticide residues in foods are very closely Data from toxicological investigations From the toxicological data a no regulated. This, of course, does not imply are a substantial part of the assessment of a observable effect level (NOEL) or a no that pesticide residues are accepted in other pesticide. The toxicological studies should observable adverse effect level (NOAEL) is media such as drinking water and air. identify possible adverse health effects of identified as the highest daily dose level For the environmental pollutants, their the compound and establish the dose at that does not produce observable effects or presence in food is not intended but may which such effects are likely to occur, and adverse effects in the most sensitive animal be caused by many different factors, of in particular identify a dose level where species. In establishing the ADI for which the contamination of food chains adverse effects are absent. The require- humans the NOEL is reduced by a and migration/formation during storage ments for toxicological investigations have (un)safety factor, which takes into account and production are significant. It should be been developed since the early 1960s, espe- the uncertainties of the results of the inves- emphasized that although the uses of vari- cially during the international cooperation tigation, the extrapolation from animals to ous highly persistent compounds such as within the Codex Alimentarius of the humans, and the variations in sensitivity PCBs and many chlorinated pesticides FAO/WHO. ADIs and suggestions of and life-style within the human population. (e.g., DDT) have been severely restricted MRLs are established at yearly meetings in When the toxicological background mater- or banned for a number of years in many a committee of independent experts, the ial is considered sufficient, a safety factor of Western countries, they can still be found so-called Joint FAO/WHO Meetings on 100 is normally used (a factor of 10 for dif- in foods, in particular in fat from fish and Pesticide Residues (JMPR). A similar ferences between species and 10 for differ- domestic animals. As a consequence of expert group, the Joint FAO/WHO Expert ences within species). Additional safety their persistence in the food chain, these Committee on Food Additives establishes factors are occasionally used, for example, compounds are also present in human tis- tolerable daily intakes of contaminants in when the biological effect is considered to sues. Concentrations of organochlorine foods. In the United States, U.S. EPA be particularly serious or when uncertainty contaminants in human reproductive tis- establishes reference doses (RfD) for pesti- exists in the evaluation of the consequences sue, adipose tissue, and blood from the cides and environmental chemicals using of a finding. Safety factors of 1000 or even general population worldwide, as well as principles similar to those ofJMPR. higher have occasionally been used. the concentrations in human breast milk The majority of the toxicological data When a clear NOAEL cannot be estab- fat, are presented in Tables Al and A2, used in the risk assessment are generated in lished on the basis of the available data, a respectively [adapted from Thomas and studies performed according to interna- lowest observed adverse effects level Colborn (1)]. tionally accepted standards (guidelines), (LOAEL) is sometimes identified and used Approval by regulatory agencies of a which for each type ofstudy state the mini- to establish an ADI. In this case, a safety pesticide demands that the health risk asso- mum requirements for an acceptable per- factor larger than 100 is always used (the ciated with exposure to the compound be formance. In addition, regulatory agencies numerical value depends on the quality of evaluated by the regulating authorities. In also require that, to be used in decision the data and the severity of the effect). Denmark, pesticides are approved for use making, the investigations are performed The ADI covers the intake during the by the Danish Environmental Protection according to Good Laboratory Practice whole lifetime, including childhood, and is Agency; and residues in foods are regulated (GLP) involving quality control and qual- defined as the daily intake of the com- through the establishment of maximal ity assurance. The different types of toxico- pound that-on the basis of all relevant residue limits (MRL) for each single pesti- logical investigations required for the information at that time-is considered cide by the Danish National Food Agency. evaluation of a pesticide involve studies on not to result in adverse health effects. ADI A substantial part of these evaluations takes acute effects, including effects on skin and is expressed in milligrams per kilogram of place within the framework of the mucous membranes, and more important body weight (bw). The use of "considered European Union. In brief, the MRLs are long-term studies on chronic effects, not to" and "on the basis of all relevant based on three parameters: including carcinogenicity following information at that point in time" under- * A toxicological evaluation of the pesti- repeated, daily exposure. Studies on effects lines that the toxicological investigations cide in question, which results in the on reproduction over a minimum of two have their inherent limitations, that a zero- establishment of an acceptable daily generations are also required, as well as risk does not exist, and that the ADI may intake (ADI) for humans special studies on teratogenicity and change in the light of new knowledge. ADI

Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 777 TOPPARI ET AL.

Table Al. Concentration of organochlorine contaminants in human reproductive tissue, adipose tissue, and blood from the general population worldwide.a Geographic location/year Number of samples Quantification analysis of sample/chemical (% positive) Units/mean SD Range Tissue Case informationb and detection limits Norway, Oslo No history of (1981-1982) exposure ppb (pg/kg) Maternal HCB 15(100) 2 2 serum/wet Caesarean 0.001 mg/kg (HCH) 20(90) 1 1 Normal delivery PCBs 15 (93.3) 10 4 Caesarean 0.002 mg/kg (PCBs) 20(100) 10 7 Normal delivery p,p'-DDE 15 (100) 19 21 Caesarean 0.002 mg/kg (DDE) 20(100) 10 8 , Normal delivery P-HCH 15(73.3) <1 Caesarean 0.001 mg/kg(bHCH) 20(30) <1 Normal delivery Umbilical HCB 12 (83.3) 2 1 cord serum Caesarean 20(85) 1 2 . Normal delivery PCBs 12 (75) 5 4 Caesarean 20(65) 3 1 Normal delivery p,p'-DDE 12(100) 10 9 Caesarean 20(100) 3 2 Normal delivery ,B-HCH 12 (16.6) <1 Caesarean 20(10) <1 Normal delivery

Yugoslavia ppb (pg/l) No history of GLC-ECD (1985-1986) Geometric Maternal exposure 1-HCH 14(100) 1.69 1.0-3.6 serum Nonpregnant 14 (100) 1.45 1.1-2.8 At delivery y-HCH 14(100) 1.77 1.2-3.6 Nonpregnant 14(100) 2.72 0.9-5.4 At delivery p,p'-DDE 14 (100) 9.31 6.1-14.4 Nonpregnant 14 (100) 7.57 2.8-11.7 At delivery p,p'-DDT 14(100) 5.87 3.2-9.7 Nonpregnant 14(100) 3.62 1.0-8.2 At delivery PCBs 14(100) 2.86 1.8-3.9 Nonpregnant 14(100) 2.01 1.1-4.65 At delivery

Yugoslavia, No history of GC-ECD North Adriatic ppb(Mg/l) exposure area (1989) Median HCB 10)100) 2 1.0-4 Blood Samples taken oa-HCH 10 (80) 2 1.0-2 from lactating P-HCH 10(80) 18 13.0-31 mothers p,p'-DDE 10 (100) 6 4.0-13 PCB 10 (100) 7 6.0-?

India, Bombay No history of GC-ECD (1987) Ppfm Amniotic fluid exposure PCBs 26(100) 0.131 0.026 0.001-1.162 Male professionals (1986-1987) ppm. PCBs 60(100) 0.837 0.01 0.005-3.33 Blood

India, Lucknow Placental No history of GLC-ECD (1979-1980) ppb tissue exposure HCH 27 (100) 39.9 10.7-97.9 Liveborn 9(100) 35.7 17.2-62.4 Stillborn a-HCH (Lindane) 27 (100) 17.1 4.1-95.6 Liveborn 9(100) 13.4 5.5-26.1 Stillborn Aldrin 27 (<100) 8 0.0-85.3 . Liveborn 9 (<100) 31.7 0.0-83.3 Stillborn p,p'-DDE 27 (100) 18.3 2.8-93.0 Liveborn 9(100) 12.4 4.7-22.3 Stillborn p,p'-DDT 27 (100) 13.8 2.0-46.3 Liveborn 9(100) 38.5 5.4-80.0 Stillborn DDT total 27(100) 39.8 7.6-162.2 Liveborn 9 (100) 60.8 21.9-93.2 Stillborn (continued)

778 Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

Table Al. (continued) Geographic location/year Number of samples Quantification analysis of sample/chemical (% positive) Units/mean SD Range Tissue Case informationb and detection limits India, Delhi and Faridabad No history of GLC-ECD (1987) ppb (ng/gI exposure HCB 7 (42.9) 12.19 8 0-64 Adipose/wet U.S. EPA Manual, 1980 HCB 4(100) 280 280 59-830 Adipose/wet

Poland, Pobnan No history of GC-ECD (1981) ppm (P9}gL exposure HCB 36 (86.1) 0.029 0.025 0.004-0.117 Testicular a-HCH 36(100) 0.018 0.014 0.004-0.058 Testicular )-HCH 36 (72.2) 0.124 0.09 0.049-0.571 Testicular y-HCH 36 (69.4) 0.022 0.02 0.004-0.083 Testicular A-HCH 36 (97.2) 0.055 0.053 0.011-0.254 Testicular E-HCH 36 (38.8) 0.094 0.045 0.039-0.177 Testicular p,p'-DDE 36 (97.2) 0.072 0.045 0.015-0.216 Testicular Heptachlor epoxide 36 (8.3) 0.131 0.047 0.083-0.194 Testicular

Massachusetts, New Bedford No history of (1985-1986) pp exposure PCBs 391 5.9 0.0-60.9 Blood Males 449 5.7 0.0-154.2 Blood Females Michigan Children, 4 year GLC-ECD (1980-1981) ppb old-no history (Webb-McCall (ng/mIl of exposure method) PCBs 205 (52.2) 4.18 3.29 1.00-19.40 Blood Fish exposure 3 ng/ml 80 (48.8) 4.82 4.81 1.00-23.30 Blood Farm exposure 3 ng/ml PBBs 205(12.7) 2.44 1.5 1.00-6.40 Blood Fish exposure 1 ng/ml 80 (21.3) 2.95 2.66 1.00-9.50 Blood Farm exposure 1 ng/ml DDT 202 (69.8) 4.36 3.87 1.00-21.40 Blood Fish exposure 1 ng/ml 79(77.2) 4.24 4.47 1.00-143.2 Blood Farm exposure 1 ng/ml

Missouri No history of HRGC-HRMS (1987) _Wt exposure 2,3,7,8-TCDD 50(100) 54.4 125.8 2-745 Adipose/whole 50(100) 0.519 1.314 0.013-8.290 Serum/whole

Texas, El Paso (1984-1985) No history of U.S. EPA, 1974 exposure Manual ppb of Analytical Methods Aldrin 112(34) 4.6 8.6 0.0-46.8 Blood BHC 112 (24) 2.5 2 0.0-16.5 Blood p,p-DDE 112 (99) 7.1 4.3 0.0-34.6 Blood Heptachlor 112 (19) 3.1 3 0.0-9.9 Blood

Texas, No history of GLC-ECD Gulf Coast exposure (1979-1980) ppm HCB 7(85.7) 0.021 0.0-0.043 Adipose 0.005 ppm (86%) p,p-DDE 7(100) 3.849 0.843-6.527 Adipose 0.004 ppm (96%) 2,4,5,2',4',5'-HCBP 7(100) 0.184 0.098-0.267 Adipose 0.005 ppm (92%) 2,3,4,2',4',5'-HCBP 7(100) 0.645 0.211-1.625 Adipose 0.005 ppm (96%) Germany, Munich No history of HRGC/HRMS exposure T4CDD 28 (100) 8 2.6-1 8.0 Adipose P5CDD 28 100) 16.4 7.7-40.4 Adipose H6CDD 28 (100) 94.7 35.7-178.2 Adipose H7CDD 28)100) 106.7 35.1-246.0 .. OCDD 28 (100) 373.2 116.5-789.1 .. T4CDF 28 (100) 2.5 0.7-12.8 ,, P5CDF 28 (100) 35.2 7.6-93.3 ,, (continued)

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 779 TOPPARI ET AL.

Table Al. (continued) Geographic location/year Number of samples Quantification analysis of sample/chemical (% positive) Units/mean SD Range Tissue Case informationb and detection limits Germany, Munich (continued) H6CDF 28 (100) 41.5 15.8-146.0 H7CDF 28 (100) 14.2 3.8-45.6 OCDF 28 (100) 4 1.2-13.5 T4CDD 28 (100) 16.4 1.0-88.9 Liver/fat P5CDD 28 (100) 20.2 7.3-58.7 H6CDD 28 (100) 166.8 56.4-615.1 H7CDD 28 (100) 1002.4 95.5-3463.1 OCDD 28 (100) 4416.2 472.7-15259 T4CDF 28 (100) 5 5 0 9-45 3 P5CDF 28 (100) 173.7 36.7-643.0 H6CDF 28 (100) 389.5 40.8-1800.7 H7CDF 28 (100) 218.9 12.2-757.0 OCDF 28 (100) 29.7 4.3-65.8

Pakistan, No history of GC/MS Quetta ppb (W/I) exposure (1987) 21(66.6) 0.396 0-1.88 Blood xc-HCH 21(90) 1.99 0-7.16 ,B-HCH 21(100) 9.26 0.53-18.98 4,4'-DDE 21(76.2) 0.94 0-4.83 4,4'-DDT

Spain, agrarian area (1985-1987) ppm (ml/kg) Adipose/ext. No history of GLC-ECD HCB 87(100) 2.99 2.24 lipid exposure p,p'-DDE 87 (100) 6.27 5.67 Lindane 87(100) 0.083 0.05 )-HCH 87 (100) 3.06 5.18 p,p'-DDD 87 (62) 0.079 0.079 p,p'-DDT 87(100) 1.5 0.89 Dieldrin 87 (100) 0.072 0.068

Australia, Sidney No history of GLC-ECD (1985-1986) ppmWnAigL exposure 292 (99.3) 3.72 0.0-26.30 Adipose/wet 0.001 ppm (81%) DDT 292 (99.3) 0.13 0.0-0.16 0.001 ppm (81-102%) Dieldrin

Canada, Ontario No history of GC/MS (1984) ppb (ng/gI exposure 141(100) 84 56 18-373 Adipose 1.4 ng/g HCB 141 (100) 84 82 14-530 3.0 ng/g )-HCH 141 (100) 33 27 2-107 1.1 ng/g Heptachlor 141 (100) 3237 2602 138-12167 1.2 ng/g epoxide 141 (99.3) 47 41 0-235 0.9 ng/g p,p'-DDE 141 (100) 84 80 7-369 1.7 ng/g Dieldrin 141 (92.2) 11 13 0-98 1.8 ng/g p,p'-DDT 141 (100) 2136 1473 197-11209 " 00lng/g Mirex PCB

Israel, No history of GC-MS Jerusalem exposure (males (1984-1985) 5-year history of infertility= ppb (ng/g) Study group) 29 3.09 6.81 0.0-28.8 Blood Study group p,p'-DDT 14 0.4 0.99 0.0-3.5 ., Control group 29(100) 16.1 11.82 2.9-43.4 ,, Study group p,p -DDE 14(100) 10.55 8.17 2.1-32.1 ,, Control group 29 7.111 6 0.0-21.6 ,, Study group o,p'-DDT 14 5.83 8.6 0-0-32 2 ,, Control group 29 2.61 3.51 0.0-15.4 ,, Study group (continued)

780 Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

Table Al. (continued) Geographic location/year Number of samples Quantification analysis of sample/chemical (% positive) Units/mean SD Range Tissue Case informationb and detection limits o,p-DDE 14 1.27 2.04 0.0-6.4 " Control group 29 2.28 2.42 0.0-9.8 " Study group Lindane 14 1.13 0.671 0.0-3.0 " Control group 29 3.65 3.71 0.0-15.9 " Study group Dieldrin 14 2.69 2.47 0.0-7.1 " Control group 29 (100) 8.31 3.83 3.0-15.8 ' Study group Heptachlor 14 (100) 11.64 3.67 17.0-21.7 " Control group epoxide 29 11.21 13.48 0.0-64.2 " Study group 14 7.94 14.69 0.0-47.3 " Control group Total PCBs

Japan No history of GC-ECD (1986-1988) ppgm exposure ,-HCH 23(100) 0.84 0.44 0.37-2.02 Adipose/fat p,p'-DDE 23 (100) 2.4 2.5 11.04-0.52 Heptachlor 23 (91.3) 0.07 0.06 0.00-0.25 epoxide 23 (100) 0.67 0.48 0.13-2.16 Chlordane 20(85) 0.08 0.08 0.00-0.27 Dieldrin

Japan, Tokushima ppb (ng/g) No history of GC-ECD (1984-1985) Geometric exposure Chlordane 22(100) 0.51 1.6 0.18-1.16 Blood Male subjects 0.01 ppb 21(100) 0.46 1.7 0.12-1.12 " Female subjects

Japan No history of GC-MS (1985) ppt(pg/g) exposure 2,3,7,8-TCDD 13 (92.3) 9 6-18 Adipose/wet (Cancer patients) 1,2,3,7,8-PCDD 13 (100) 15 3-36 1,2,3,4,7,8-HCDD 13 (69.2) 8 5-14 1,2,3,6,7,8-HCDD 13 (92.3) 70 26-220 1,2,3,7,8,9-HCDD 13(76.9) 12 4-44 1,2,3,4,6,7,9-HCDD 13 (69.2) 28 14-63 1,2,3,4,6,7,8-HCDD 13 (92.3) 77 29-180 08CDD 13 (92.3) 230 26-1100 PCDDs total 410 160-1400 2,3,7,8-T4CDF 13(100) 9 3-12 2,3,4,7,8-P5CDF 13(100) 26 4-71 1,2,3,4,7,8-/1,2,3,4,7,9-H6CDF 13 (84.65) 15 4-24 1,2,3,4,7,9-H6CDF 13 (84.6) 14 3-28 2,3,4,6,7,8-H6CDF 13 (23.1) 8 4-16 PCDFs total 63 7-120 Abbreviations: GC-ECD, gas chromatography-electron capture detector GLC-ECD, gas-liquid chromatography-electron capture detector GC-MS, gas chromatography-mass spectrometry HRGC-HRMS, high-resolution GC-MS. aAdapted from Thomas and Colborn (1). bGeneral information regarding the subjects analyzed. is not a danger-limit, but rather a safe value and used in such a way that the lowest possi- exceeding the ADI, as many samples nor- for an acceptable exposure to the pesticide ble amount of residue is produced. The mally are without detectable residues. in question. Exposures that lead to exceed- MRL is never established at a level higher The toxicological evaluations of conta- ing the ADI on a short-term basis are not than needed even if the established ADI minants and industrial chemicals are in expected to result in any health hazard. value would allow a higher residue content. principle performed following the same GAP is defined as the nationally autho- In practice, this means that the intake of principles that are used for the pesticides. rized safe method for application of a pesti- most pesticides by the general population is However, as exposure to such compounds is cide that under local conditions, such as well below the ADI. In the evaluation of the not intended, but in many cases is unavoid- climate, is found necessary for effective crop health risk, the possible total intake of the able, tolerable daily intakes (TDIs) are protection. The MRL is based on field pesticide is calculated as if the concentration established rather than ADIs. spraying trials and subsequent determina- in all the food in which it can be present is For many environmental chemicals the tions of residues. These investigations at the MRL for each single food item. This toxicological database is less comprehen- include different methods of application, means that exceeding the MRL in one single sive than for most pesticides. This is true including those using the highest dosages sample does not automatically result in for many chemicals that were introduced

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 781 TOPPARI ET AL.

Table A2. Human breast-milk fat concentrations worldwide compared with the lowest observed endocrine or reproductive effect levels in animal studies. a Human data Animal data Average levels in Chemical and units human breast milk fat High levels, location Doses, exposure route Impairment Dieldrin 0.05 ppm 1.78 ppm 20 mg/kg/bw. Induced testoserone, Australia Single injection to ten 3- 16ax- and 16p-hydroxylases 1.00 ppm week-old rats Iraq 1.00 ppm Uruguay Heptachlor and its epoxide 0.05 ppm 2.50 ppm 250 ,umole/kg/bw. Induced testosterone, Spain Single injection to 3- 1 6a- and 16P-hydroxylases 0.48 ppm week-old rats Italy Chlordane (oxychlordane 0.08 ppm >2.00 ppm 0.16 mg/kg wt/day. Elevation of plasma cortico- and trans-nonachlor) Mexico and Iraq Dietary exposure to pregnant mice sterone concentrations, when throughout gestation measured on day 400 in both male and female offspring. Total DDT 1.00 ppm >100 ppm Guatemala o,p'-DDT 0.25 ppm 2 ppm. Caused feminization Dissolved in corn oil and (a thickened ovarylike cortex injected into yolk of gull embryos in left testis).

HCB 0.10 ppm 7.00 ppm 300 ,umole/kg/bw. Induced testosterone, Greece Single injection to 3- 16ax- and 16p-hydroxylases. week-old rats. 1-HCH 1.00 ppm 6.50 ppm 50 mg/kg/bw (ppm). Caused reduction in testes weight. Chile Dietary exposure to weaned rats for 13 weeks. y-HCH 0.05 ppm 0.89 ppm 8 mg/kg (ppm). Caused reduction in testes weight, Italy ip daily in glycerine- degeneration of seminiferous suspension to adult male tubules and affected both rats for 10 days. spermatocytes and spermatids Total PCB 1.00 ppm 3.6 ppm 8 mg Aroclor 1254/kg Males exposed only during Canada (ppm). Dissolved in 0.2 lactation exhibited reduced (Hudson Bay) ml peanut oil. Lactating fertility as adults. Specifically, dams were exposed via decreased number of implants their diet on days 1,3,5,7, and decreased number of embryos and 9 of lactation when mated normal females 2,3,7,8-TCDD 2.00 ppt 1.45 ppb 0.064 jig/kg (ppb)- Male offspring had decreased Vietnam Dietary exposure to pregnant sperm counts when exposed (1 970s) rat on day 15 of gestation perinatally in utero and via lactation aAdapted from Thomas and Colborn) (1). in the society before the stricter regula- based on the monographs prepared by the and the World Health Organization. Most tions of new chemicals were enforced in expert groups of the JMPR and on the attention has been paid to those com- many countries during the 1970 to 1980s. reviews performed by the U.S. EPA as pounds that have not already been dis- published in the IRIS database (U.S. EPA's cussed in "Environmental Chemicals with Toxicological Summaries on Integrated Risk Information System). In Known Estrogenic Effects" and "Exposure Pesticides and Other addition, valuable information can be of Humans to Environmental Chemicals Chemicals Implicated as found in the Environmental Health with Estrogenic Activity and Their Effects Hormones Criteria (EHC) documents published by on Male Reproductive Health." Environmental the International Programme on Chemical Preliminary estimates of dietary intakes The following summaries on the most Safety (IPCS). These monographs are pub- are included (unpublished data). It is antici- important toxicological properties of vari- lished under the joint sponsorship of the pated, but not necessarily true for all of the ous compounds that have been implicated United Nations Environment Programme, compounds, that diet is the major route of as environmental hormones are mainly the International Labour Organisation, exposure for the general population. It is

782 Environmental Health Perspectives - Vol 104, Supplement 4 - August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

known that the drinking water in Denmark daily intake of 0.000003 mg/kg bw/day to pregnant hamsters. Cleft palate was meets the general limit standard of 0.1 can be estimated. 2,4-D was most recently induced in A/JAX mice at 15 mg/kg fg/liter, and that for most of the com- evaluated by JMPR in 1975 when an ADI bw/day; lower doses were not tested. Other pounds the levels of exposure are expected of 0 to 0.3 mg/kg bw was established (4). strains of mice were less sensitive. Higher to be very low. However, for some of the An RfD for noncarcinogenic effects of doses (approximately 200 mg/kg bw/day) herbicides, recent analysis of groundwater 2,4-D has been established by the U.S. EPA induce frank teratogenic effects in rats. A samples has in some cases revealed the pres- (5) at 0.01 mg/kg bw/day based on a no qualitative association between 2,4,5-T ence of 2,4-dichlorophenoxyacetic acid effect level of 1.0 mg/kg bw/day in rat sub- exposure and human birth defects has been (2,4-D) and atrazine near the limit value. chronic bioassays. At 5.0 mg/kg bw/day, suggested. Terata and fetotoxic effects have For pesticides and PCBs these prelimi- hematologic, hepatic and renal toxicities not been observed in monkeys up to a dose nary estimates of dietary intakes are based were observed. The safety factor was 100. of 40 mg/kg bw/day (5). 2,4,5-T does not mainly on the results and experiences of Histopathological examinations correlated accumulate in the mammalian organism. more than 30 years of pesticide control of well with kidney organ weight changes Alachlor. For alachlor, the dietary Danish and imported foodstuffs. The showing cortical and subcortical pathology. intake estimate of less than 0.006 mg/per- results have been continuously published Increases in ovarian weights and T4 levels son/year would correspond to a daily intake (in Danish) in reports from the Danish were not considered to be treatment of less than 0.0000003 mg/kg bw/day for National Food Agency. The latest report related. Increases were also noted in thy- a 60-kg person. Alachlor has not been was published in 1994, covering the results roid/pituitary weights at high dose levels; evaluated by JMPR. of the samples analyzed in 1993 (2). The however, no histopathological changes were An RfD for noncarcinogenic effects of latest sales statistics of the chemicals were found (5). Chronic toxicity and reproduc- alachlor was established by the U.S. EPA at published 1994 (3). tion studies of 2,4-D exposure indicated no 0.01 mg/kg bw/day based on a no effect adverse effects at dietary levels up to 500 level of 1 mg/kg bw/day in a 1-year dog- Herbicides ppm in dogs (approximately 14.5 mg/kg feeding study. At 3 mg/kg bw/day, hemo- The exact exposure to the herbicides bw/day), up to 1250 ppm in rats (approxi- lytic anemia was observed; and in an earlier described later in this section is not known. mately 62.5 mg/kg bw/day) (6), or at levels study using higher dose levels, changes To a limited extent, the herbicides have of 1000 ppm in drinking water (50-100 indicative of liver toxicity were recorded. been included in the Danish pesticide con- mg/kg bw/day) in pregnant rats (exposed The safety factor was 100. No signs of tes- trol program since the early 1980s. When through gestation and for 10 months fol- ticular toxicity were observed in any of analyzed, these herbicides were found to be lowing parturition) or their offspring these studies (5). almost absent in food samples during the (exposed for up to 2 years after weaning) In a 2-year chronic toxicity/carcino- 1970s in spite of considerable use. (7). 2,4-D does not accumulate in the genicity study of alachlor in rats, a NOEL Herbicides are used on plant food crops mammalian organism. Further toxicological for systemic toxicity was found at 2.5 only very early in the growth season and information on 2,4-D exposure can be mg/kg bw/day. At 15 mg/kg bw/day, molt- are therefore less likely to be present in the found in EHC 29 (8) and EHC 84 (9). ing of retinal pigmentation, increased final product. This view is supported by 2,4,5- Trichlorophenoxyacetic Acid. mortality rate in females, and abnormal the observations of other countries. MRLs For 2,4,5-trichlorophenoxyacetic acid disseminated foci in the liver of males were for the herbicides in fruits and vegetables (2,4,5-T), the dietary intake estimate of seen. In an earlier 2-year feeding study in have been set at the limit of detection, typ- less than 0.006 mg/person/year would cor- rats, 14 mg/kg bw/day and higher dose lev- ically 0.05 mg/kg. The above-mentioned respond to a daily intake of less than els resulted in degenerative ocular and observations lead to the assumption that 0.0000003 mg/kg bw/day for a 60-kg per- hepatic changes as well as other pathologi- the average herbicide concentration in all son. The most recent evaluation of 2,4,5-T cal gross and microscopic findings in the fruits and vegetables, after rinsing, peeling, by JMPR is from 1981 when an ADI of 0 thyroid, kidneys, brain, spleen, heart, and preparating for consumption, is at least to 0.03 mg/kg bw was established (10). prostate, and ovaries (5). 1000 times lower than the MRL, e.g., less An RfD for noncarcinogenic effects of In a three-generation reproduction than 0.05 fg/kg. In Denmark the yearly 2,4,5-T was established by the U.S. EPA at study of alachlor in rats using dietary levels intake of fruits and vegetables has been esti- 0.01 mg/kg bw/day and is based on a no of 0, 3, 10, and 30 mg/kg bw/day, no sig- mated at 60 kg per capita for each food cat- effect level of 3 mg/kg bw/day in a 2-year nificant adverse effects on the reproductive egory. This would correspond to an intake feeding study in rats and a three-generation performance of adult rats were observed at of less than 0.006 mg of each herbicide per reproduction study in rats. At 10 mg/kg any dose tested. Compound-related effects person per year. bw/day, increased urinary coproporphyrin were seen on kidneys (increased weights 2,4-Dich1orophenoxyacetic Acid. was observed in the chronic study and a and pathological changes in parents and 2,4-Dichlorophenoxyacetic acid and tendency toward reduced neonatal survival progeny) in the high-dose males and derivatives are currently being used as was seen in the three-generation reproduc- females. Lower ovary weights were noted in herbicides. The dietary intake estimate of tion study. The safety factor was 300 (5). the high-dose females of each parental gen- less than 0.006 mg/person/year would Other studies have also shown effects eration and in pups from the third genera- correspond to a daily intake of less than on reproduction in mice, rats, hamsters, tion. The NOEL for systemic toxicity was 0.0000003 mg/kg bw/day for a 60-kg per- and monkeys. A NOEL of 8 mg/kg 10 mg/kg bw/day (5). son. If it is assumed that drinking water bw/day for reduced fetal body weight was In a developmental toxicity study in contains 2,4-D at a level near the limit reported for NMRI mice (11). Fetal mor- rats, soft stools, hair loss, anogenital stain- value of0.1 fg/liter, and the daily consump- tality was observed after administration of ing, and maternal death were observed tion is set at 2 liters for a 60-kg person, a 2,4,5-T at 40 mg/kg bw/day (highest dose) after 400 mg/kg bw/day, whereas the

Environmental Health Perspectives Vol 104, Supplement 4 * August 1996 783 TOPPARI ET AL.

NOEL for maternal toxicity was 150 mg/kg (mainly benign) pituitary tumors. At this increases in the incidences of mammary bw/day. Based on a slight decrease in mean dose level the thyroid weight was increased fibroadenomas and adenocarcinomas. fetal body weight and a slight increase in throughout the study. There was no effect In beagle dogs administered atrazine for mean postimplantation loss, the NOEL with 10 ppm in the diet. Amitrole was con- 1 year in the diet, a dose level of 4.97 and LOAEL for developmental toxicity sidered not to be teratogenic in the rat at mg/kg bw/day was without toxicity. At a were 150 and 400 mg/kg/day, respectively. dose levels up to 1000 mg/kg bw/day. level of 33.65 mg/kg bw/day, clinical signs In the rabbit, the NOEL for developmen- Slight maternal toxicity and fetal and and pathological changes related to cardiac tal toxicity was found to be equal to or embryotoxicity (reduced fetal body weight/ toxicity were found (5). greater than 150 mg/kg bw/day (5). litter, reduced skeletal ossification and Atrazine is able to induce hormonal Amitrole. No residues of amitrole have increased incidence of enlarged and/or dark imbalances in rodents. Most work has been been detected in food after recommended fetal thyroids) were seen after dose levels of directed toward the effects of atrazine on use (12). The previously mentioned pre- 500 and 1000 mg/kg bw/day. the pituitary-gonadal axis. Steroid hormone liminary dietary intake estimate of less than In the hamster, up to 100 ppm of amit- metabolism was found to be impaired by 0.006 mg/person/year would correspond role in the diet produced no evidence of atrazine. At 120 mg/kg bw/day for 7 days, to a daily intake of less than 0.0000003 carcinogenicity, and no effect was observed atrazine increased the wet weight of the mg/kg bw/day for a 60-kg person. on the thyroid. anterior pituitary (hyperemia and hypertro- Amitrole was evaluated for acceptable In the rabbit, embryo and fetotoxicity phy) and reduced the activities of several daily intake by JMPR in 1993, when a were observed after a dose level of40 mg/kg steroid-metabolizing enzymes in the rat. temporary ADI of 0 to 0.0005 mg/kg bw bw/day. The no effect level was considered The deethylated metabolite of atrazine was was established based on a no effect level of to be 4 mg/kg bw/day. equally potent. Prenatal treatment with 0.5 mg/kg bw/day in a long-term toxicity/ No adequate reproduction study is atrazine or deethylatrazine (16 mg/kg bw, carcinogenicity study in rats. At a dose available. From a limited study in rats at sc) did not alter pituitary metabolism in level of 5 mg/kg bw/day, an increase in the dietary concentrations ranging from 25 to male pups, but atrazine increased 5-a- incidence of thyroid tumors and an increase 1000 ppm, effects on reproductive capabil- steroid reductase activity in the female in (mainly benign) pituitary tumors were ity were observed at a dose level of 500 ppm pups. Treatment pre- and postnatally with observed. Due to the noncompleteness of and above. Reduction of liver weight and atrazine and its metabolite decreased 3-ax- the available reproduction study, a safety thyroid hyperplasia was observed at the hydroxysteroid dehydrogenase activity, and factor of 1000 was used (13). lowest level studied, 25 ppm (equivalent to atrazine decreased 5-a-steroid reductase In mice, rats, and sheep, but not in 1.3 mg/kg bw/day). activity in the male pups. Both compounds Syrian hamsters, dogs, or cattle, amitrole Amitrole does not accumulate in the decreased the number of androgen-specific affects the thyroid. In the rat, a no effect mammalian organism. Additional toxico- binding sites in the prostate. Neither com- level was established at 2 ppm in the diet logical information can be found in EHC pound had any effect on female pituitary for 6 to 13 weeks (equivalent to 0.1 mg/kg 158 (12). androgen metabolism. In vitro studies have bw/day) based on increased iodine uptake Atrazine. For atrazine, the preliminary demonstrated inhibition of rat pituitary (shortly after injection), increased thyroid dietary intake estimate of less than 0.006 androgen metabolism by atrazine (14). weight, and histopathological changes of mg/person/year would correspond to a In a three-generation reproduction the thyroid (goiter and clearly activated daily dietary intake of less than 0.0000003 study with atrazine in rats, body weights thyroids). The mechanism involves inhibi- mg/kg bw/day for a 60-kg person. If it is were significantly lower for high-dose ani- tion of thyroid peroxidase, resulting in assumed that the drinking water contains mals (34.97 mg/kg bw/day) throughout decreases in circulating levels of T4 and T3, atrazine at a level near the limit value of the study. Body weight gains were also which stimulate the pituitary to increase 0.1 fg/liter, and the daily consumption is significantly depressed at the high dose. secretion of thyroid-stimulating hormone set at 2 liters for a 60-kg person, a daily The NOEL for parental toxicity was 3.5 (TSH), which in turn may cause thyroid intake of 0.000003 mg/kg bw/day can be mg/kg bw/day. At 34.97 mg/kg bw/day an hypertrophy, hyperplasia, and neoplasia. estimated. Atrazine has not been evaluated equivocal decrease in body weights of the Long-term toxicity/carcinogenicity by JMPR. second generation male pups was observed. studies have been performed in mice, rats, An RfD for noncarcinogenic effects of Therefore, the NOEL for reproductive and golden hamsters. In the mouse, liver atrazine was established by the U.S. EPA at toxicity was also 3.5 mg/kg bw/day (5). and thyroid tumors were observed after 0.035 mg/kg bw/day based on a no effect Several developmental toxicity tests have administration of 1000 mg/kg bw/day by level of 3.5 mg/kg bw/day in a 2-year rat been performed in rats and rabbits. In these gavage. In a special carcinogenicity study, feeding study. Mean body weights were studies, embryotoxicity and fetotoxicity in which pups were treated for a period of significantly depressed in males and females have been observed only at dose levels 90 weeks at a level of 500 ppm in the diet, a receiving 25 and 50 mg/kg bw/day. The where maternal toxicity was also evident. In slight increase in liver tumors was observed. absolute weight of liver and kidney in addition, developmental toxicity studies in In another 18-month carcinogenicity study males receiving 50 mg/kg bw/day was sig- rats of two metabolites ofatrazine, hydroxy- in mice, levels of 0, 1, 10, or 100 ppm in nificantly lower than that in controls. The atrazine and diaminochlortriazine, did the diet did not increase the incidence of safety factor was 100 (5). Acinar hyperpla- not reveal a specific potential concerning tumors. At 100 ppm an increase in thyroid sia of the mammary gland and epithelial embryo- or fetotoxicity (5). weight was observed. hyperplasia of the prostate were increased Atrazine does not accumulate in the In the rat, 100 ppm in the diet for 24 in males receiving 50 mg/kg bw/day when mammalian organism. months produced an increase in the inci- compared to controls. In females receiving Metribuzin. For metribuzin the dence of thyroid tumors and an increase in 25 or 50 mg/kg bw/day, there were dietary intake estimate of less than 0.006

784 Environmental Health Perspectives a Vol 104, Supplement 4 - August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

mg/person/year would correspond to a feeding study in rats, the pituitary weight detection limit is 0.1 mg/kg, which is also daily intake of less than 0.0000003 mg/kg relative to the body weight was decreased the MRL in most crops. However, in some bw/day for a 60-kg person. Metribuzin has after 40 mg/kg bw/day. fruits the MRL is 2 mg/kg. Considering that not been evaluated by JMPR. In two two-generation reproduction residues were found in only a few samples An RfD for noncarcinogenic effects was studies in the rat, the lowest NOEL for and that intake of residues primarily will established by the U.S. EPA at 0.025 reproductive effects was identified at 32.5 originate from fruits, it is assumed that the mg/kg bw/day based on a no effect level of mg/kg bw/day. At 100 mg/kg bw/day average concentration in all fruits, after 2.5 mg/kg bw/day in a 2-year dog-feeding reduced litter size was seen. In the parents, rinsing, peeling, and preparing for con- study. At 37.5 mg/kg bw/day mortality, increased kidney weights and renal lesions sumption, is at least 1000 times lower than decreased body weights and liver and kid- of the proximal tubules were observed after the MRL, e.g., less than 0.002 mg/kg. In ney effects were observed. The safety factor doses of 32.5 mg/kg bw/day. Teratology Denmark the yearly intake of fruits has been was 100 (5). studies in rats showed no malformations, estimated at 60 kg per capita. This implies In a 2-year chronic toxicity/carcino- whereas fetotoxicity (decreased mean fetal an intake of less than 0.120 mg of carben- genicity study in rats, the NOEL was body weight) was observed after doses of dazim per person per year corresponding to 5.0 mg/kg bw/day. Following a dose level 1000 mg/kg bw/day, a dose level that also a daily intake of less than 0.000006 mg/kg of 15.0 mg/kg bw/day, decreased body produced maternal toxicity. However, in bw/day for a 60-kg person. weight gain and pathologic changes in the another teratology study in rats, reduced In 1983, JMPR established an ADI of liver, kidneys, uterus, and mammary glands skeletal maturity and increased vascular 0 to 0.02 mg/kg bw for benomyl. In 1985, were observed. fragility were reported after doses of 20 an ADI was established of 0 to 0.01 mg/kg Metribuzin was not carcinogenic to mg/kg bw/day. In the rabbit, no terata were bw for carbendazim. These ADI values rats. Enlarged thyroids were reported in observed, whereas fetotoxicity (decreased were based on no effect levels for benomyl male rats administered 25 and 75 mg/kg fetal weight and increased number of fetal of 30 mg/kg bw/day for teratology studies bw/day for 3 months (5). Metribuzin runts) was seen at a maternally toxic dose in the rat and 125 mg/kg bw/day for was not carcinogenic in a 2-year mouse of 500 mg/kg bw/day (5). chronic studies in the rat. The no effect study. At 480 mg/kg bw/day, changes In one 2-year chronic bioassay of level for carbendazim was 2.5 mg/kg were observed in hematological parameters trifluralin using F344 rats, significant bw/day in subchronic dog studies (16-18). and liver weight. The NOEL was 120 increases were found in the incidences of An RfD for noncarcinogenic effects of mg/kg bw/day. bladder papillomas and renal pelvis carci- benomyl was established by the U.S. EPA Studies on developmental toxicity in nomas at the highest dose level tested in at 0.05 mg/kg bw/day based on a no effect rats and rabbits have not revealed terato- female and male rats. In addition, a level of 5 mg/kg bw/day in a three-genera- genicity at doses up to 100 or 15 mg/kg significant increase in the incidence of fol- tion reproduction study in rats. At 25 bw/day, respectively. In a three-generation licular cell tumors of the thyroid gland mg/kg bw/day, decreased pup weaning reproduction study in rats, a NOEL of 15 (adenomas and carcinomas combined) weights were observed. The safety factor mg/kg bw/day for reproductive effects and occurred at the highest dose tested in male was 100 (5). maternal toxicity was established (5). rats. Four other rodent chronic bioassays of Toxicological monographs on benomyl Metribuzin does not accumulate in the trifluralin in the diet have been performed. and carbendazim have also been published mammalian organism. These included a 2-year study in Sprague- by IPCS (15,19). Trifluralin. For trifluralin, the dietary Dawley rats, a 78-week study in Osborne- In the reproductive system, benomyl intake estimate of less than 0.006 mg/ Mendel rats, a 78-week study in B6C3F1 and its metabolite carbendazim cause, after person/year would correspond to a daily mice, and a 2-year study in B6C3F1 mice. an initial testicular swelling, a decrease in intake of less than 0.0000003 mg/kg Trifluralin did not produce significant testis and epididymis weight in adult male bw/day for a 60-kg person. Trifluralin has increases in the incidence of tumors in any rats, a reduction in caudal sperm reserves, a not been evaluated by JMPR. of these studies (5). decrease in sperm production, and a An RfD for noncarcinogenic effects of decrease of male fertility. At higher doses, trifluralin was established by the U.S. EPA Fungicides there is generalized disruption of all stages at 0.0075 mg/kg bw/day based on a no Benomyl (and Its Prineipal Metabolite of spermatogenesis. Benomyl does not effect level of 0.75 mg/kg bw/day in a 1- Carbendazim). Benomyl is widely used as affect copulatory behavior, seminal vesicles, year dog-feeding study. At 3.75 mg/kg a fungicide together with carbendazim. In sperm mobility, or related reproductive bw/day, increased liver weights and an both instances the active principle is hormones. The lowest benomyl concentra- increase in methemoglobin were observed. methyl-2-benzimidazolecarbamate and tion shown to induce statistically signi- The safety factor was 100 (5). MRLs are expressed as carbendazim. ficant spermatogenic effect in male rats was At higher dose levels administered to the The main source of exposure for the 45 mg/kg bw/day. The NOEL for these dog for 6 months (10 mg/kg/day), enlarged general population is the residue of benomyl effects was 15 mg/kg bw/day (15,19). The livers, discolored kidneys, corneal vascular- and carbendazim in food crops. Dietary testicular effects of benomyl treatment in ization, hemolytic anemia, and increased exposure analysis in the United States and the adult rat were shown to be reversible alkaline phosphatase concentrations were the Netherlands yielded an expected mean after a recovery period of 70 days. seen (5). daily intake of about one-tenth the ADI for In one study in which prepubertal male Several subchronic and chronic bio- benomyl and carbendazim (0.001-0.002 rats (33 days of age) were administered 200 assays in the mouse and rat have produced mg/kg bw/day) (15). In 1993, residues were mg benomyl/kg bw/day for 10 days, no NOELs that are higher than the one found in 1 of 72 samples of apples and 1 of effects on sperm parameters and testicular observed in the dog (5). In a 3-month 30 samples of bananas in Denmark. The histopathology were induced (15).

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 785 TOPPARI ET AL.

Several studies in rats and mice have the diet, equal to 7.0 mg/kg bw/day, based Maneb produced hepatocellular adeno- demonstrated teratogenicity of benomyl on increased serum TSH and decreased T4 mas in the mouse in a 79-week long-term and carbendazim. A NOEL of 30 mg/kg values at 250 ppm in the diet. toxicity/carcinogenicity study at a dose bw/day for benomyl has been established Mancozeb was not carcinogenic in the level of 2400 ppm in the diet. A no effect in the rat (5). mouse in two long-term toxicity/carcino- level was established at 60 ppm in the diet, Benomyl and carbendazim caused liver genicity studies. At the highest dose used, equivalent to 11 mg/kg bw/day, whereas tumors in two strains of mice that have a 1000 ppm in the diet (equivalent to 130 decreased body weight and decreased high spontaneous rate of liver tumors, mg/kg bw/day) decreased body weight and thyroxin levels were observed at 240 ppm. whereas carbendazim did not produce decreased T3 and T4 values were recorded. In a 31-month long-term toxicity/car- tumors in a mouse strain with a low spon- In a two-year long-term toxicity/carcino- cinogenicity study in the rat, a no effect taneous incidence of tumors. Carcino- genicity study in the rat, a no effect level of level of 300 ppm in the diet, equivalent to genicity studies with both compounds in 125 ppm in the diet, equivalent to 4.8 20 mg/kg bw/day, was observed. At 1000 the rat were negative (10,20). mg/kg bw/day, was observed. At 750 ppm ppm in the diet decreased body weight, Benomyl and carbendazim do not accu- in the diet, decreased body weight gains, decreased T4 values, and increased thyroid mulate in the mammalian organism. decreased T3 and T4 values, increased weights were seen. There was no evidence Ethylenebisdithiocarbamates: thyroid weights, thyroid follicular cell of carcinogenicity. Mancozeb, Maneb, Metiram, and Zineb. hypertrophy, hyperplasia, and nodular In two two-generation reproduction Ethylenebisdithiocarbamates (EBDCs) are hyperplasia were seen. Increased incidences studies, the no effect level was observed at fungicides primarily used in the production of thyroid follicular cell adenomas and/or 75 ppm in the diet, equivalent to 5.6 of fruits and vegetables. Residues are fre- carcinomas were noted at this dose level. mg/kg bw/day. At 300 ppm, increased quently found at levels below the MRLs, In two 2-generation reproduction organ to body weight ratios for liver and which typically are between 0.5 mg (the studies, the overall no effect level was kidney, and follicular cell hyperplasia of detection limit) and 2 mg/kg. If a concen- observed at 120 ppm in the diet, equiva- the thyroid were seen. tration of 1 mg/kg is used as an average lent to 7.0 mg/kg bw/day. At 150 ppm Two teratogenicity studies in rats pro- MRL for fruits and vegetables, and it is decreased body weights were seen and at duced embryo fetotoxicity (increased early assumed that the average concentration in 1200 ppm microscopic changes were resorptions, increased postimplantation all fruits and vegetables, after rinsing, peel- observed in the thyroid, kidney, and pitu- losses and a decrease in viable fetuses) only ing and preparing for consumption, is at itary gland. Other changes found were at dose levels at which maternal toxicity least 100 times lower than the MRL, an increased relative weights of liver, kidney was clearly evident (13). intake of less than 1.2 mg per person per and thyroid; decreased gestation and lacta- The data on metiram would support an year can be obtained from 60 kg of fruits tion, body weight, and feed consumption; ADI of 0 to 0.03 mg/kg bw based on a no and 60 kg of vegetables. This would corre- decreased premating body weight and effect level of 2.5 mg/kg bw/day for thy- spond to a daily intake of less than 0.00006 feed consumption. roid effects in dogs and the no effect level mg/kg bw/day for a 60-kg person. Several teratogenicity studies in rats and of 3.1 mg/kg bw/day in the long-term rat These compounds were evaluated for rabbits have produced embryo and fetotox- study, using a safety factor of 100. acceptable daily intake by JMPR in 1993. icity only at dose levels where maternal In the mouse, rat, dog, and monkey, A group ADI of 0 to 0.03 mg/kg bw was toxicity was clearly evident. At a high dose metiram affects the thyroid. In two 13- established for each compound alone or in level of 512 mg/kg bw/day, teratogenic week studies in rats, the no effect levels combination because of a) the similarity effects were seen (13). of metiram were 80 and 100 ppm in the of chemical structure of the EBDCs; The data on maneb would support an diet, equivalent to 5.8 and 6.0 mg/kg b) the comparable toxicological profiles of ADI of 0 to 0.05 mg/kg bw, based on the bw/day, respectively, based on decreased the EBDCs based on the toxic effects of no effect level of 5.0 mg/kg bw/day for serum T4 and an increase in absolute thy- ethylenethiourea (ETU), which forms part thyroid effects in rat studies using a safety roid weight at 300 and 960 ppm, respec- of the terminal residue to which con- factor of 100. tively. In dogs, follicular cell hyperplasia of sumers of products treated with the In the rat, dog, and monkey, maneb the thyroid was observed after 1000 ppm EBDCs are exposed; and c) the fact that affects the thyroid. In a 13-week study in in the diet for 52 weeks. The no effect level parent EBDC residues cannot be differen- rats, the no effect level of maneb was was 80 ppm, equivalent to 2.5 mg/kg tiated using presently available analytical 80 ppm in the diet equal to 5.0 mg/kg bw/day. In the monkey, decreased serum procedures (13). bw/day, based on an increase in absolute T3 and T4 levels, increased thyroid These compounds are rapidly absorbed, thyroid weight and thyroid follicular cell weights, and minimal follicular cell hyper- extensively metabolized, and rapidly hyperplasia at 400 ppm. In two studies in plasia of the thyroid were observed after 15 excreted (90%) within 24 hr in rodents. dogs, follicular cell hyperplasia in the thy- and 75 mg/kg bw/day administered by ETU is the major metabolite. roid was observed after 400 ppm in the gavage for 26 weeks, whereas 5 mg/kg The data on mancozeb would support diet for 13 weeks and 1000 ppm in the diet bw/day was without effect. an ADI of 0 to 0.05 mg/kg bw, based on a for 52 weeks, respectively. The overall no Metiram was not carcinogenic to the no effect level of 4.8 mg/kg bw/day in a rat effect level was 200 ppm, which is equiva- mouse in an 88-week long-term toxicity/ long-term toxicity/carcinogenicity study lent to 6.4 mg/kg bw/day. In the monkey carcinogenicity study at dose levels up to using a safety factor of 100. an increased thyroid weight was observed 1000 ppm in the diet. A no effect level was In the rat, mancozeb affects the thy- after 300 ppm in the diet for 6 months, established at 300 ppm in the diet, equal to roid. In a 13-week study in rats, the no while 100 ppm in the diet, equivalent to 24 mg/kg bw/day, whereas decreased body effect level of mancozeb was 125 ppm in 7.3 mg/kg bw/day, was without effect. weight was observed at 1000 ppm.

786 Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

In a 111-week long-term toxicity/car- were observed. From the limited data avail- Long-term dietary exposure of humans cinogenicity study in the rat, a no effect able, a dose level of 50 mg/kg bw/day to HCB caused an epidemic of porphyria level of 80 ppm in the diet, equivalent to appeared to be without significant adverse cutanea tarda (PCT) in Turkish citizens 3.1 mg/kg bw/day, was observed. At reproductive effects. who accidentally consumed bread made 320 ppm in the diet, muscular atrophy Zineb was not teratogenic in mice at from grain treated with HCB. In children was observed. There was no evidence dose levels up to 2000 mg/kg bw/day dur- less than 1 year of age, pink sore disease of carcinogenicity. ing critical periods of organogenesis, nor was observed along with 95% mortality. In In a three-generation, two litter per were maternal or embryo/fetotoxicity addition to the PCT-associated symptoms generation reproduction study in rats, no observed. The no effect level for terato- of skin lesions, hypertrichosis, and hyper- adverse effects were revealed on reproduc- genicity and embryo/fetotoxicity was 630 pigmentation, the exposure caused neuro- tive parameters. The no effect level was mg/kg bw/day in the rat. At the maternally toxicity and liver damage. Follow-up observed at 40 ppm in the diet, which was toxic dose level of 2000 mg/kg bw/day, studies reported PCT symptoms, reduced equivalent to 1.8 mg/kg bw/day. At 320 zineb was teratogenic, resulting in a sig- growth and arthritic changes in the ppm decreased parental body weight and nificant increase in hydrocephalus and appendages of children who were directly food consumption were recorded in the Fo skeletal anomalies (13). or indirectly (i.e., through breast milk) and F1 generations. Hexachlorobenzene. The use of hexa- exposed. Accurate exposure data (dose and Metiram was not teratogenic in rats chlorobenzene (HCB) as a seed dressing to duration) are lacking (5). HCB is a potent and rabbits at any dose tested during criti- prevent fungal disease on grains was dis- inducer of hepatic microsomal enzymes. cal periods of organogenesis. The no effect continued in most countries in the 1970s. Several studies have indicated that HCB levels for embryo/fetotoxicity were 80 HCB continues to be released into the affects the immune system. mg/kg bw/day in the rat based on slight environment as a byproduct and contami- A number of investigations have indi- decreases in litter size and weight after dose nant in many other chlorinated chemicals cated that repeated exposure to HCB can levels of 160 mg/kg bw/day, and 40 mg/kg including chlorinated solvents. HCB is affect male reproduction in various species, bw/day in the rabbit, based on decreases in highly stable in the environment and accu- but only at relatively high doses. Thus, 30 mean fetal weights at 120 mg/kg bw/day. mulates in the food chains. Humans are mg/kg bw/day for 21 days reduced serum At both these dose levels maternal toxicity almost exclusively exposed through the testosterone levels in male mice, probably was evident (13). diet, especially meat, fish, and poultry; and due to induction of hepatic microsomal In the rat and dog, zineb affects the HCB accumulates in human adipose tis- enzymes. Decreased fertility was observed thyroid. Dietary administration of zineb to sue. In Denmark 0.1 ppm of HCB was in male rats after 250 mg/kg bw/day for 5 rats for 6 weeks indicated a no effect level found in the fat from mothers' breast milk days. Histologic changes in the testes, lead- of 500 ppm, equivalent to 25 mg/kg in 1986 (21). Typical concentrations found ing to retarded maturation has been noted bw/day, based on morphological changes in various foodstuffs of animal origin are in the pig after 50 mg/kg bw/day for 90 of the thyroid gland and reduced iodine around 0.01 mg/kg fat. Similar values are days. In female monkeys HCB caused uptake at 5000 ppm. Rats treated by gav- found for fish. From the continuous sur- degenerative changes in the reproductive age for 4 weeks exhibited slight hyperplasia veillance of HCB in food, a maximal intake tissues. In female rats HCB produced ele- of the thyroid at a dose level of 1000 ofless 0.73 mg per person per year has been vated serum progesterone levels and no mg/kg bw/day, resulting in a no effect level estimated. This corresponds to an intake of estrogenic effects were observed. of 250 mg/kg bw/day. In a limited study in less than 0.000033 mg/kg bw/day. In a four-generation rat reproduction dogs, follicular cell hyperplasia of the thy- In 1978, JMPR withdrew the previous study, reduced litter sizes, increased num- roid was observed after 10,000 ppm in the ADI of0 to 0.0006 mg/kg bw/day (22). ber of stillbirths, and reduced pup survival diet for 52 weeks, whereas 2000 ppm had An RfD for noncarcinogenic effects was were observed after a dose of 4 mg/kg no effect. established by the U.S. EPA at 0.0008 bw/day. A NOAEL was established at Zineb was not carcinogenic when given mg/kg bw/day based on a no-effect level of 1 mg/kg bw/day. to mice at a dose of 460 mg/kg bw/day 0.08 mg/kg bw/day in a 130-week chronic Oral administration ofHCB has induced from postnatal day 7 until weaning, fol- feeding study in rats. The study involved tumors in the liver, thyroid, parathyroid, lowed thereafter with dietary administra- feeding male and female rats diets contain- and kidney in three rodent species. tion of 1300 ppm until 18 months of age. ing HCB for 90 days prior to mating and HCB was negative in most in vitro and In a two-year study in rats, dietary levels of until 21 days after parturition (at weaning), in vivo mutagenicity assays (5). 500 ppm and higher revealed goitrogenic whereafter the offspring that had been Tributyltin Compounds. Tributyltin effects. At or above 1000 ppm renal con- exposed to HCB and metabolites in utero compounds were previously used in large gestion, nephritis, and nephrosis increased and from maternal nursing received the quantities in marine antifouling paints on mortality and diminished growth rate. compound in their diets for the remainder ships, boats, and mariculture pen nets. There was no evidence of carcinogenic of their lifetime (130 weeks). At a dose However, their use is now restricted in potential. Neither of these two studies is level of 0.3 mg/kg bw/day, liver changes most countries. considered adequate for the study of long- were observed. At higher doses increases in Exposure to tributyltin compounds is term toxicity/carcinogenicity according to pup mortality, hepatic toxicity including not well documented; however, the dietary the modern standard. increased liver porphyrin levels, and severe exposure is expected to be indeed very low. Treatment of rats with zineb at doses of chronic nephrosis (males only) were In 1991, JMPR established an ADI of 50 to 960 mg/kg bw/day suggested adverse observed. The safety factor was 100 (5). 0 to 0.0005 mg/kg bw for tributyltin com- effects on the reproduction. Sterility, HCB-induced porphyria has been reported pounds (fentin compounds) based on a decreased fertility, and resorption of fetuses for all species examined, except the dog. no-effect level of 0.5 mg/kg bw/day in

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 787 TOPPARI ET AL.

short-term rat studies with respect to preparing for consumption is at least 100 fasciculata of the adrenals (males and general, as well as specific, effects on the times lower than the MRL, an intake of females) were seen. The no effect level was immune system (23,24). less than 2.4 mg per person per year can be 7.0 mg/kg bw/day. An RfD for noncarcinogenic effects of obtained from 60 kg of fruits and 60 kg of Vinclozolin did not show carcinogenic- tributyltin compounds was established by vegetables. This would correspond to a ity in long-term studies in rats and mice the U.S. EPA at 0.00003 mg/kg bw/day daily intake of less than 0.00012 mg/kg after doses as high as 278 mg/kg bw/day based on a no effect level of 0.025 mg/kg bw/day for a 60-kg person. (rats) and 912 mg/kg bw/day (mice). In bw/day in a chronic feeding study in rats. Vinclozolin was evaluated for accept- the long-term mouse study, increases in At 0.25 mg/kg bw/day, an effect on host able daily intake by JMPR in 1986 (25), absolute and relative weights of the liver resistance to T spiralis was reported. The where a temporary ADI of0 to 0.04 mg/kg were noted in females at the two highest safety factor was 1000 (5). bw was estimated based on a no effect level dose levels, 287 and 912 mg/kg bw/day. In a parallel lifetime carcinogenicity/ of 7 mg/kg bw/day in a 6-month dog feed- Increased liver and testes weights were seen chronic toxicity study in rats, the only ing study. A safety factor of 200 was used. in male mice at the highest dose level of observed immunological alterations were The ADI was made temporary as 3,5- 818 mg/kg bw/day. After a complete an increase of IgM and a decrease of IgG dichlorophenylcarbamoyl-2-propionic acid microscopic examination no toxicologically titers in the females at a dose of 2.5 mg/kg was identified as a major plant metabolite significant differences were observed bw/day after 12 and 24 months of treat- that had not been fully evaluated with between treated and control animals. The ment. No effect was observed after a dose respect to its toxicity, and additional no effect level was 87 mg/kg bw/day. In of 0.25 mg/kg bw/day (5,24). Similar studies were requested. In 1988 (26) the long-term rat study, absolute weights immunological effects have been observed JMPR established an ADI of 0 to 0.07 of several organs were decreased at the two in short-term rat studies. In a 4-week feed- mg/kg bw for vinclozolin and all metabo- highest dose levels. The incidence of ing study, thymic atrophy was reported at lites containing the 3,5-dichloroaniline macroscopic and microscopic findings were dietary concentrations as low as 1 mg/kg moity, as additional data ensured the low randomly distributed among all test groups bw/day. In a companion study, rats fed 1 acute toxicity and absence of mutagenicity and did not appear to be related to treat- or 4 mg/kg bw/day in the diet for at least of this plant metabolite. The safety factor ment with vinclozolin. The no effect level 6 weeks displayed impaired cell-mediated was 100. was 27 mg/kg bw/day. immunity (5). An RfD for noncarcinogenic effects of Special studies in mice and rabbits did Short-term studies suggest that trib- vinclozolin was established by the U.S. not reveal any teratogenic potential of vin- utyltin decreases the activity of the pitu- EPA at 0.025 mg/kg bw/day based on an clozolin. In the mouse study it was found itary-thyroid axis. Structural effects were estimated no effect level of 2.5 mg/kg that mice receiving 6000 ppm or higher in observed in the pituitary and thyroid bw/day from the same dog study used by their diets did not have any implantations. glands, and changes in circulating hor- JMPR. The safety factor was 100 (5). The no effect level was 1 10 mg/kg bw/day. mones were observed. However, in long- After absorption vinclozolin is rapidly In a three-generation rat reproduction term studies most of these effects seemed and extensively biotransformed and study, no effects were observed on alter- to be absent. The mechanism of action is excreted as conjugated metabolites. The ation in fertility, gestation length, litter not known (5,24). major metabolite found in rat urine was size, sex ratio of pups, fetal birth weight, Tributyltin was found to be fetotoxic N-(3,5-dichlorophenyl)-2-methyl-2,3,4- weight gain of pups, or survival of pups. and teratogenic in mice and rats at a level trihydroxybutanoic acid amide (25). No treatment-related malformations or at which overt maternal toxicity was also Vinclozolin has a low order of acute developmental defects were noted, and no observed (18 mg/kg/day), with no effects toxicity and was negative in a number of in treatment-related pathological findings observed at 9 mg/kg/day (5). vitro and in vivo mutagenicity assays. were seen in either parent rats or pups at One carcinogenicity study in rats In a 6-month feeding study, groups of necropsy. The NOAEL was estimated reported increased incidences of benign dogs were given daily dosages of vinclo- to be higher than the highest dose tested, tumors of the pituitary gland (not dose zolin equal to 0, 7.0, 20, 41, and 135 152 mg/kg bw/day (25). related) and adrenal medullary tumors mg/kg bw/day (males) or 0, 7.4, 21, 41, In contrast, recent reproduction/toxicity (pheochromocytomas) at a high dose. These and 141 mg/kg bw/day (females). Hemo- studies conducted in rats have shown tumor types appear usually at high and vari- lytic anemia was observed in high-dose Leydig cell tumors and atrophic ventral able background incidences, and the signi- males and females, while other clinical prostate and seminal vesicles after chronic ficance ofthe finding is questionable (24). chemistry parameters were not affected. exposure of adult male rats to vinclozolin. Dicarboximides (Vinclozolin). Vin- Dose-related increases in absolute and rela- In addition, anogenital distance in male clozolin (3-(3,5-dichlorophenyl)-5-ethenyl- tive weights of adrenals were noted in ani- offspring was reduced to a femalelike size 5-methyl-2,4-oxazolidinedione) is used as a mal feed with dose levels of 20 mg/kg (27). Subsequently, another toxicological fungicide on fruits, vegetables, vines, and bw/day or higher. Other organ weight study (28) showed strong demasculinizing ornamental plants and grasses. Residues changes were noted, including decreased effect of vinclozolin in the male offspring slightly above the Danish MRLs of 1 to 2 relative pituitary weights in females at dose of mice treated with 100 or 200 mg/kg mg/kg are occasionally found in some levels of 21 mg/kg bw/day or higher. Upon bw/day from gestational day 14 to postna- fruits and vegetables. If a concentration of microscopic examination, changes related tal day 3; in addition to the femalelike 2 mg/kg is used as an average MRL for to the anemia observed at the highest dose anogenital distance many other abnormali- fruits and vegetables and it is assumed that were recorded and, in additition, atrophy/ ties were observed, such as retention of the average concentration in all fruits and stromal proliferation of the prostate nipples, hypospadias, epididymal granulo- vegetables, after rinsing, peeling, and (males) and vacuolization of the zona mas, maldescent of the testis, and frequent

788 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

presence of vaginal pouch. These findings insecticides. Thus, a daily intake of less adverse effect on reproduction have been suggested that vinclozolin possesses anti- than 0.00003 mg/kg bw/day for a 60 kg reported (31). androgenic activity. Recent biochemical person is assumed. The results of numerous carcinogenicity studies have demonstrated that the demas- Carbaryl. Carbaryl is approved as an studies in mice and rats have mainly been culinizing effects of vinclozolin may be insecticide for use on plants and fruit trees. negative. However, these studies are old attributed to two intermediary metabolites, Food is considered the major exposure and do not meet modern standards. In 2-(3,5-dichlorophenyl)-carbamoyl)oxy- route for carbaryl. Residue findings ranging the majority of genotoxicity assays, car- 2-methyl-3-butenoic acid (MI) and from trace to 0.05 mg/kg food have been baryl did not have any DNA-damaging 3', 5'-dichlo-2-hydroxy-2-methylbut-3- reported. In United States the intake in properties (31). enanilide (M2), which inhibit the andro- 1969 was estimated at 0.003 mg/person/ Malathion. The use of malathion is gen receptor function, although in certain day (31). The MRL for cereals is 0.5 rather limited in food crops. No residues conditions these metabolites may act as mg/kg. Assuming a concentration of 0.5 were reported in 1993 when a limit of 1 agonists (29,30). mg/kg as an average for fruits and vegeta- mg/kg food was used. If it is assumed that bles, and assuming that the average con- residues are present in 1/12 of fruits and Insecticides centration in fruits and vegetables after vegetables and assuming that the average A number of the insecticides mentioned rinsing, peeling, and preparing for con- concentration in the food items after rinsing, below are no longer approved for use. sumption is at least 100 times lower than peeling and preparing for consumption is at However, several of the compounds are the MRL, an intake of less than 0.6 mg per least 1000 times lower than 1 mg/kg, an very persistent and can be found in the person per year can be obtained from intake of less than 0.01 mg per person per environment as general background pollu- 60 kg of fruits and 60 kg of vegetables. year can be obtained from 10 kg of fruits tion. Therefore, residues may still be found This would correspond to a daily intake of and vegetables. This would correspond to a in foods. In addition, residues may be con- less than 0.00003 mg/kg bw/day for a daily intake of less than 0.0000005 mg/kg tained in foods imported from countries 60-kg person. bw/day for a 60-kg person. where some of the compounds are still in Carbaryl was last evaluated by JMPR Malathion was last evaluated by JMPR use. In particular, the organochlorine pesti- for acceptable daily intake in 1973. An for acceptable daily intake in 1966. An ADI cides are bioaccumulating in the food ADI of 0 to 0.01 mg/kg bw has been estab- of0 to 0.02 mg/kg bw was established (33). chains and are found widely distributed in lished based on the following no effect An RfD for noncarcinogenic effects was fat-containing foods of mammalian or levels: 10 mg/kg bw/day in the rat; 1.8 established by the U.S. EPA at 0.02 mg/kg marine origin. In contrast to this environ- mg/kg bw/day in the dog; and 0.06 mg/kg bw/day based on a no effect level of 0.16 mentally based contamination of foods, it bw/day for cholinesterase (ChE) inhibition mg/kg bw/day in a subchronic human is characteristic that the development of in humans (31,32). feeding study. At a dose of 0.34 mg/kg several different types of new insecticides An RfD for noncarcinogenic effects was bw/day, depression of red blood cell during the last few years has resulted in a established by the U.S. EPA at 0.1 mg/kg cholinesterase activity was seen. The safety shift to more specific and limited uses of bw/day based on a no effect level of 9.6 factor was 10 (5). the single approved insecticide. This means mg/kg bw/day in a chronic feeding study In a 2-year chronic toxicity/carcino- that residues of each single insecticide in rats. At a dose of 15.6 mg/kg bw/day, genicity study in rats, a NOEL of 5 mg/kg are not widespread but rather found in a kidney and liver toxicity was seen. The bw/day was observed. At a dose of 50 limited number ofcrops. safety factor was 100 (5). mg/kg bw/day, decreased brain cholin- f-HCH, dieldrin, heptachlor, hep- Carbaryl has been shown to affect mam- esterase and body weight were reported. tachlor epoxide, and lindane (y-HCH) are malian reproduction and perinatal develop- In a reproduction study in rats, reduced included in the Danish pesticide control ment adversely in a number of species. number oflive pups and reduced pup body program. With a few exceptions they are Effects include impairment of fertility, weight were seen at a dose of 240 mg/kg never found in fruits and vegetables, decreased litter size, and reduced postnatal bw/day, the only dose tested. No terato- including cereals. Typical concentrations in viability. Malformations have also been genic effect was seen in the rat after ip fat of animal and marine species is about observed. Adverse reproductive and devel- injection of 900 mg/kg bw/day (5). 0.01 mg/kg of fat. The intake from heavy opmental effects were noted only at doses Methomyl. Methomyl is approved for consumption of animal fat and fish is esti- that caused maternal toxicity. Data indicate use on apples, pears, and some berries. mated at less than 0.7 mg per person per that the reproductive and developmental Residues found in GAP trials are from 0.5 year. This would correspond to a daily processes of mammals are not especially to 1.0 mg/kg raw food. In considering the intake of less than 0.00003 mg/kg bw/day sensitive to carbaryl compared with the reduction after rinsing, peeling, and prepar- for a 60-kg person. susceptibility ofthe adult organism (31). ing for consumption, it is assumed that the Chlordane and oxychlordane, methoxy- In the mouse, doses of carbaryl up to average concentration in fruits is at least chlor, mirex, and toxaphene are not 35 mg/kg bw/day for 5 days did not affect 1000 times lower than 1 mg/kg, and that included in the Danish pesticide control weights of testes and accessory sex glands only 1/6 of fruits are contaminated. This program. These compounds are not used or uptake of progesterone by the prostate would correspond to an intake of less than in Denmark. Therefore, the knowledge on gland. However, at 68 mg/kg bw/day, 0.01 mg of methomyl per person per year. exposure is rather limited. From concentra- increased hepatic androgen hydroxylase The daily intake would then be 0.0000005 tions reported in human mothers' milk activity was indicated (31). mg/kg bw/day for a 60-kg person. (21), it seems reasonable to assume, that There are conflicting data about effects Methomyl was evaluated by JMPR for the exposures to these compounds are not of carbaryl on sperm counts and changes in acceptable daily intake in 1989. An ADI of higher than to the previously mentioned sperm morphology in plant workers. No 0 to 0.03 mg/kg bw was established (34).

Environmental Health Perspectives - Vol 104, Supplement 4 - August 1996 789 TOPPARI ET AL.

An RfD for noncarcinogenic effects was NOEL for reproductive parameters was determined that the reproductive effects of established by the U.S. EPA at 0.025 2.1 mg/kg bw/day. methoxychlor are mediated in part by an mg/kg bw/day based on a no effect level of Dicofol was not teratogenic in mice, rats increase in prolactin release, which in turn 2.5 mg/kg bw/day in a 2-year feeding or rabbits. Embryotoxicity and fetotoxicity influences the hypothalamic levels of study in dogs. At 10 mg/kg bw/day, kidney were observed at maternally toxic dose GnRH. This may be considered an and spleen toxicity was seen. Histopatho- levels (35). early effect of methoxychlor on the rat logic changes were observed in the liver Methoxychlor. The dietary intake of reproductive system. and bone marrow at 50 mg/kg bw/day. methoxychlor has been estimated at less than Cummings and Gray (39) found that The safety factor was 100 (5). 0.00003 mg/kg bw/day for a 60-kg person. methoxychlor affects the decidual cell The NOEL observed in the dog study Methoxychlor was last evaluated by response of the rat uterus, suggesting a is further supported by lifetime studies in JMPR for acceptable daily intake in 1977. direct effect of the compound on the rats and mice, and a three-generation An ADI of 0 to 0.1 mg/kg bw was estab- uterus with no effects on uterine weight, reproduction study in rats where higher lished (20). serum progesterone levels, or corpora lutea NOELs were seen (5). An RfD for noncarcinogenic effects was maintenance. Long-term exposure to Dicofol In 1993 Dicofol residues were established by the U.S. EPA at 0.005 methoxychlor reduced fertility and induced found in a limited number of apples, citrus mg/kg bw/day based on a no effect level of fetotoxicity. The effects of reduced fertility fruits, and berries at levels below the 5.01 mg/kg bw/day in a rabbit teratology and fetotoxicity were noted in a three-gen- reporting limit of 1 mg/kg. If it is assumed study. At 35.5 mg/kg bw/day, excessive eration reproduction study. Although the that residues are present in 50% of fruits loss of litters due to maternal toxicity was available data from these three studies are and assuming that the average concentra- seen. The safety factor was 1000 (5). limited, it is apparent that methoxychlor at tion in fruits after rinsing, peeling, and Teratology studies in rats as well as a 1000 ppm produced reproductive effects in preparing for consumption is at least 100 three-generation reproduction study in rats the form of reduced fertility index, reduced times lower than 1 mg/kg, an intake of less and a 2-year chronic rat study all showed litter size, and reduced viability index. than 0.3 mg per person per year can be no effect levels higher than that observed Bal (40) reported inhibition of sperma- obtained from 60 kg of fruits. This would for the rabbit. togenesis, degeneration of spermatogonia correspond to a daily intake of less than Methoxychlor is considered to have and spermatocytes, and cytoplasmic vacuo- 0.000015 mg/kg bw/day for a 60-kg person. estrogenic activity. Kupfer and Bulger (36) lation in the epithelium of the epididymis Dicofol was evaluated by JMPR for found that both methoxychlor and metabo- in male rats after the administration of 100 acceptable daily intake in 1992. An ADI of lites have estrogenlike activity with several to 200 mg/kg bw/day methoxychlor. A 0 to 0.002 mg/kg bw was established based metabolites having proestrogen activity. decrease in seminal vesicle and caudal epi- on the no effect level of 0.22 mg/kg Gray et al. (37) investigated the effects didymal weight and caudal sperm count as bw/day in a rat long-term study. At 2.2 of methoxychlor on the pubertal develop- well as delayed puberty were observed in mg/kg bw/day, histopathological changes ment and reproductive function in the male neonatal rats administered 25 to 200 in the liver and vacuolation of adrenal and female rat by dosing rats from gesta- mg/kg bw/day methoxychlor for one gen- cortical cells were observed (35). tion, weaning, lactation, through puberty eration, indicating that the endocrine func- Dicofol is structurally related to DDT, with either 25, 50, 100, or 200 mg/kg tion of the testes and pituitary gland were and therefore tends to accumulate in the bw/day of methoxychlor. In females, an affected (37). Cooke and Eroschenko (41) mammalian organism, however at a rate acceleration of vaginal opening, abnormal also noted that the development of the much less than that of DDT and its estrous cycle, inhibition of luteal function, neonatal male rat reproductive tract was metabolites. and a blockage of implantation was inhibited by methoxychlor administration, Dicofol primarily affects the liver in observed. In males, an inhibition of somatic as shown by a decrease in serum testos- experimental animals. It is an inducer of growth and accessory sex gland weight, ele- terone levels and decreased DNA content of the hepatic mixed-function oxidase (MFO) vated pituitary and serum prolactin levels, the seminal vesicles, bulbourethral glands, activity in rats and mice. In the rat, an and a suppression of testicular Leydig cell and the ventral prostate. Rats fed 2000 increase in the incidence of thyroid follicu- function were found. Some of these effects ppm methoxychlor for 90 days exhibited lar cell hypertrophy in one study could not occurred at levels as low as 25 mg/kg decreased prostate size and cell content be reproduced in other studies. In the dog, bw/day. These observations are consistent (42). Goldman et al. (38) hypothesized a NOEL was found to be 0.82 mg/kg with the earlier reports that methoxychlor that part of methoxychlor's effects on male bw/day (1-year study) based on liver mimics estrogen both in vivo and in vitro. reproductive function may be mediated by changes and reduced cortisol response to Goldman et al. (38) investigated the a prolactinemic effect, since administration ACTH at 5.7 mg/kg bw/day (35). subchronic effects of methoxychlor on the of 25 or 50 mg/kg/day methoxychlor to 21- Dicofol increased the incidence of liver rat (Long-Evans hooded) reproductive sys- day-old male rats caused an increase in adenomas in male mice at high doses, but tem by dosing for 8 weeks with 25 or 50 serum prolactin levels and an increase in not in female mice. Dicofol was not mg/kg of methoxychlor by oral gavage. No hypothalamic gonadotropin-releasing carcinogenic in the rat. effect was observed on the pituitary weight, hormone levels. In a two-generation reproduction study serum LH, FSH, or prolactin levels and the Several chronic oral carcinogenicity in rats, the NOEL was 0.5 mg/kg bw/day pituitary LH or FSH concentrations. bioassays have been conducted with based on increased incidences of ovarian Pituitary prolactin levels were increased at methoxychlor, the results of which have stromal cell hyperplasia and hepatocellular both dose levels. There was an increase in been equivocal. changes at 2.5 mg/kg bw/day. Viability of GnRH levels in the mediobasal hypothala- A number of studies show that meth- offspring was reduced at higher doses. The mus at the high-dose level. The authors oxychlor does not accumulate in the body

790 Environmental Health Perspectives - Vol 104, Supplement 4 - August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS to any appreciable degree, but accumula- estrogen and progesterone were simul- on endocrine control systems. P-HCH tion of methoxychlor in fat has been taneously given to pregnant mice that were did not displace 17,-estradiol from its observed following administration of very fed lindane during early pregnancy, the receptors. The mechanism of the effect high dietary levels. Methoxychlor is metab- fetal development became normal. It thus is unclear. olized in the liver to readily excretable polar appeared that lindane caused a hormone Positive or marginally positive tumori- compounds. The results of several studies deficiency, resulting in reproductive and genic responses, characterized as benign indicate that methoxychlor does not induce developmental failure (44). hepatomas or hepatocellular carcinomas, hepatic microsomal enzymes (5). /3-Hexachlorocyclohexane. f-Hexa- have been observed in two strains of Lindane (y-Hexachlorocyelohexane or chlorocyclohexane (P-HCH) is a byprod- mice (45). y-HCH). The intake of lindane has been uct in the manufacture of lindane. 3-HCH DDT and Its Metabolites, DDE and estimated at less than 0.00003 mg/kg has been found in the air over oceans at DDD. The use of DDT in the Western bw/day for a 60-kg person. 0.004 to 0.13 ng/m3. P-HCH is the most world is now severely restricted. However, Lindane tends to accumulate in body persistent HCH isomer; bioconcentration it is still extensively used in some develop- fat of mammalian species. More than 90% takes place in invertebrates, fish, birds, and ing countries against malaria. Technical of the intake of lindane is estimated to man. In mammals the compound is stored DDT contains mainly the p,p'-isomer originate from foods. Studies of the total in adipose fat (45). (75-80%); o,p'-DDT constitutes 10 to diet in the United States estimated the Food is the main exposure route for the 25% of the technical product. The DDT intake up to 0.00005 mg/kg bw/day in general population. Total dietary studies metabolite DDE is more persistent than 1970 and up to 0.000003 mg/kg bw/day from United Kingdom in 1966 to 1967 the parent compound, and is therefore the in 1980 (43). found 0.003 mg/kg, in 1975 to 1977 form of DDT normally found in foods and Lindane was last evaluated by JMPR 0.0005 mg/kg, and in 1981 <0.0005 human tissues. Following the severe restric- for acceptable daily intake in 1989. An mg/kg of total diet. The average daily tions in the use of DDT in the late 1960s, ADI of 0 to 0.008 mg/kg bw was estab- intake in the United States in 1982 to a dramatic decline has occurred in the con- lished based on the following no-effect lev- 1984 ranged from less than 0.0000001 to tamination of foods and humans. For els: 0.75 mg/kg bw/day in the rat; 1.6 0.0000004 mg/kg bw/day for various age example, levels in herring have declined mg/kg bw/day in the dog (34,43). groups (45). from 5 to 7 mg/kg fat in 1973 to 0.5 to 1 An RfD for noncarcinogenic effects was ,B-HCH is included in the Danish mg/kg fat in 1993 (2). In 1986, DDE lev- established by the U.S. EPA at 0.0003 pesticide control program. As mentioned els in human milk in Denmark ranged mg/kg bw/day based on a no effect level of above, the intake has been estimated at from 0.29 to 1.21 mg/kg fat (21). 0.33 mg/kg bw/day in a chronic feeding less than 0.00003 mg/kg bw/day for a DDT and its metabolites are included study in rats. At 1.55 mg/kg bw/day, kid- 60-kg person. in the Danish pesticide control program. ney and liver toxicity was seen. The safety No ADI or RfD has been established In 1993, the typical levels found in animal factor was 1000 (5). for f-HCH. The most sensitive organ to and fish lipids were about 0.02 mg/kg. It In a 2-year bioassay in dogs, a NOEL the toxicity of P-HCH is the liver. In the has been estimated that an individual who for liver effects was determined to be 1.6 rat, hypertrophy and proliferation of consumes a large amount of fish is exposed mg/kg bw/day. Lindane is a well-known smooth endoplasmic reticulum and to less than 2.5 mg of DDT and its inducer of the liver MFO system. Lindane increased activity of microsomal enzymes metabolites per year. This corresponds to induced hyperplastic nodules and/or hepa- were seen after a dose 2.5 mg/kg bw/day less than 0.0001 mg/kg bw/day. tocellular adenomas in mice, especially in for 90 days. After 12.5 mg/kg bw/day, JMPR in 1984 established an ADI at 0 males given high doses in long-term changes in the gonads (testicular atrophy) to 0.02 mg/kg bw/day for any combination studies. Lindane did not increase the inci- were reported, but associated hormonal of DDT, DDD, and DDE (17). dence of liver tumors in rats (43). changes showed no consistent endocrine An RfD for noncarcinogenic effects was Lindane had no teratogenic effect in effect. There were no adverse effects established by the U.S. EPA at 0.0005 mice, rats, dogs, and pigs. Fetotoxic and/or reported after 0.1 mg/kg bw/day. In a mg/kg bw/day based on a no effect level of maternal toxic effects were observed with long-term study, liver enlargement and his- 0.05 mg/kg bw/day in a 27-week rat feed- doses of 10 mg/kg bw/day and above. A tological changes were seen at 0.5 mg/kg ing study. At 0.25 mg/kg bw/day, liver dose of 5 mg/kg bw/day was considered a bw/day (45). In a two-generation repro- lesions (hepatocellular hypertrophy) were no effect level. In a three-generation reproduction study in rats, parental liver effects observed. The safety factor was 100 (5). duction study in rats, lindane had no effect were also reported. No effects on reproduc- In a three-generation rat reproduction on reproduction or maturation at doses up tion were seen after 0.1 mg/kg bw/day, study, offspring mortality increased at all to 5 mg/kg bw/day. At 2.5 mg/kg bw/day, whereas 0.5 mg/kg bw/day resulted in dose levels, the lowest of which corre- liver changes indicative of MFO enzyme increased mortality and infertility. The sponds to about 0.2 mg/kg bw/day. Three induction occurred in the offspring of the parental animals at 2.5 mg/kg bw/day other reproduction studies (rat and mouse) third generation. The no effect level in this showed decreased weight of ovaries and showed no reproductive effects at much study was 1.25 mg/kg bw/day (43). increased weight of adrenal glands and higher dose levels (5). When given to pregnant mice at various uterus in females. In males, the testes Nine feeding studies, including two stages of pregnancy, lindane (3.6-10.4 showed a reduced number of Leydig cells. multigenerational studies, have been mg/kg bw/day) caused reproductive failure A weak estrogenlike effect after a dose of 25 conducted in mice. Only one of these and fetotoxicity. The effects were partly mg/kg bw/day for 5 days has been described studies, conducted for 78 weeks, showed corrected by co-administration of estrogen, in female mice and rats. The uterus was the no indication of DDT tumorigenicity. while progesterone had no effect. When target organ, and there were no clear effects However, both hepatocellular adenomas

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 791 TOPPARI ET AL.

and carcinomas were observed in six mouse bio-assays. At 0.273 mg/kg bw/day, 0.175 mg/kg bw/day. In a three-generation studies. Benign and malignant lung tumors regional liver hypertrophy was observed in reproduction study in rats the NOEL was were observed in two studies in which mice females. The safety factor was 1000 (5). 0.25 mg/kg bw/day. At 0.5 mg/kg bw/day, were exposed both in utero and throughout In a 24-month chronic toxicity study in increased pup mortality was recorded (5). their lifetime. Doses leading to the increased the mouse, a NOEL of 0.15 mg/kg bw/day Heptachlor epoxide has induced liver tumor incidence ranged from 0.15 to 37.5 was observed based on hepatocellular carcinomas in two strains of mice of both mg/kg bw/day. Three studies using rats swelling and necrosis in males and sexes. In rats no significant carcinogenic and doses from 25 to 40 mg/kg bw/day increased liver weight in males and females effect was seen. produced an increased incidence of benign at 0.75 mg/kg bw/day (5). Chlordane has Most studies have indicated that liver tumors. Another study was negative, been studied in four mouse and four rat heptachlor epoxide is not mutagenic. as were three additional assays in which long-term carcinogenesis bioassays. Dose- Dieldrin. Dieldrin is the major lower doses were given (5). related incidences of liver carcinoma con- metabolite of aldrin. Both compounds Studies of DDT exposure in hamsters stitute the major finding in mice. In the have been severely restricted or banned have not shown an increased tumor inci- rat, chlordane produced a significant since the early 1970s in a number of coun- dence. Unlike mice and humans, hamsters increase in adenomas of the liver of male tries. Dieldrin accumulates in the mam- accumulate DDT in tissue but do not rats. Although no tumors were observed in malian organism. Aldrin is rarely found in metabolize it to DDD or DDE. Studies of female rats, hepatocellular swelling was food. Dieldrin levels in food (total diet DDT in dogs and monkeys have not significantly increased. Other studies in the study, United Kingdom) have been declin- shown a carcinogenic effect. However, the rat have been negative (5). ing from 0.004 mg/kg in 1966 to 1967 to length of these studies was insufficient to Most studies have found chlordane 0.0005 mg/kg in 1981 (46). In 1986, assess the carcinogenicity of DDT (5). not to be genotoxic. No adequate studies dieldrin levels in human milk in Denmark p,p'-DDE exposure increased the inci- on reproduction and teratology are avail- ranged from less than 0.01 to 0.04 mg/kg dence of liver tumors including carcinomas able (5). fat (21). The intake has been estimated in males and females of two strains of mice Heptachlor was evaluated by JMPR in at less than 0.00003 mg/kg bw/day for a and in hamsters. In the rat, a dose-depen- 1991 when an ADI of 0 to 0.0001 mg/kg 60-kg person. dent, but not statistically significant, bw was established for heptachlor and hep- JMPR in 1977 established an ADI at 0 increase in thyroid tumors in females was tachlor epoxide combined (23). An RfD to 0.0001 mg/kg bw/day for the combined not considered convincing evidence of for noncarcinogenic effects was established total ofdieldrin and aldrin (22). carcinogenicity (5). DDE was mutagenic by U.S. EPA at 0.0005 mg/kg bw/day An RfD for noncarcinogenic effects was in mouse lymphoma cells and Chinese based on a no effect level of 0.15 mg/kg established by the U.S. EPA at 0.00005 hamster (V79) cells, but not in Salmonella. bw/day in a 2-year chronic feeding study in mg/kg bw/day based on a no effect level of p,p'-DDD induced an increased inci- rats. At 0.25 mg/kg bw/day, increased liver 0.005 mg/kg bw/day in a 2-year feeding dence of lung tumors in male and female weight was seen in males. The safety factor study in rats. At 0.05 mg/kg bw/day, mice, liver tumors in male mice, and was 300 (5). increased liver weight and liver lesions thyroid tumors in male rats (5). In an inadequate three-generation repro- (parenchymal cell changes including focal Chlordane, Heptachlor, and Hepta- duction study, no effects were observed proliferation and focal hyperplasia) were chlor Epoxide. At one time, these insecti- after 0.5 mg/kg bw/day (5). observed. The safety factor was 100 (5). cides were widely used in the United State, Heptachlor induced hepatocellular car- Liver toxicity of dieldrin has also been but now the use of chlordane is suspended, cinomas in male and female mice in two observed in the mouse, hamster, dog, and and that of heptachlor restricted. Technical studies. No indication of treatment-related monkey in chronic feeding studies (46). chlordane contains a mixture of various increase of tumors has been reported in In most of the reproduction studies chlordane isomers and heptachlor (chlori- chronic studies with rats. (over one to six generations) carried out nated cyclodienes). Oxychlordane and hep- Most studies have indicated that with aldrin or dieldrin on mice and rats, tachlor epoxide are persistent metabolites heptachlor is not mutagenic. the major effect was an increased mortality of chlordane and heptachlor, respectively. Heptachlor epoxide was evaluated by rate in pups not yet weaned. Reproductive Levels of heptachlor/heptachlor epoxide JMPR in 1991 when an ADI of 0 to performance was only affected at doses in human milk fat in the 1 970s ranged from 0.0001 mg/kg bw was established for causing maternal intoxication. It was con- 0.10 to 0.35 ppm in Europe and America, heptachlor and heptachlor epoxide com- cluded, that 0.1 mg/kg bw/day was a whereas chlordane and oxychlordane levels bined (23). An RfD for noncarcinogenic NOAEL in the rat (46). are low and mostly below the detection effects was established by the U.S. EPA at Dieldrin was carcinogenic in seven limit in Europe (21). As mentioned earlier, 0.000013 mg/kg bw/day based on a lowest strains of mice when administered orally. the intakes have been estimated at less than effect level of 0.0125 mg/kg bw/day in a Dieldrin increased the incidence of 0.00003 mg/kg bw/day for a 60-kg person 60-week chronic feeding study in dogs. hepatomas and hepatic carcinomas. It also for each ofthese compounds. The effect seen was increased liver-to-body produced significant increases in the inci- Chlordane was evaluated by JMPR in weight in both males and females. The dence of pulmonary adenomas, pulmonary 1986 when an ADI of 0 to 0.0005 mg/kg safety factor was 1000 (5). carcinomas, and lymphoid tumors. How- bw was established (25). An RfD for non- A NOEL of 0.025 mg/kg bw/day was ever, in the rat, seven studies with four carcinogenic effects of chlordane has been established from a two-generation repro- strains of rats fed 0.1 to 285 ppm dieldrin established by the U.S. EPA at 0.00006 duction study in dogs. At 0.075 mg/kg and varying in duration of exposure from mg/kg bw/day based on a no effect level of bw/day, liver lesions were observed in 80 weeks to 31 months did not produce 0.055 mg/kg bw/day in a rat chronic pups. Reduced pup survival was seen after positive results for carcinogenicity (5).

792 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

Although positive in a few assays in and carcinomas) was seen in both male and Aldicarb was not found to induce signi- vitro for clastogenic activity, most studies female rats. ficant increases in tumor incidence in mice in vitro and in vivo on the genotoxicity of Toxaphene was mutagenic in bacteria, or rats in feeding studies or mice in a skin- aldrin and dieldrin have yielded negative but negative in a modified dominant lethal painting study. In the feeding studies there results (46). assay ofmale mice (5). were, however, significant changes in the Mirex. Mirex was commonly used in incidence of pituitary tumors in female rats the United States in the 1960s and 1970s Nematocides and fibrosarcomas in the male mouse (5). to control fire ants; it was also used in plas- Aldicarb. The intake of aldicarb through Aldicarb has not been shown to be tics, rubber, paints, paper, and electrical foods is estimated to be very low and genotoxic in any of a variety of in vitro and goods, but its use is now banned. The insignificant. Aldicarb does not accumulate in vivo genotoxicity assays (35). intake has been estimated at less than in the body. 1,2-Dibromo-3-chloropropane. No 0.00003 mg/kg bw/day for a 60-kg person. In 1992 JMPR established an ADI at information on exposure to 1,2-dibromo- An RfD for noncarcinogenic effects was 0.003 mg/kg bw based on a no effect level 3-chloropropane (DBCP) is available. established by the U.S. EPA at 0.0002 of 0.025 mg/kg bw in humans. The no DBCP has not been evaluated by mg/kg bw/day based on a no effect level of effect level was based on decrease of acetyl JMPR and an oral RfD is not available 0.07 mg/kg bw/day in a chronic feeding cholinesterase (Ache) activity in erythro- from the U.S. EPA. study in rats. The effects seen at at a dose cytes at 0.05 mg/kg bw (35). An inhalation RfD for noncarcinogenic level of 0.7 mg/kg bw/day were liver hyper- An RfD for noncarcinogenic effects was effects was established by the U.S. EPA at trophy and thyroid cystic follicles. established by the U.S. EPA at 0.001 0.0002 mg/m3 based on a no effect level of Histologic examinations revealed dose- mg/kg bw/day based on a no effect level of 0.17 mg/m3 in a 13-week, subchronic rab- related changes in the parathyroid gland, 0.01 mg/kg bw/day in acute oral exposure bit inhalation study. At 1.7 mg/m3 testicu- kidney, liver, spleen, and thyroid. The studies in humans. Essentially all hazards lar effects were observed. The safety factor safety factor was 300 (5). Effects of mirex from aldicarb exposure are associated with was 1000 (5). In that study, the average on the liver and thyroid have been reported cholinesterase inhibition. At 0.025 mg/kg sperm count as well as the percentage of in many other studies. bw/day, clinical signs of acetylcholinester- live sperm of the rabbits was significantly Effects on the testis have been reported. ase inhibition were observed. The safety less than those of the controls after 7 weeks Histologic evaluation of the testis revealed factor was 10 (5). of exposure, and remained decreased for the hypocellularity, decreased spermatogenesis, Available evidence both from experi- duration of the exposure period and and luminal nucleated and giant cells, all of mental animals and from humans exposed through week 42 after the exposure. To which are characteristic of testicular degen- to aldicarb suggests that neurobehavioral assess the effects of DBCP on fertility, eration. Reproductive and developmental effects are short lasting with no accumula- exposed male rabbits were mated to unex- effects (decreased fertility, fetal cataracts, tion of effects over time. Thus, the doses posed female rabbits at weeks 14 and 41 of edema) have been reported in several producing effects following repeated daily the study. DBCP did not affect the libido studies. Perinatal and postnatal exposure to exposure are comparable to those following of the exposed male rabbits during week mirex resulted in significantly decreased a single dose. Also, a comparable degree of 14; however, the five males exposed to 17 pup survival (5). cholinesterase inhibition is seen at the same mg/m3 DBCP were infertile. During week Mirex is an inducer of liver microsomal dose levels, whether delivered in one acute 41 (27 weeks postexposure), all rabbits mixed function oxidases. dose or by subchronic or chronic dosing (5). exposed to lower concentrations of DBCP Toxaphene (Camphechlor). Toxa- The results of a 2-year feeding study in produced normal litters, and two of the five phene is a mixture of several hundred poly- the rat demonstrated that the exposed ani- males exposed to 17 mg/m3 regained fertil- chlorinated terpenes. Toxaphene is a global mals did not differ significantly from con- ity (i.e., increased sperm count) and pro- pollutant spread by long-range air trans- trols for any of a number of toxicological duced normal litters. The FSH serum levels portation. Although hardly used in parameters, except the inhibition of were also significantly elevated at 14 weeks Sweden, it was detected in two pooled cholinesterase. Therefore, the NOAEL for in the males exposed to 1.7 mg/m3 DBCP. human milk samples; the level calculated systemic toxicity is greater than or equal to Increased FSH serum levels were consistent from 22 peaks was 0.1 ppm in the milk fat. 0.3 mg/kg bw/day (5). with a marked decrease in sperm count, The intake has been estimated at less In two multigeneration reproduction whereas serum levels of testosterone were than 0.00003 mg/kg bw/day for a 60-kg studies in rats, the lowest NOEL for feto- unchanged. The only gross lesion observed person. Toxaphene was evaluated by JMPR toxicity was found at 0.3 mg/kg bw/day. under macroscopic examination was the in 1973. No ADI was established (22). No No reproductive effects were noted at any small size of the testes. A histopathologic RfD has been established by U.S. EPA. dose tested. Decreased body weight of sec- examination revealed changes in the repro- Dietary toxaphene administered to ond-generation pups was observed at 0.7 ductive system. These effects included mice at doses higher than 50 ppm caused mg/kg bw/day. Reduced pup viability was atrophy of the testes, epididymides, and dose-related increased incidences of hepa- observed at lactation day 4 at higher doses. accessory sex glands including the prostate. tocellular carcinomas and adenomas in Developmental toxicity studies in rats The testes weight was significantly decreased both sexes (5). and rabbits revealed no congenital malfor- to 50% of control values (week 14) in the In rats receiving toxaphene at doses of mations but showed feto/embryotoxicity at group exposed to 1.7 mg/m3 and to 75% of 556 and 1112 ppm for males and 540 and maternally toxic doses. The lowest NOEL control values (week 8) in the group exposed 1080 ppm for females for 80 weeks, a sta- for developmental effects was 0.125 mg/kg to 17 mg/m3. Severe testicular atrophy was tistically significant dose-related increased bw/day in the rat and 0.25 mg/kg bw/day characterized by nearly complete or complete incidence of thyroid tumors (adenomas in the rabbit. loss of spermatogenic cells in seminiferous

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 793 TOPPARI ET AL.

tubules. Following the recovery period, or 5000 ppm of bisphenol-A and female interaction with the Ah receptor, and they tubular regeneration was observed in the mice (50/group) were fed 0, 5000, or are potent inducers of certain cytochrome testes ofthe affected rabbits (5). 10,000 ppm of bisphenol-A. Male mice at P450 isozymes (CYPlAl and CYP1A2). A number of occupational studies on 5000 ppm and female mice at 5000 and The toxicity of such PCB congeners can be exposure to DBCP demonstrated this com- 10,000 had reduced body weights. At 1000 expressed in terms of TEFs, that is, pound to be a potent testicular toxicant in and 5000 ppm, there was an increase in the expressed as a fraction of the toxicity of humans. The azoospermic men had ele- number of multinucleated giant hepatocytes TCDD (51,52). Other PCB congeners vated FSH, but normal LH and testos- in male mice. This effect was not consid- presumably act by different mechanisms terone levels. Testicular biopsy showed ered adverse, and an NOAEL in mice was and are potent inducers of a different set of atrophy of seminiferous epithelium and established at 130 mg/kg bw/day (5). cytochromes (CYP2B1 and CYP2B2). In tubules lined by Sertoli cells only. Few reli- The only toxic effect seen in beagle addition, there are PCB congeners that are able exposure data are available from any of dogs fed 1000 to 9000 ppm of bisphenol-A intermediate in this respect, i.e., they elicit a these studies. Limited follow-up studies of in the diet for 90 days was an increase in mixed spectrum of effects. Our knowledge affected worker populations indicate that mean liver weight in the high-dose group. of possible interactions between the various paternal exposure to DBCP sufficient to In a two-generation feeding study of groups of PCBs is still very limited (50). produce oligoospermia or azoospermia did bisphenol-A (100-9000 ppm) in rats, PCB levels in foods (herring and cod not detectably increase the rate of congeni- decreases in body weight in the parent gen- liver) and human milk have declined in the tal malformations or impair the health sta- eration at 9000 ppm and in the F1 genera- Nordic countries since the 1970s (2,50). tus of offspring conceived during or after tion at greater than or equal to 1000 ppm PCB levels in fish now seem to have stabi- DBCP exposure (5). were observed. In mice, a dose level of lized at 10% of the levels found in the DBCP is, however, a potential mutagen 1250 mg/kg bw/day was associated with beginning of the 1970s. In fact, the levels capable of inducing a dominant-lethal fetotoxicity and maternal toxicity, but in herring are now below 0.1 mg/kg. It has effect in mice. there was no significant increase in the been estimated that the intake of PCBs Chronic inhalation carcinogenesis incidence of malformations at any dose from heavy fish consumption is less than bioassays in rats and mice have established level. In rats, dose levels of less than or 25 fg per person per day, corresponding to DBCP as a carcinogen, with high inci- equal to 1280 mg/kg/day were not toxic 0.0004 mg/kg bw/day for a 60-kg person. dences of tumors appearing in the nasal and did not cause malformations to the A Nordic group of experts concluded cavity and on the tongues of rats and in the fetus (5). that the present database does not allow a nasal cavity and lungs of mice (5). Polychlorinated Biphenyls. The poly- traditional risk assessment to be performed, chlorinated biphenyls (PCBs) are a series of i.e., it was not possible to recommend a Industrial Chemicals and 209 congeners. They were marketed under tolerable daily intake of either total PCB Environmental Poliutants various trade names (Aroclor, Clophen A, or any individual congener. Furthermore, Bisphenol-A. Bisphenol-A is used as a plas- and Kanechlor). Within each series there the evaluation suggested that the present tic monomer in the manufacture of epoxy, were different average chlorine contents in exposure of Nordic populations is of the polycarbonate, and polyester-styrene the mixtures. PCBs had diverse uses in same order of magnitude as that at which resins. Little is known about exposure to electrical transformers and other electrical subtle health effects may occur in children bisphenol-A. Recently, bisphenol-A was appliances (49), but their use has been exposed in utero and possibly through found to be released from the epoxy-lac- severely restricted in most countries since breast-feeding (see "Exposure of Humans quer coating in food cans. Concentrations the 1970s and 1980s. Waste containing to Environmental Chemicals with from 0.004 to 0.023 mg per kg of food >50 ppm PCBs should be treated as haz- Estrogenic Activity and Their Effects on have been detected in canned food (47). ardous waste and destroyed by combustion Male Reproductive Health"). Further A tolerable daily intake of bisphenol-A (50). Many of the PCB congeners are very studies would be necessary to clarify was established at 0.05 mg/kg bw by the stable in the environment and accumulate whether such effects actually occur (50). In EU Scientific Committee for Food and a in marine, avian, and mammalian species. addition, when the concentrations of the specific limit for migration of bisphenol-A Foodstuffs, in particular fatty fish, meat, dioxinlike PCBs are taken into account, from food packaging materials has been set and dairy products, are important sources the joint risk from PCBs in mothers' milk at 3 mg per kg of food (48). of human exposure to PCBs. Until and certain other foods appears to be of An RfD for chronic oral exposure of recently, most analytical procedures for similar importance as the dioxin content. bisphenol- A was established at 0.05 mg/kg PCBs in food and human tissue have mea- Most of the animal studies with PCB bw/day by the U.S. EPA (5). sured total PCBs and do not reflect differ- mixtures have been performed using com- In both evaluations an LOAEL of 1000 ences in composition. Modern analytic mercial mixtures that differ in composition ppm in the diet (corresponding to 50 techniques are now able to provide con- from the PCB mixtures to which humans mg/kg bw/day) of rats in a 103-week gener-specific information, but so far only are exposed through food. Although ani- chronic oral bioassay was used applying a a limited database has emerged. mal feeding studies demonstrate the car- safety factor of 1000. Groups of 50 rats of The biological effects caused by the var- cinogenicity, immunotoxicity, and each sex were fed diets containing 0, 1000, ious congeners differ, not only in potency reproductive toxicity of commercial PCB or 2000 ppm of bisphenol-A. All treated but also qualitatively. Several non- and preparations, it is not known which of the groups of rats had reduced body weights. mono-ortho-substituted PCB congeners PCB congeners in such preparations are Food consumption was also reduced. induce effects similar to those caused by responsible for these effects. In the same study, male mice (50/ chlorinated dioxins and dibenzofurans, i.e., PCBs (Aroclor 1260, Kanechlor 500 group) were fed diets containing 0, 1000, the toxicity is probably mediated through and Clophen A 60, mixtures containing

794 Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

more highly chlorinated biphenyls) have decreased brain dopamine, which has absorption of PCBs in humans after oral produced hepatocellular carcinomas in been observed in monkeys orally exposed exposure, and indirect evidence of oral three strains of rats and two strains of mice. to Aroclor 1016 (5). The issue of male absorption of PCBs by humans is available Aroclor 1254 (lower chlorine percentage) reproductive effects was not addressed in from studies of ingestion of contaminated only produced hepatocellular adenomas in these studies. fish by the general population. PCBs dis- mice and rats (5). Most genotoxicity assays In the rat, Aroclor 1016 produced tribute preferentially to adipose tissue and of PCBs have been negative. elevated serum levels of adrenal cortical concentrate in human breast milk due to PCB exposure has been reported to hormones, and Aroclor 1016 was immuno- its high fat content (5). cause several effects on different processes toxic in mice. The estrogenic activity of hydroxylated in reproduction and the development of Aroclor 1016 (3.8 mg/kg bw/day for PCB metabolites has been discussed in the the embryo/fetus/offspring. Most effects 237 days) adversely affected reproduction "Environmental Chemicals with Known are specific to fertility and to the develop- in mink. The investigators noted that the Estrogenic Effects." ment of offspring. In addition, teratogenic adverse effects on reproduction did not Two episodes of ingestion of PCB-con- effects, such as hydronephrosis and cleft appear to be due to an effect on spermato- taminated rice oil have been reported: the palate, have been produced in mice. Effects genesis, since PCB-treated male mink have Yusho incident of 1968 in Japan and the seen on male reproduction include decreases had acceptable levels of reproduction Yu-Cheng incident of 1979 in Taiwan (see in fertility, matings and pregnancies; testic- when mated to untreated females in other "Exposure of Humans to Environmental ular spermatozoa concentration; ventral studies (5). Chemicals with Estrogenic Activity and prostate weight; and seminal vesicle weight. In rhesus monkey infants whose mothers Their Effects on Male Reproductive The underlying cause of some of the repro- were or had been exposed to Aroclor 1248 Health"). There is strong evidence indicat- ductive effects of PCBs may be alterations during gestation and lactation, behavioral ing that the health effects seen in Yusho in hormonal levels and/or receptor affini- testing showed hyperactivity and retarded victims were due to ingestion of polychlori- ties/levels. Decreased levels of gonadal hor- learning ability at 6 to 24 months of age. nated dibenzofurans, rather than to PCBs mones can be explained by enhanced However, at 44 months of age the monkeys themselves (5). metabolism of steroids, which are the nor- were hypoactive and at 4 to 6 years of age Polychlorinated Dibenzo-p-dioxins mal substrates for microsomal enzymes they showed impairment in learning and and Dibenzofurans. Polychlorinated induced by PCBs (50). In addition, effects memory tests. These effects were reported dibenzo-p-dioxins and polychlorinated on the thyroid gland and its function have at doses of about 0.006 mg Aroclor dibenzofurans consist of 75 and 135 dif- frequently been reported (52). 1248/kg bw/day to the mothers (50). ferent congeners, respectively. The most Perinatal toxicity and long-term neu- Adult female rhesus monkeys fed 0, 0.1 toxic and well-studied congener is TCDD. robehavioral effects of Aroclor 1016 have and 0.2 mg/kg bw/day ofAroclor 1248 for These compounds are not used commer- been evaluated in infant monkeys. Aroclor up to 14 months showed skin changes, cially but are formed as unwanted byprod- 1016 was administered to groups of eight such as hyperpigmentation and alopecia, ucts in a variety of chemical and thermal adult female rhesus monkeys via the diet at characteristic signs of PCB intoxication. processes, such as production of chlori- levels of 0.007 and 0.028 mg/kg bw/day Increased menstrual duration was noted. In nated compounds, incineration processes, for approximately 22 months. Exposure a breeding trial a highly increased rate of paper and pulp bleaching, and emissions began 7 months prior to breeding and con- abortion was observed, and liveborn infants from steel foundries and from motor vehi- tinued until offspring were weaned at age 4 showed clinical signs of PCB toxicity and cles (53). These chemical and thermal months. No exposure-related effects were died at weaning. Thymic atrophy was a processes lead to continuous formation seen in the mothers, who all had uncom- common lesion in such infants. The study and release of PCDDs and PCDFs into plicated pregnancies. Mean birth weights included three recovery breeding periods the environment. of the infants in the 0.028 mg/kg bw/day after dosing: at 22 months, 36 months, Although many of the PCDD/PCDFs dose group were lowered. Effects occurring and 55 months after the initiation of are emitted into the environment, only in the offspring of these monkeys consisted Aroclor 1248 dosing, respectively. Results congeners with chlorine substitution, at of hairline hyperpigmentation at greater of this prolonged recovery period revealed least in positions 2, 3, 7, and 8, are found than or equal to 0.007 mg/kg bw/day, and impairment of reproductive function in in biological material. These stable, decreased birth weight at 0.028 mg/kg female rhesus monkeys lasting for more lipophilic substances accumulate in food bw/day. The results of neurobehavioral than 4 years after dosing ceased. In the chains. The stable PCDD/PCDFs share a tests in the monkey offspring at 14 months groups of infants for which birth weight common mechanism of action. They have and 4 to 6 years of age demonstrated data are available, a significant reduction in been shown to act via a cytosolic receptor, adverse learning deficits at the 0.028 mean birth weight for PCB-exposed the Ah receptor. The Ah receptor regulates mg/kg/day maternal dose. However, evalu- infants was evident. Severe immunological a number of cellular proteins, such as ation of these data is complicated by possi- deficiencies have also been observed in cytochrome P4501A1 (CYP lAI), and ble inconsistencies in the outcome of the monkeys after PCB exposure. most of the toxic responses following learning tests. Learning was impaired and Studies of individual congeners and TCDD exposure are believed to be medi- facilitated on different problems; perfor- PCB mixtures of higher chlorine content ated through the Ah receptor. To express mance was improved in the low-dose in animals indicate, in general, that PCBs the combined toxicity of the PCDD/ group, and no significant differences were are readily and extensively absorbed (75- PCDF mixtures to which humans are seen between either test group and the con- >90%) in rats, mice, monkeys, and ferrets. exposed, the concept of TCDD equivalen- trol group. There is some evidence to sug- A study of a PCB mixture containing 54% cies (TCDD-TEQs) has been developed. gest that deficits in learning are related to chlorine provides direct evidence of In a variety of short-term tests, the toxicity

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 795 TOPPARI ET AL.

of each single congener has been expressed through the binding to the Ah receptor. household products and in agricultural and as a fraction of the well-known toxicity of This binding is followed by translocation industrial applications. They are also used TCDD, the so-called TEFs (54). to the dioxin-responsive enhancer (DRE) as spermicides in contraceptive foams, jel- Food is considered to be the main elements on DNA for the expression of lies, and creams. The alkyl group typically source of exposure to PCDD/PCDFs. The many of the species- and tissue-specific, is a branched nonyl-, octyl- or dodecyl- most important food items are milk and toxicological and biological responses, chain. The ethoxylate chain may have 1 to milk products, meat and meat products, including increased expression of growth 100 repeating units. Nonylphenol ethoxy- eggs, and fish and fish products. Data from factors. However, there is some evidence lates (NPEs; nonoxynols) are the most Germany (1991) have estimated an average that some effects, e.g., alteration in thyroid common forms of the APEs. daily intake of TCDD-TEQs of 130.3 hormone metabolism, are not necessarily Alkylphenol, alkylphenol mono- and pg/person/day, corresponding to approxi- Ah-receptor-mediated. diethoxylates, and their carboxylic acids are mately 2 pg/kg bw/day for a 60-kg person The effect of TCDD on male repro- the products of the microbial breakdown (0.000000002 mg/kg bw/day) (55). duction in the rat has already been dis- of APEs and are present in surface waters Comparable results are obtained from a cussed in "Environmental Chemicals with and sediments. The alkylphenols are rela- more limited investigation of Danish food- Known Estrogenic Effects." TCDD is both tively persistent and bioaccumulate in the stuffs (56). TCDD-TEQs were estimated antiestrogenic and antiandrogenic. A com- lipid of living organisms (see section in samples of mothers milk from Denmark, prehensive review of a number of endo- "Environmental Chemicals with Known Sweden, and Norway in 1987. The levels crine and other effects of TCDD has been Estrogenic Effects"). In addition to the in the three Nordic countries were very given by Pohjanvirta and Tuomisto (57). obvious exposure from the use of spermi- similar, the mean values ranged from 17 to Pentachlorophenol. Pentachloro- cides and contact with products containing 20 pg/g on a fat basis (54). phenol (PCP) has been widely used as a APEs, it is possible that humans can be In 1988 a Nordic expert group estab- wood preservative. Dietary exposure to exposed to APE degradation products lished a TWI of 35 pg/kg bw/week for PCP is thought to be limited. An average through the water supply, sewage sludge TCDD (5 pg/kg bw/day) (54). The evalu- intake of 0.00015 mg/kg bw/day has been used for fertilizer, and aquatic organisms ation was based on a NOAEL of 1000 estimated. PCP may be present in indoor used as food. However, virtually nothing is pg/kg bw/day found in a chronic long- and outdoor air, but data on inhalation known about the magnitude of human term toxicity/carcinogenicity study in rats exposure are scarce (58). exposure to these compounds. and a three-generation reproduction study A RfD for noncarcinogenic effects was Nonoxynol-9 (9 ethylene oxide units), in rats. The safety factor used was 200. In established by the U.S. EPA at 0.03 mg/kg which is used as a spermicide, was rapidly both studies effects were observed following bw/day based on a no effect level of 3 absorbed from the vagina of rats (59). administration of 10,000 pg/kg bw/day. In mg/kg bw/day in a 2-year chronic feeding There is also evidence that nonoxynol-9 can the long-term study, liver toxicity and study in rats. At 10 mg/kg bw/day pigmen- be metabolized to nonylphenol, which after hyperplastic nodules were observed in tation of liver and kidneys was observed in glucuronidation was excreted through urine female rats. At a 10-fold higher dose, female females. The safety factor used was 100 and bile (60). Intravaginal administration rats developed hepatocellular carcinomas. In (5). The main organs or systems affected in of nonoxynol-9 in rats caused damage to the reproduction study 10,000 pg/kg animals after long-term exposure to low the vaginal epithelium and acute inflamma- bw/day resulted in decreased litter size, levels of PCP were the liver, kidney, tion in the vagina, cervix, and uterus (61). decreased fetal and neonatal survival, and nervous system, and immune system (58). When administered to rats during the first decreased weight gain in surviving pups. Numerous studies have investigated the week of pregnancy, nonoxynol was found An overwhelming biochemical and tox- teratogenicity of PCP administered orally to be embryotoxic (62). icological database exists on PCDDs and to rodents. These studies did not reveal ter- Meyer et al. (63) observed fetotoxicity PCDFs. These compounds produce a wide atogenic effects; however, feto/maternal of nonoxynol-9 at a high oral dose level spectrum of responses in experimental toxicity was seen at 30 mg/kg bw/day. (500 mg/kg bw/day) that also caused animals and, presumably, in humans. The Since PCP apparently does not cross the maternal toxicity. When given epicuta- toxic effects include teratogenic, repro- placental barrier, the observed fetotoxicity neously, no effect was seen. Nonoxynol-30 ductive, behavioral, neuroendocrine, may be a reflection ofmaternal toxicity (5). was without effects. Nonoxynol-9 was not immunotoxic, and hepatotoxic effects. The In mice, PCP exposure increased the mutagenic in the Ames test (63). tumorigenic actions seen in some rodents incidences of hepatocellular adenomas and Octylphenol and nonylphenol were most likely are not the result of a geno- carcinomas, adrenal medulla pheochromo- found to be estrogenic both in vitro and in toxic, but rather the resulst of a promo- cytomas and malignant pheochromocy- vivo (see "Environmental Chemicals with tional, and possibly hormonal, mechanism tomas, and/or hemangiosarcomas and Known Estrogenic Effects"). (53). The biochemical effects include hemangiomas in one or both sexes. In two Phthalate Esters. Since the 1920s induction of a number of cytochromes, chronic oral rat studies no significant phthalate esters have been used as plasticiz- most noticeably CYPlAl and CYP1A2. increase in tumor incidence as compared ers for polyvinyl chloride. The most widely Indirect effects, such as alterations of hor- with controls was observed. used phthalate ester is di(2-ethylhexyl)- monal metabolism leading to, for example, Mutagenicity data provides some indica- phthalate (DEHP) (comprising 50% of all altered sex hormone levels and lowered tion that PCP has clastogenic potential (5). phthalate ester plasticizers). It may account thyroid hormone levels may have an Alkylphenol Ethoxylates. Alkylphenol for 40% or more of the plastic. In addition influence on growth and development of ethoxylates (APEs) are nonionic surfac- to this ester, a range of phthalate esters is the organism. Most of the toxic and bio- tants widely used as detergents, emulsifiers, available and used in different plastics, logical responses are thought to be initiated wetting agents, and dispersing agents in depending on the intended use of the

796 Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

product. Examples are dibutyl phthalate mice may be more sensitive than hamsters addition, DBP showed some evidence of (DBP) and butylbenzyl phthalate (BBP). and monkeys. The effect was reversible. clastogenic activity in mammalian cells. It Phthalate esters are produced in enormous MEHP has toxic effects on Sertoli cells is hydrolyzed to monoesters. There is also quantities. They are present in many com- in vitro (65). evidence that DBP induces peroxisome mercially available plastics and are leach- Dietary levels of 0, 0.01, 0.1, and proliferation (5). able or volatilizable from such plastics. The 0.3% DEHP were administered to male An RfD for noncarcinogenic effects of environmental fate of phthalate esters after and female CD-I mice that were exam- butylbenzyl phthalate (BBP) was estab- their release from plastics has not been ined for adverse fertility and reproductive lished by the U.S. EPA at 0.2 mg/kg widely studied (64). effects using a continuous-breeding proto- bw/day based on an NOAEL of 159 mg/kg DEHP exists widely in the environ- col. DEHP was a reproductive toxicant bw/day in a 6-month feeding study in rats. ment and is found in most samples, in both sexes, significantly decreasing fer- The effect observed after 470 mg/kg including air, precipitation, water, sedi- tility and the proportion of pups born bw/day was increased liver weight. The ment, soil, and biota. Levels up to 300 alive per litter at the 0.3% level, and safety factor was 1000 (5). In the previ- ng/m3 have been found in urban air. River inducing damage to the seminiferous ously mentioned study all rats given 1417 sediment levels up to 70 mg/kg have been tubules. DEHP has been observed to be mg/kg bw/day BBP had small testes; 5/11 reported. DEHP also has been found in both fetotoxic and teratogenic. had soft testes; and 1/11 had a small various types of food, such as fish, shellfish, DEHP is a potent inducer of hepatic prostate and seminal vesicle. In addition, eggs, and cheese. The average dietary expo- peroxisomal enzyme activity. This effect is testicular lesions characterized by atrophy sure to DEHP was estimated at 300 fg/per- strongly associated with the hepatocarcino- of seminiferous tubules and aspermia were son/day in the United States in 1974. genic effect of DEHP. In contrast to rat also observed. At the lower dose levels Blood transfusions and other medical treat- hepatocytes, DEHP metabolites do not there were no observable effects on male ments using plastic devices may lead to produce peroxisome proliferation in reproductive organs. A male mating trial involuntary human exposure to DEHP human hepatocytes (65). In a long-term study was performed concomitantly with (65). When all sources were considered, carcinogenicity study, groups of rats were the toxicity study. Testicular atrophy was the total dietary intake of DEHP in the fed with 0, 300, or 600 mg DEHP/kg observed in male rats after 10 weeks of United States was estimated to be 5.8 bw/day for 103 weeks. Similarly, groups of exposure to 2875 mg/kg bw/day). In an mg/person/day. Maximum exposure to mice received DEHP in the diet for 103 older rat study no effects were reported total phthalate from packaging sources weeks. No clinical signs of toxicity were after 250 mg/kg bw/day, while liver alone in the United Kingdom was esti- observed in either rats or mice. A statisti- weights were increased in animals fed diets mated at 4.37 mg/person/day [cited by cally significant increase in the incidence of containing 500, 750, or 1000 mg/kg Sharman et al. (66)]. Sharman et al. (66) hepatocellular carcinomas and the com- bw/day, respectively, for 90 days. A mild found total phthalates in dairy products (in bined incidence of carcinomas and adeno- decrease in growth rate was reported for the United Kingdom) ranging from 4 to mas was observed in female rats and both the two top dose groups. 20 mg/kg in cheese and butter (DEHP sexes of mice. In a 14-day rat study, BBP at a dose from 0.6-3 mg/kg). Studies indicate that DEHP is not a level of 375 mg/kg bw/day produced The extent of human exposure to directly acting mutagen. MEHP, the significant increases in liver and kidney phthalate esters in Denmark is not known. monoester form of DEHP and a metabo- weights and kidney pathology (proximal In Denmark, DEHP is not allowed for use lite, was positive in several assays for muta- tubular regeneration). in plastic materials intended to come into genicity and clastogenicity (65). In male rats administered 160, 480, or contact with foods. Petersen (67) found a An RfD for noncarcinogenic effects of 1600 mg/kg bw/day BBP for 14 days, bio- mean concentration of DEHP lower than DBP was established by the U.S. EPA at 0.1 chemical or morphological changes in the 0.05 mg/liter in retail whole milk from mg/kg bw/day based on an NOAEL of 125 liver as well as effects on testes weights Denmark. mg/kg bw/day in a subchronic to chronic were not observed in the 160 mg/kg/day The European Union Scientific Com- feeding study in rats. The effect observed dose group. However, at 480 mg/kg mittee for Food has established an ADI of after 600 mg/kg bw/day was increased mor- bw/day, liver enzyme activities were 0.025 mg/kg bw/day for DEHP (67). An tality. The safety factor was 1000 (5). increased and testicular atrophy was RfD for non-carcinogenic effects of DEHP Fetotoxicity was observed when mice were observed in some of the rats. BBP was was established by the U.S. EPA at 0.02 fed 2100 mg/kg bw/day DBP throughout tested at dietary levels of 0, 6000, and mg/kg bw/day based on an LOAEL of 19 gestation. An increase in terata of border- 12,000 ppm in long-term carcinogenicity mg/kg bw/day in a subchronic to chronic line statistical significance was observed in studies in mice and rats. feeding study in guinea pigs. The effect progeny ofthis treatment group. BBP induced a statistically significant observed was increased liver weight. The DBP causes degeneration of the semi- increase in mononuclear cell leukemia in safety factor was 1000 (5). niferous tubules, probably as a result of female rats; the response in male rats was MEHP, the monoester form of DEHP increased urinary excretion of zinc. No inconclusive and there was no such is the principal metabolite of DEHP (65). data on carcinogenicity are available. response in mice. BBP did not induce lung Several studies have shown testicular DBP did not induce mutations in adenomas in strain A mice administered atrophy after DEHP administration (high bacteria. It was mutagenic in the mouse 24 ip injections of 160, 400, or 800 mg/kg doses). Younger rats seem to be more lymphoma forward mutation assay only in bw. BBP is not mutagenic in in vitro susceptible than older ones, and rats and the presence of metabolic activation. In assays (5).

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 797 TOPPARI ET AL.

Appendix B: Identification and Assessment and dead pups, number of pups with of the Effects of Chemicals on anomalies are recorded, and, if necessary, histological examinations of dead or sacri- Reproduction and Development, Focusing ficed animals are performed. The number on the Effects on Males of live pups on day 4 and at weaning are recorded and the following indices are often calculated: * Mating index: copulation/estrous cycles The following effects have to be considered: * Ministry of Health, The National Food required * Impairment of male (and female) Agency of Denmark (70) * Fecundity index: pregnancies/copulations reproductive functions or capacity, i.e., * Nordic Council of Ministers and * Male fertility index: males impregnat- adverse effects on libido, sexual behav- National Institute of Occupational ing females/males exposed to fertile ior, any aspect of spermatogenesis (or Health, Denmark (71) nonpregnant females oogenesis), or hormonal activity or * Hansen and Meyer (72) * Female fertility index: females conceiv- physiological response that would inter- * European Chemical Industry, Ecology ing/females exposed to fertile males fere with the capacity to fertilize, fertil- and Toxicology Centre (ECETOC) (73) * Incidence of parturition: parturitions/ ization alone, or development of the One-, Two-, and Multigeneration pregnancies fertilized ovum up to and including Studies. Generation studies examine suc- * Live birth index: viable pups born/pups implantation. cessive generations to identify possible born * Induction of noninheritable harmful effects of a substance on fertility of male * Survival index for 24 hr, 4 days, 12 effects on the progeny, i.e., in the and female animals; pre-, peri-, and post- days, and 21 days broadest sense any effect interfering natal effects on the ovum, fetus, and prog- The number of litters per female per with normal development, both before eny, including teratogenic and mutagenic generation varies in the different guidelines and after birth up to puberty, should effects; and peri- and postnatal effects on from one to two. be included. Both morphological mal- the mother. Fertility assessment by continuous formation(s) and functional distur- Various international organizations and breeding has been used to study the deple- bances (e.g., hormonal, neurological) countries have drafted guidelines for these tion of oocytes from the ovary in mice should be evaluated. tests, with a number of common require- exposed to procarbazine. In this study, pre- The risk assessment includes: ments. The preferred species are the rat and natally treated mice were housed continu- * Hazard identification the mouse; other species may be used if rele- ously with untreated male mice, and the * Dose-effect assessment (estimation of a vant (e.g., differences in the toxicokinetics cumulative number of offspring was mea- possible NOAEL) between the preferred species and man to sured by removing the female when she was * Extrapolation (prediction of adverse clarify ambiguous results or to further study noticeably pregnant and then returning the effects in other species, particularly observed effects). The test substances are female to the male's cage immediately after in man) administered by the appropriate route to the birth of her pup in order to establish a * Prediction of safe levels of exposure groups of animals (number of animals per pattern of forced repetitive breedings. in man group should be sufficient to yield about 20 In proposed guidelines from the U.S. pregnant animals at or near term). The EPA (74), samples of sperm from the distal Experimental Studies in chemical is administered over at least one cauda epididymis (or the proximal vas def- Laboratory Animals spermatogenic cycle and the last stages of erens) shall be collected for the evaluation Many different experimental methods are oocyte maturation before the parent genera- of the percentage of progressively motile being used for the investigation ofthe repro- tion animals are mated. The exposure of the sperm and sperm morphology. In addition, ductive toxicity of chemicals. Several tests females is continued throughout the mating the entire cauda epididymis shall be are standardized and guidelines have been period and the gestation up to weaning of minced in saline to enumerate the total issued by various governmental agencies the last generation. At least three treatment number ofsperm. and international organizations. Other tests groups and a control group (untreated or A number of useful tests of male repro- are still undergoing scientific evaluation. vehicle in the highest dose used) should be ductive toxicity, which are discussed in The group of tests discussed below is used. Ideally, unless limited by the physico- detail, for example, by Thomas (75), are not a comprehensive listing of all available chemical nature or biological effects of the listed below: tests but rather a presentation of represen- test substance, the highest dose level should * Testis: size in situ, weight, consistency, tative examples. (The presentation of the induce toxicity but not mortality in the morphometry, spermatid reserves, individual tests and the test strategies are parental animals. The low dose should ide- DNA flow cytometry, biochemical based on the references given at the end of ally not induce any observable adverse assays, gene expression this appendix) effects on the parents or offspring. * Epididymis: weight and histology, The animals are observed daily for clin- number of sperm in distal half, sperm Tests for Reproductive Toxicity ical changes. Body weight is recorded motility, sperm morphology, biochemi- The description of the tests is based on the weekly for the parent animals and for off- cal assays, gene expression following references: spring normally at birth, on days 4, 7, and * Accessory sex glands: weight and histol- * OECD (68) 14, at weaning, and thereafter every week. ogy, biochemical assays, gene expression * Commission of the European Pregnancy rate, duration of pregnancy, * Semen: total volume, gel-free volume, Communities (69) number of pups per litter, number of live sperm concentration, sperm motility,

798 Environmental Health Perspectives - Vol 104, Supplement 4 - August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

sperm morphology, hamster egg pene- rabbits) are used to establish a dose-effect to reflect the entire complex and intricate tration, cervical mucus penetration relationship. The pregnant female rats are function of behavior. For testing behavior, * Sperm (special features): video/cine- exposed during the period of organogene- therefore, a range of parameters-a test micrography, membrane structure, sis, i.e., between day 6 when implantation battery-is used to identify changes in metabolism occurs and day 15. (The corresponding individual functions. * Fertility: ratio-exposed/pregnant periods for mice and rabbits are days 6-15 The most frequently used animal women; number of embryos per preg- and days 6-18, respectively). This period species are the rat and the mouse. The nant female; ratio-viable embryos/ has been found to be the most sensitive to guidelines generally recommend groups of corpora lutea; number of 2- to 8-cell the induction of structural, anatomical 20 animals with dosing from day 15 ofges- eggs; number ofabnormal eggs malformations (the corresponding sensitive tation to day 21 postgestation, i.e., span- * Endocrine monitoring: testis-testos- period for humans is between days 18 and ning fetogenesis and the entire lactation terone, dihydrotestosterone, estradiol, 60 of pregnancy). Days 6 to 15 are the period. (However, the recommended dos- estrone, thyroid gland-thyroxin, indicated dosing period for pregnant rats. ing period does not cover all events since triiodothyronine pituitary-FSH, LH, However, this may vary depending on the the CNS is also susceptible to abnormal prolactin, TSH substance administered or whether the development during the period of organo- End points that are used to indicate effect on a specific organ is to be studied. genesis. Consequently, dosing is often reproductive dysfunction include (76): The animals are observed daily for started earlier, for example, on day 1 or day * Decreased libido, impotence clinical changes. Body weight and food 6 of the pregnancy). After birth, the num- * Abnormal sexual behavior consumption are recorded throughout ber of progeny is recorded and the litters * Sperm abnormalities: decreased num- the gestation. The uterus is removed by are adjusted so that each contains the same ber or motility, morphology cesarean section and the uterus and the number of pups. To determine whether * Subfecundity: abnormal sex organs fetuses are examined the day before antici- the chemical substance tested affects the and/or pubertal development, infertility pated birth. (The dam is examined macro- offspring directly through the mother's * Early fetal loss scopically for any structural abnormalities milk or indirectly by a change in milk pro- * Late fetal loss (stillbirth) or pathological changes that may have duction or as a result of a change in the * Intrapartum death influenced the pregnancy). If dosing is ini- behavior of the exposed mother, cross-fos- * Death in the first week tiated before or at the time of implantation, tering may be employed. Cross-fostering is * Decreased birth weight the preimplantation loss, i.e., the number a method by which litters from exposed * Prematurity/postmaturity at birth of embryos lost prior to implantation, mothers are reared by control mothers and * Altered sex ratio is evaluated. vice versa. * Multiple births, birth defects The total number of implantations, Studies to identify abnormal develop- * Infant death i.e., living embryos, dead embryos, and ment are conducted on individual animals * Childhood morbidity resorption (embryos that die early and are over short or long periods. In the rat, The Conventional Teratologgy Study. reassimilated, corresponding to early abor- studies last until weaning (although this The teratology study is the in vivo method tions in humans) are noted. The degree of period does not cover the entire period of for studying embryo-fetal toxicity as a con- resorption (i.e., the extent to which the brain development as the brain does not sequence of exposure during pregnancy embryo has been resorbed) is recorded to attain an approximate adult stage until the (e.g., growth retardation, anatomical varia- establish the time of death of the embryo age of 6 weeks, which corresponds to 12 to tions, teratogenicity, lethality). Various during the pregnancy. 15 years of age in humans). international organizations and countries The fetuses are sexed, weighed, and Methodology employed in behavioral have drafted guidelines for these tests. examined for gross malformations. Retarded teratology is described in several reviews. These guidelines have a number of com- growth and effects on visceral and skeletal Behavioral teratology tests may generally be mon requirements. The preferred species development are evaluated, including the grouped into tests of physical development, include rodents (e.g., rat, mouse) and non- degree of ossification ofthe bones. simple reflexes, motor function, develop- rodents (e.g., rabbit). Other species may be The Perinatal and Postnatal Studies. ment of the senses, spontaneous activity, used if relevant (e.g., differences in the tox- Prenatal exposure to chemicals may lead to learning and memory, and functions of the icokinetics between the preferred species a range of functional disturbances in the neurotransmitter systems. and man to clarify ambiguous results or to offspring. For example, lead and methyl- further study observed effects). mercury affect brain development but In Vivo Screening Tests Young mature virgin females are arti- effects on fertility, the immune system, As the capacity for toxicity testing cannot ficially inseminated or mated with males. metabolism of foreign substances, and keep up with the number of chemicals in The time of mating is established by obser- development as a whole have been observed. the modern society, there is an increasing vation of mating (e.g., rabbits), identi- Behavioral teratology identifies changes demand for new toxicological tests of fication of a plug (mixture of sperm and in behavior due to effects on the central shorter duration, using fewer resources. cellular material from the vagina of rats nervous system (CNS) and the peripheral Especially for older chemicals where the and mice), vaginal smear (in rats) or by nervous system (PNS). As behavior is patent rights no longer exist, or for chemi- noting the time of insemination (e.g., for affected by the function of other organs cals introduced into the market years ago pigs and rabbits). Normally, three dose such as liver, kidneys, and the endocrine when no or few toxicity data were levels and a control group (untreated or system, toxic effects on these organs in off- required, it is essential to develop tools for vehicle control; the group size is 20 preg- spring may also be reflected in general obtaining data for safety assessment. In nant animals for rats and mice, and 12 for changes in behavior. No single test is able recent years such new screening tests for

Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 799 TOPPARI ET AL.

reproductive and developmental toxicity the (expected) adverse effect on spermato- causing embryonic or fetal death is have been developed. By definition, a genesis and fertility (the use of lower dose evaluated. Dominant lethals are generally screening test is limited in scope compared and shorter premating treatment seem to be accepted to be the result of chromosomal to a conventional test. Data from a screen- possible causes for these negative findings). aberrations (structural and numerical ing test indicating a possible toxic potential On the basis of the experience with anomalies), but gene mutations and toxic of a substance identify the substance as one the tests, the OECD has prepared propos- effects cannot be excluded. ofhigh priority for further evaluation. als for two new guidelines (421 and 422), A short presentation of two examples of to be included in the Test Guideline In Vtro Tests screening tests for reproductive and devel- Programme. They are being considered for During recent years many in vitro test opmental toxicity is given below. inclusion as new base set tests among the systems have been proposed as alternatives The in Vivo Teratology Screening tests used for new chemicals in EU because to whole animal testing for developmental Test. This test was introduced as an alter- of a recognition of the need to obtain ini- toxicity. These tests are not able to replace native method to the above-mentioned tial information on reproduction and animal testing, but can reduce the number teratogenicity tests. The hypothesis under- developmental toxicity at an early phase of of chemicals to be tested with live animals. lying this test is that most prenatal effects hazard identification. In vitro techniques may also prove useful do not just produce specific defects but The purpose of the test is to generate in the screening of complex chemical mix- also are manifested in the postnatal period limited information concerning the effects tures, e.g., in product development (pre- as a lack ofviability and reduced growth. of a chemical on male and female repro- screen) or in the screening of closely Pregnant mice (rats can also be used) ductive performance such as gonadal related chemicals. In addition, in vitro are exposed to a test substance from day 8 function, mating behavior, conception, tests can be used as a tool for pinpointing to day 12 of the pregnancy. A control development ofconceptus, and parturition. the mechanisms underlying a known group is not exposed. One dose level is The test is not considered as an alternative embryo-fetotoxic effect, and as such pro- employed (the minimum toxic dose for the to, nor a replacement for, the existing vide information that can improve the mother animal). The mother animals are generation and teratology studies. interpretation of the results and conse- weighed during the period of exposure. The dosing of the animals is initiated 2 quently the extrapolation from laboratory After birth the litter is weighed on the first weeks prior to mating and continued until animal experiments to humans. No speci- and third days. Stillborn young and young the end of the study on postnatal day 4. fic test is mentioned here but a reference is that die after birth are dissected and exam- The number of animals per group (gener- given to a recent review. ined for defects. The test focuses on mal- ally at least three test groups and a control formations as the end points of concern. group) is at least 10 animals of each sex Test Strategies Results from a validation study have shown (expected to provide at least 8 pregnant Strategies for the assessment of reproductive that effects on offspring viability or body females per group). Effects on fertility and toxicity of chemical substances has been weight indicate a potential for teratogenic birth are registered. Live pups are counted proposed (71,76). The toxicity testing effects, i.e., malformations. A test guideline and sexed and litters weighed on days 1 strategies deal (at present) with evidence for this test has not been established. and 4 postpartum. The parameters include, for reproductive toxic effects, i.e., hazard Rep roductio n/Develop mental among others, a detailed histological exam- assessment, but potency should be consid- Toxicity Sereening Tests. Recently OECD ination on the ovaries, testes, and epi- ered at a later stage. An example is given in has introduced guidelines for screening didymides of at least the highest dosed and Table BI (77). tests for reproductive toxicity, i.e., the the control animals. Reproduction/Developmental Toxicity Significance of Experimental Screening Test as a part of the Screening Other Toxicity Tests Data and Their Relevance to Information Data Sets (SIDS) for older In Vivo Tests Man (Extrapolation) chemicals that were produced in large vol- It has been stated that the ultimate proof umes (68). The Combined Repeat Dose Other toxicity tests than those mentioned that a substance is a human teratogen can and Reproduction/Developmental Toxicity above can reveal effects that indicate a come only from information on the conse- Screening Test is a combination of the potential of a chemical to interfere with quences of human exposure. This state- "Repeated Dose Oral Toxicity-Rodent: normal reproduction. Thus, in all the toxi- ment is valid regarding toxicological effects 28-Day Study" and a reduced one-genera- cological tests involving repeated dosing, in general. However, the experience within tion study, whereas the Preliminary including the carcinogenicity test, the the area of developmental toxicology (and Reproduction Toxicity Screening Test is a gonads and accessory sex organs are carcinogenicity) comparing data from reduced one-generation study. subjected to pathological examination human exposure and data from dosing The reduced one-generation test has including histopathology. experimental animals (irrespective of the been validated with, among other chemicals, As an example, a rodent test for genetic difficulties comparing two different kinds ethylene glycol monoethyl ether (EGME) toxicology (OECD guideline for testing of data sets), and the biology as expressed and cyclophosphamide (CP). EGME showed chemicals), the Rodent Dominant Lethal by Calabrese (78) serve as the basis for the both systemic and reproductive/develop- Test should be mentioned. Male animals use of laboratory animals in predicting pos- mental effects similar to those previously are exposed to the test substance and sible consequences for humans. Calabrese reported using standard protocols. The mated to untreated virgin females. The var- says the following: study on CP demonstrated most of the ious germ-cell stages can be tested sepa- Cellular structure and biochemistry are known toxicological properties of CP, rately using sequential mating intervals. remarkably alike across the entire animal including developmental toxicity, but not The induction of dominant lethal effects kingdom, starting with the lipoprotein cell

800 Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

Table B1. Reproductive toxicity testing strategy. consequence of an exposure to chemicals at any period, if the test program and the Data Results Recommended testing specific tests are correctly designed. Human data Clear evidence of developmental No further testing However, several questions remain: toxicity or impaired fertility Is the testing program sufficient to Limited evidence of Teratology study and/or - developmental toxicity behavioral teratology study identify the possible consequences of - impaired fertility Fertility study exposure? Full toxicological data set Clear evidence of developmental No further testing * Are the effectiveness/sensitivity of the in experimental animals toxicity or impaired fertility specific toxicological methods sufficient No indication of reproductive toxicity No further testing (e.g., number of animals involved per - but neurotoxicity in adult Behavioral teratology study number of species tested per number of Some indication of reproductive toxicity See below groups per dosage regimen). Do we need Acute or repeated Testicular effects Fertility study (males dosed), new or additional toxicological tests? dose toxicity studies sperm-test * Are the parameters used relevant and Ovarian effects Fertility study (females dosed) sufficient? Will the specific test benefit Neurotoxicity in adult animals Fertility study combined with behavioral teratology study from the introduction of additional High acute toxicity Generation study or OECD parameter(s) or do we need additional ReproTox screening test parameters (end points)? Reproductive toxicity studies Clear evidence of developmental No further studies toxicity or impaired fertility Other Chemicals Subjected Developmental toxicity at Studies using specific end to Only a Limited Number of maternally toxic doses points and lower doses Toxicological Tests Equivocal evidence of impaired Repeat study fertility New chemicals are subjected to a testing Effects on sperm quality only Epidemiological study of program in the European Union involving exposed men different levels of testing according to the In vitro teratogen study Teratogen effect Teratology study tonnage on the market. Most chemicals Teratogen effect plus SAR and No further studies have only been tested in a limited range of kinetics indicate effect Data on kinetics/dynamics Accumulation in fetus, competitive Studies on developmental tests according to the requirements on the metabolism, etc. toxicity base set level. Only a few new chemicals Hormone analogue, accumulation will reach a tonnage level triggering the in testes Fertility study (relevant sex demand for a full scale testing program dosed) (level 2). No toxicological data No indications Initial toxicity testing The existing (older) chemicals are followed by studies for reproductive toxicity tested according to a strategy that considers the actual use of the chemical/actual expo- Abbreviations: SAR, structure-activity relationship; OECD, Organisation for Economic Cooperation and Development. sure and the potential of health hazard. Thus, a few of these chemicals have been subjected to some or more toxicological membrane, which affects the absorption of in rodent studies and their relevance for tests, while only a very limited set of data is xenobiotics into the cell to metabolic the man, the experience is limited. available for many of the chemicals. processes like glycolysis, the Krebs' cycle, However, it has been stated that effects on Therefore, the following problems should and numerous other aspects of intermedi- in ary metabolism. The similarity among fertility rodents seem to be a good indi- be discussed: animals on the cellular level is so apparent cator for effects in humans, and most work * What is the possibility of introducing that it serves as the basis upon which on contraceptive agents in humans stems one or more screening tests to obtain scientists have extrapolated or inferred from original studies in rodents. In particu- data sufficient for an acceptable functions from one species to another. lar, agents causing toxic damage to the assessment of the potential effect on testis in animals seem to have a similar reproduction? Thus, concerning findings in experimental effect in humans (79). * What are the possibilities of introduct- studies in laboratory animals, demonstrating additional parameters in the existing developmental toxicity (valid for cancer Present Situation ing tests at the base level to obtain as well) is indicative of a potential human Chemicals that Have Been additional information on the potential response. Subjected to a Full-scale of the specific chemical to impair the Concerning the predictive value of in Toxicological Teting Program reproduction? vitro results and their relevance for the man, Introduction of additional parameters the extrapolation between the response in The toxicology testing program includes or even new test methodology within isolated parts of the organism and that of testing of the chemical, covering exposure the area of reproductive toxicology will the intact organism has to be considered. in all periods of the lifespan, including require more experimental work to help Concerning the predictive value of the reproductive cycle. Thus, in princi- us better understand the underlying effects on fertility that were demonstrated ple, it should be possible to detect any biological mechanisms.

Environmental Health Perspectives * Vol 104, Supplement 4 * August 1996 801 TOPPARI ET AL.

REFERENCES 1. Thomas KB, Colborn T. Organochlorine endocrine disruptors 21. Jensen AA. Levels and trends of environmental chemicals in in human tissue. In: Chemically-induced Alterations in Sexual human milk. In: Chemical Contaminants in Human Milk and Functional Development: The Wildlife/Human (Jensen AA, Slorach S, eds). Boca Raton, FL:CRC Press, Connection (Colborn T, Ckement C, eds). Adv Mod Environ 45-198; (1991). Toxicol 21:365-394 (1992). 22. IPCS. Summary ofToxicological Evaluations Performed by the 2. LST. Pesticide Residues in Danish Foods, 1993 (in Danish). Joint FAO/WHO Meeting on Pesticide Residues [JMPR]. LST Publ No 228. S0borg, Denmark:Ministry of Health, WHO/PCS/ 94.1. Geneva:International Programme on National Food Agency of Denmark, 1994. Chemical Safety, 1993. 3. MST. Pesticide Statistics. Sales 1990, 1991, 1992: Treatment 23. WHO. Pesticide Residues in Food- 1991 Evaluations. Part II: Frequencies, 1992 (in Danish). Information Series from the Toxicology. WHO/PCS/92 52. International Programme on Danish Environmental Protection Agency, No 1. Chemical Safety. Geneva:World Health Organization, 1992. Copenhagen:Danish Environmental Protection Agency, 1994. 24. EHC 116. Tributyltin Compounds. Environmental Health 4. WHO. 1975 Evaluations of Some Pesticide Residues in Food. Criteria 116. International Programme on Chemical Safety. WHO Pesticide Residues, Series 5 (FAO/ AGP/1975/M/13). Geneva:World Health Organization, 1990. Geneva:World Health Organization, 1976. 25. FAO. Pesticide Residues in Food: 1986 Evaluations. Part II: 5. U.S. EPA. U.S. EPA's Integrated Risk Information System Toxicology. Food and Agriculture Organization Plant Product (IRIS Database) on CD-ROM (through October 1994). Protect Paper 78/2. Rome:Food and Agriculture Organization, CHEM-BANK, Silverplatter CD ROM system 3.11. 1987. Norwood, MA:Silverplatter Information, 1994. 26. FAO. Pesticide Residues in Food: 1987 Evaluations. Part II: 6. Hansen WH, Quaife ML, Habermann RT, Fitzhugh OG. Toxicology. Rome:Food and Agriculture Organization, 1988. Chronic toxicity of 2,4-dichlorophenoxyacetic acid in rats and 27. van Ravazenzwaay B. Dicsussion of Prenatal and Reproductive dogs. Toxicol Appl Pharmacol 20:122-129 (1971). Toxicity of Reg No 83-258 (Vinclozolin). Washington:U.S. 7. Bj0rklund N, Erne K. Toxicological studies of phenoxyacetic Environmental Protection Agency, 1992. herbicides in animals. Acta Vet Scand 7:364-390 (1966). 28. Gray LEJ, Ostby JS, Kelce WR. Developmental effects of an 8. EHC 29. 2,4-Dichlorophenoxyacetic Acid (2,4-D). environmental antiandrogen: the fungicide vinclozolin alters Environmental Health Criteria 29. International Programme sex differentiation of the male rat. Toxicol Appl Pharmacol on Chemical Safety. Geneva:World Health Organization, 129:46-52 (1994). 1984. 29. Kelce WR, Monosson E, Gamcsik MP, Laws SC, Gray LEJ. 9. EHC 84. 2,4-Dichlorophenoxyacetic Acid-Environmental Environmental hormone disruptors: evidence that vinclozolin Aspects. Environmental Health Criteria 84, International developmental toxicity is mediated by antiandrogenic metabo- Programme on Chemical Safety. Geneva:World Health lites. Toxicol Appl Pharmacol 126:276-285 (1994). Organization, 1989. 30. Wong C, Kelce WR, Sar M, Wilson EM. Androgen receptor 10. FAO. Pesticide Residues in Food: 1981 Evaluations. Food and antagonist versus agonist activities of the fungicide vinclozolin Agriculture Organization Plant Product Protect Paper 42. relative to hydroxyflutamide. J Biol Chem 270:19998-20003 Rome:Food and Agriculture Organization, 1982. (1995). 11. Neubert D, Dillman I. Embryotoxic effects in mice treated 31. EHC 153. Carbaryl. Environmental Health Criteria 153. with 2,4,5-trichlorophenoxyacetic acid and 2,3,7,8-tetra- International Programme on Chemical Safety. Geneva:World chlorodibenzo-p-dioxin. Naunyn Schmiedebergs Arch Health Organization, 1993. Pharmacol 272:243-264 (1972). 32. WHO. 1973 Evaluations of Some Pesticide Residues in Food. 12. EHC 158. Amitrole. Environmental Health Criteria 158. WHO Pesticide Residues Series 5. (FAO/AGP/1973/M/9/1). International Programme on Chemical Safety. Geneva:World Geneva:World Health Organization, 1974. Health Organization, 1994. 33. WHO. Evaluation of Some Pesticide Residues in Food. 13. WHO. Pesticide Residues in Food-1993 Evaluations. Part II: WHO/Food Add/67 32. (FAO/PL CP/ 1S5). Geneva:World Toxicology. WHO/PCS/94 4. International Programme on Health Organization, 1967. Chemical Safety. Geneva:World Health Organization, 1994. 34. FAO. Pesticide Residues in Food: 1989 Evaluations. Part II: 14. IARC. Occupational Exposures in Insecticide Application and Toxicology. Food and Agriculture Organization Plant Product Some Pesticides, Vol 53. In: IARC Monographs on the Protect Paper 100/2. Rome:Food and Agriculture Evaluation of Carcinogenic Risks in Humans. Organization, 1990. Lyon:International Agency for Research on Cancer, 1991. 35. WHO. Pesticide Residues in Food-1992 Evaluations. Part II: 15. EHC 148. Benomyl. Environmental Health Criteria 148. Toxicology. WHO/PCS/93 34. International Programme on International Programme on Chemical Safety. Geneva:World Chemical Safety. Geneva:World Health Organization, 1993. Health Organization, 1993. 36. Kupfer D, Bulger WH. Metabolic activation of pesticides with 16. FAO. Pesticide Residues in Food: 1983 Evaluations. Food and proestrogenic activity. Fed Proc 46:1864-1869 (1987). Agriculture Organization Plant Product Protect Paper 61. 37. Gray LEJ, Ostby J, Ferrell J, Rehnberg G, Linder R, Cooper R, Rome:Food and Agriculture Organization, 1985. Goldman J, Slott V, Laskey J. A dose-response analysis of 17. FAO. Pesticide Residues in Food: 1984 Evaluations. FAO methoxychlor-induced alterations of reproductive development Plant Product Protect Paper 67. Rome:Food and Agriculture and function in the rat. Fundam Appl Toxicol 12:92-108 Organization, 1985. (1989). 18. FAO. Pesticide Residues in Food: 1985 Evaluations. Part II: 38. Goldman JM, Cooper RL, Rehnberg GL, Hein JF, McElroy Toxicology. Food and Agriculture Organization Plant Product WK, Gray LEJ. Effects of low subchronic doses of methoxy- Protect Paper 72/2. Rome:Food and Agriculture Organization, chlor on the rat hypothalamic-pituitary reproductive axis. 1986. Toxicol Appl Pharmacol 86:474-483 (1986). 19. EHC 149. Carbendazim. Environmental Health Criteria 149. 39. Cummings AM, Gray LE. Methoxychlor affects the decidual International Programme on Chemical Safety. Geneva:World cell response of the uterus but not other progestational parame- Health Organization, 1993. ters in female rats. Toxicol Appl Pharmacol 90:330-336 20. FAO. Pesticide Residues in Food: 1977 Evaluations. Food and (1987). Agriculture Organization Plant Product Protect Paper 10 40. Bal HS. Effect of methoxychlor on reproductive systems of the Suppl. Rome:Food and Agriculture Organization, 1978. rat. Proc Soc Exp Biol Med 176:187-196 (1984).

802 Environmental Health Perspectives - Vol 104, Supplement 4 * August 1996 MALE REPRODUCTION AND ENVIRONMENTAL CHEMICALS

41. Cooke PS, Eroschenko VP. Inhibitory effects of technical grade 60. Walter BA, Agha BJ, Digenis GA. Disposition of methoxychlor on development of neonatal male mouse repro- [14C]nonoxynol-9 after intravenous or vaginal administration ductive organs. Biol Reprod 42:585-596 (1990). to female Sprague-Dawley rats. Toxicol Appl Pharmacol 42. Shain SA, Shaeffer JC, Boesel RW. The effect of chronic inges- 96:258-268 (1988). tion of selected pesticides upon rat ventral prostate homeosta- 61. Tryphonas L, Buttar HS. Genital tract toxicity of nonoxynol-9 sis. Toxicol Appl Pharmacol 40:115-130 (1977). in female rats: temporal development, reversibility and sequelae 43. EHC 124. Lindane. Environmental Health Criteria 124. of the induced lesions. Fundam Appl Toxicol 2:211-219 International Programme on Chemical Safety. Geneva:World (1982). Health Organization, 1991. 62. Tryphonas L, Buttar HS. Effects of the spermicide nonoxynol- 44. Sicar S, Lahiri P. Lindane (,y-HGH) causes reproductive failure 9 on the pregnant uterus and the conceptus of rat. Toxicology and fetoxicity in mice. Toxicology 59:171-177 (1989). 2:177-186 (1988). 45. EHC 123. a- and 3-Hexachlorocyclohexanes. Environmental 63. Meyer 0, Andersen PH, Hansen EV, Larsen JC. Health Criteria 123. International Programme on Chemical Teratogenicity and in vitro mutagenicity studies on nonoxynol- Safety. Geneva:World Health Organization, 1992. 9 and -30. Fundam App1 Toxicor62:236-238 (1988). 46. EHC 91. Aldrin and Dieldrin. Environmental Health Criteria 64. Woodward KN. Phthalate Esters: Toxicity and Metabolism. 91. International Programme on Chemical Safety. Geneva: Vols 1, 2. Boca Raton, FL:CRC Press. 1988. World Health Organization, 1989. 65. EHC 131. Diethylhexyl Phthalate. Environmental Health 47. Brotons JA, Olea-Serrano MF, Villalobos M, Pedraza V, Olea Criteria 131. International Programme on Chemical Safety. N. Xenoestrogens released from lacquer coatings in food cans. Geneva:World Health Organization, 1992. Environ Health Perspect 103:608-612 (1995). 66. Sharman M, Read WA, Castle L, Gilbert J. Levels of di-(2-eth- 48. EU Commission. Synoptic Document No 7. Draft of ylhexyl)phthalate and total phthalate esters in milk, cream, but- Provisional List of Monomers and Additives Used in the ter and cheese. Food Add Contam 11:375-385 (1994). Manufacture of Plastics and Coatings Intended to Come into 67. Petersen JH. Survey of di-(2-ethylhexyl)phthalate plasticizer Contact with Foodstuffs (updated 15 May 1994). Document contamination of retail Danish milks. Food Add Contam CS/PM/2356. Commission of the European Communities, 8:701-706 (1991). Directorate-General III, Industry, Industrial Affairs. III: 68. OECD. Guidelines for Testing of Chemicals. Consumer Goods Industries, Foodstuffs-Legislation and Paris:Organisation for Economic Co-operation and Scientific and Technical Aspects. Brussels:European Union Development, 1993. Commission, 1994. 69. Commission of the European Communities. Reproductive 49. EHC 140. Polychlorinated Biphenyls and Terphenyls. 2nd ed. Toxicity. The Toxicology of Chemicals, Series 2, Vol 1. Environmental Health Criteria 140. International Programme Luxembourg:Commission of the European Communities, on Chemical Safety. Geneva:World Health Organization, 1993. 1993. 50. Ahlborg UG, Hanberg A, Kenne K. Risk Assessment of 70. LST. Embryo-Foetal Damage and Chemical Substances. Polychlorinated Biphenyls (PCBs). Environmental Report in Working Party Report. LST publication No 181. S0borg, the Nord Series. Nord 26. Copenhagen: Nordic Council of Denmark:National Food Agency of Denmark, 1989. Ministers, 1992. 71. Hass U, Brandorff NP, Brunborg G, Ekstrbm T, Hansen EV, 51. Ahlborg UG, Becking GC, Birnbaum LS, Brouwer A, Derks Jakobsen BM, Jelnes JE, Meyer 0, Taskinen H, Wiger R. HJGM, Feeley M, Golor G, Hanberg A, Larsen JC, Liem Occupational Reproductive Toxicity. Methods and Testing AKD, Safe SH, Schlatter C, Wrn F, Younes M, Yrjnheikki E. Strategies for Hazard Assessment of Workplace Chemicals. Toxic equivalency factors for dioxin-like PCBs. Chemosphere Copenhagen:Nordic Council of Ministers and National 28:1049-1067 (1994). Institute of Occupational Health, 1994. 52. Safe SH. Polychlorinated biphenyls (PCBs): environmental 72. Hansen E, Meyer 0. Animal models in reproductive toxicol- impact, biochemical and toxic responses, and implication for ogy. In: Handbook of Laboratory Animals Science. Vol II: risk assessment. Crit Rev Toxicol 24:87-149 (1994). Animal Models (Svendsen P, Hau J, eds). Boca Raton, 53. Van den Berg M, De Jongh J, Poiger H, Olsen JR. The toxico- FL:CRC Press, 1994. kinetics and metabolism of polychlorinated dibenzo-p-dioxins 73. ECETOC. Identification and Assessment of the Effects of (PCDDs) and dibenzofurans (PCDFs) and their relevance for Chemicals on Reproduction and Development. Monograph toxicity. Crit Rev Toxicol 24:1-74 (1994). No 5. Brussels:European Chemical Industry, Ecology & 54. Ahlborg UG, Hakansson H, Wxrn F, Hanberg A. Nordisk Toxicology Centre, 1983. dioxinriskbed6mning. [Nordic Risk Assessment of Dioxins.] 74. U.S. EPA. Health Effect Tests Guidelines. OPPTS 870.3800. Environmental Report in the Nord Series. Nord 49. Reproduction and Fertility Effects. Draft. EPA 712-C-94-208. Copenhagen:Nordic Council of Ministers, 1988. Washington:U.S. Environmental Protection Agency, 1994. 55. Frst P, Beck H, Theelen R. Assessment of human intake of 75. Thomas JA. Toxic responses of the reproductive system. In: PCDDs and PCDFs from different environmental sources. Casarett and Doull's Toxicology: The Basic Science of Poison Toxic Subst J 12:133-150 (1992). (Amdur MO, Doull J, Klaassen CD, eds). New York: 56. LST. Dioxiner i danske Levnedsmidler. Publikation nr. 170. Pergamon Press, 1991;484-520. S0brg, Denmark:Danish Environmental Protection Agency, 76. Gray TJB. Application of in vitro systems in male reproductive 1988. toxicology. In: Physiology and Toxicology of Male 57. Pohjanvirta R, Tuomisto J. Short-term toxicity of 2,3,7,8-tetra- Reproduction (Foster PMD, Lamb JC IV, eds). New chlorodibenzo-p-dioxin in laboratory animals: effects, mecha- York:Academic Press, 1988; 137-153. nisms, and animal models. Pharmacol Rev 46:483-549 (1994). 77. Commission of the European Communities. Risk Assessment 58. EHC 71. Pentachlorophenol. Environmental Health Criteria of Notified New Substances. Technical Guidance Document 71. International Programme on Chemical Safety. Geneva: (in Support of the Risk Assessment Directive. 93/67/EEC), World Health Organization, 1987. 1993. 59. Chvapil M, Eskelson CD, Stiffel V, Owen JA, Droegemueller 78. Calabrese EJ. Principles of Animal Extrapolation. New W. Studies on nonoxynol-9. II: Intravaginal absorption, distri- York:John Wiley & Sons, 1983. bution, metabolism and excretion in rats and rabbits. 79. Barlow SM, Sullivan FM. Reproductive Hazards of Industrial Contraception 22:325-339 (1980). Chemicals. New York:Academic Press, 1982.

Environmental Health Perspectives * Vol 104, Supplement 4 - August 1996 803