Strengths and limitations of using repeat-dose toxicity studies to predict effects on fertility

Regulatory Toxicology and Pharmacology 48 (2007) 241–258 www.elsevier.com/locate/yrtph

Strengths and limitations of using repeat-dose toxicity studies to predict effects on fertility M.P. Dent

*

Unilever Safety and Environmental Assurance Centre, Unilever Colworth, Bedfordshire MK44 1LQ, UK Received 8 February 2007 Available online 12 April 2007

Abstract The upcoming European chemicals legislation REACH (Registration, Evaluation, and Authorisation of Chemicals) will require the risk assessment of many thousands of chemicals. It is therefore necessary to develop intelligent testing strategies to ensure that chemicals of concern are identified whilst minimising the testing of chemicals using animals. Xenobiotics may perturb the reproductive cycle, and for this reason several reproductive studies are recommended under REACH. One of the endpoints assessed in this battery of tests is mating performance and fertility. Animal tests that address this endpoint use a relatively large number of animals and are also costly in terms of resource, time, and money. If it can be shown that data from non-reproductive studies such as in-vitro or repeat-dose toxicity tests are capable of generating reliable alerts for effects on fertility then some animal testing may be avoided. Available rat sub-chronic and fertility data for 44 chemicals that have been classified by the European Union as toxic to fertility were therefore analysed for concordance of effects. Because it was considered appropriate to read across data for some chemicals these data sets were considered relevant for 73 of the 102 chemicals currently classified as toxic to reproduction (fertility) under this system. For all but 5 of these chemicals it was considered that a well-performed sub-chronic toxicity study would have detected pathology in the male, and in some cases, the female reproductive tract. Three showed evidence of direct interaction with oestrogen or androgen receptors (linuron, nonylphenol, and fenarimol). The remaining chemicals (quinomethionate and azafenidin) act by modes of action that do not require direct interaction with steroid receptors. However, both these materials caused in-utero deaths in pre-natal developmental toxicity studies, and the relatively low NOAELs and the nature of the hazard identified in the sub-chronic tests provides an alert for possible effects on fertility (or early embryonic development), the biological significance of which can be ascertained in a littering (e.g. 2-generation) study. From the chemicals reviewed it would appear that where there are no alerts from a repeat-dose toxicity study, a pre-natal developmental toxicity study and sex steroid receptor binding assays, there exists a low priority for animal studies to address the fertility endpoint. The ability for these types of tests to provide alerts for effects on fertility is clearly dependent on the mode of action of the toxicant in question. Further work should therefore be performed to determine the ‘failure rate’ of this type of approach when applied to a larger group of chemicals with diverse modes of action.  2007 Elsevier Inc. All rights reserved. Keywords: REACH; Fertility; Integrated testing strategy; Hazard identification; Repeat-dose; Two-generation; Screening tests; Mode of action

1. Introduction The forthcoming European chemicals legislation REACH (Registration, Evaluation, and Authorisation of Chemicals) will necessitate the re-evaluation of any chemical that is produced in or imported into the European *

Fax: +44 (0) 1234 274 744. E-mail address: [email protected]

0273-2300/$ - see front matter  2007 Elsevier Inc. All rights reserved. doi:10.1016/j.yrtph.2007.04.001

Union at levels of 1 tonne per annum (tpa) or more (EU, 2006). The safety studies required for this process will depend largely on the volume of the chemical produced or imported, with chemicals produced in large quantities requiring the most extensive data packages. The reproductive and developmental studies that are recommended under REACH are the OECD 421 screening test, the OECD 414 pre-natal development study, and the OECD 416 2-generation reproduction study. However, other

242

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

pre-existing data will also need to be taken into account to prevent the unnecessary use of animals. For example, developmental toxicity studies performed to historical guidelines may be available, but many of these older tests involved the treatment of pregnant females for a shorter period during gestation than the current OECD 414 guideline. Furthermore, other good quality modern studies may be available which are not performed to an OECD guideline, such as the Continuous Breeding Protocol used by the US National Toxicology Programme. While these types of test do not match the current OECD requirements, they can provide valuable data to assess the potential for a chemical to affect the reproductive system or development. The finalised REACH legislation requires reproductive toxicity testing to be performed once production/import reaches the 10 tpa threshold. At this level, if there is no evidence from available information on structurally related substances, structural activity relationship, or quantitative structural activity relationship ((Q)SAR) estimates, or invitro indicators that the substance may be a developmental toxicant, then an initial OECD 421 screening test is conducted. Alternatively an OECD 422 screening test (which combines the reproductive screen of the OECD 421 study with some extra assessments of general toxicity) could be used. The screening test does not need to be performed if either a prenatal developmental toxicity study (OECD 414) or a two-generation study (OECD 416) is available. Further, at this tonnage level one of the definitive tests should be performed instead of the screening test if there are indications of reproductive target organ toxicity from repeated-dose toxicity tests or a close structural relationship to a known developmental/reproductive toxicant. The next tonnage trigger (100 tpa) requires a developmental toxicity study (OECD 414) unless already performed, while the two-generation study is required once production or import reaches the 1000 tpa level (unless alerts triggered this study at a lower tonnage level). The registrant therefore has a fundamental decision to make at 10 tpa; to perform the screening test at this tier then perform a prenatal developmental toxicity study if produc-

tion/import exceeds 100 tpa in the future, or if it suspected that this will be the case go straight to the prenatal developmental toxicity study and bypass the screening test altogether. While it could be argued that both these studies are necessary because the OECD 421 screening test assesses endpoints not addressed in the OECD 414 developmental toxicity, the value of the screening test is diminished when data from a definitive developmental toxicity study are available. The endpoints not addressed in the pre-natal development study (OECD 414) that are covered by the screening test (OECD 421) are fertility, mating performance, parturition, early post-natal development and maternal care and lactation. The repeat-dose toxicity studies that are required at 100 tpa (28-day or 90-day repeatdose toxicity tests OECD 407 or 408) can provide useful information that may point to the potential for effects on reproduction. This is because repeat-dose toxicity studies include organ weights and histological examination of the gonads and accessory sex organs, providing useful information on potential effects on fertility or endocrine effects. A brief summary of the types of effects that may point to a potential for effects in another reproductive endpoint is shown in Table 1. The decision whether to perform a screening test when it is suspected that a pre-natal development study will eventually be needed therefore depends on the certainty with which effects on fertility and post-natal survival and development can be predicted from a combination of a pre-natal development study and the required repeat dose repeatdose toxicity study. For most chemicals this will be a repeat dose study of 90-days’ duration or less. The aim of this investigation was to identify chemicals where significant amounts of published reproductive and repeat-dose toxicity data are available. The objective was to look for concordance between studies with a view to: 1. Identifying chemicals where data from pre-natal developmental toxicity and sub-chronic repeat-dose toxicity studies have not been predictive of effects on mating performance and fertility.

Table 1 Prediction of effects across studies Study

Examples of findings that may give alerts for. . . Fertility/mating performance

Pre-natal development

Postnatal survival/development

Pre-natal development study (OECD 414)

Early implantation loss; effects on reproductive organs (e.g. hypospadias, cryptorchidism)

Structural changes fully assessed in study

Certain abnormalities may indicate a possible effect on postnatal survival; in-utero mortality; significant growth retardation in the absence of maternal toxicity

Repeat-dose toxicity study (OECD 407 or 408)

Reproductive organ pathology; organ weight changes in reproductive or accessory organs

Neurotoxicity; pathology of the endocrine system

Pathology of the endocrine system

Developmental and reproductive toxicity screening test (OECD 421/422)a

Assessed in study

Smaller live litter size at birth; observance of abnormal offspring

Assessed in study (but only up to day 4 of age)

a The relative insensitivity of this investigation (low animal numbers, loss of information due to cannibalism, and short duration) mean that effects may need to be substantial in order to be detected.

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

2. Evaluating whether in such cases common modes of action can be identified and how this knowledge may be used in intelligently designing testing programmes to identify potential hazards to development and fertility.

1.1. Identification of candidate chemicals and information gathering It was decided to select candidate chemicals from those that have been classified and labelled in Europe according to the Classification, Packaging, and Labelling of Dangerous Substances Directive (67/548/EEC). These are materials that are generally well studied and for which data are likely to be available in the public domain. Depending on the endpoint(s) affected, the severity of the effect and the credibility of the supporting data, materials that classified and labelled according to this directive as toxic to reproduction are given one of the following reproductive category (Rep Cat) codes: 1. Substances known to impair fertility or cause developmental toxicity in humans (i.e. there is sufficient evidence to establish a causal relationship to the material and effects in humans). 2. Substances which should be regarded as though they impair fertility or cause developmental toxicity in humans (i.e. there is sufficient evidence to provide a strong presumption of effects, e.g. relevant animal studies or other information). 3. Substances which cause concern for human fertility or cause concern for humans owing to possible developmental effects (i.e. sufficient information to provide suspicion of effects but insufficient to place in Category 2). Substances that impair fertility in humans or should be regarded as such (Rep Cat 1 or 2) are given the risk phrase R60: May impair fertility. Substances that cause developmental toxicity in humans or should be regarded as such (Rep Cat 1 or 2) are given the risk phrase R61: May cause harm to the unborn child. Substances which cause concern for human fertility (Rep Cat 3) are given the risk phrase R62: Possible risk of impaired fertility. Substances which cause concern for humans owing to possible developmental effects (Rep Cat 3) are given the risk phrase R63: Possible risk of harm to the unborn child. The most recent Amendment to Technical Progress (ATP) at the time of writing was ATP 29 (published in OJ L 152 of 30/04/2004). A list of materials that have been classified as toxic to reproduction up to and including the 29th ATP was therefore compiled. A search was then performed for relevant data published on materials classified as Rep Cat 1, 2, or 3 for fertility (R60 or R62). Databases interrogated included the US National Toxicology Program (NTP) database (www.ntp.niehs.nih.gov) and the US National Library of Medicine’s Toxnet database

243

(www.toxnet.nlm.nih.gov). The chemical name and CAS number was used to search for reproductive toxicity studies and repeat-dose toxicity studies performed by the NTP. The Toxnet database was searched using the chemical name, CAS number, and keywords such as ‘reproduction’, ‘fertility’, ‘development’, ‘two-generation’, ‘sub-chronic toxicity’, ‘chronic toxicity’, and ‘histology’. Candidate materials were then selected according to the following criteria: Materials that were classified as Rep Cat 1, 2, or 3 for fertility (R60 or R62) for which an animal study including histology of the sex organs of treated animals was available. The data for these materials was used to investigate whether the histological findings alone could be predictive of functional effects upon fertility seen in rodent reproductive studies. If no effects indicative of reproductive toxicity were found, the next stage was to ensure that a study of functional effects on fertility was available. This information was necessary to ensure that the species in question is susceptible to the reproductive effects of the test material. Without this it was not possible to conclude that a reproductive toxicology study may help identify a human hazard or not. Therefore, where possible, for each chemical a judgement was made as to whether rodent functional effects could be predicted in a repeat-dose toxicity study using appropriate doses. Where possible, this judgement is based on well conducted and reported sub-chronic rodent toxicity studies published in the literature. It was decided to restrict the exercise to sub-chronic data since chronic toxicity data will be available for very few chemicals under REACH. Where these are not available, histology from the F0 (parental) animals on a reproduction study may have been used. For many substances there are no such publications. In these instances, public data from reliable sources such as monographs from the United States Environmental Protection Agency (EPA), the World Health Organisation (WHO) or the European Chemicals Bureau (ECB) was used. For some well known toxicants information was taken from toxicology text books. For some materials no information could be found. Where possible, materials that were expected to exhibit similar intrinsic hazards based upon structural similarities were considered together. Thus, dinitrotoluenes were grouped together and assessed using data for 2,4-dinitrotoluene. R-2,3-Epoxy-1-propanol was read-across from 2,3-epoxypropan-1-ol (glycidol). Lead compounds, chromium compounds, and cadmium compounds were grouped together under their respective metals. 2-methoxyethyl acetate and 2-ethoxyethyl acetate were subject to read-across from 2-methoxyethanol or 2-ethoxyethanol, respectively, as the toxicity to reproduction of these materials is expected to be similar (WHO, 1990). Hexan-2-one was evaluated on the basis that its metabolite, 2,5-hexanedione is a known testicular toxicant (Abdel-Rahman et al., 1976). With the exception of these materials, where information could be found on a substance it was not included in Table 2.

244

Table 2 The ability for subchronic tests to predict effects on fertility CAS#

Classification

Could rodent functional effects If yes, nature of effect R60 R62 R61 R63 be predicted in a sub-chronic study? (reference)

Carbon disulphide

75-15-0

—

3

—

3

Yes (Gondzik, 1970)

Linuron (ISO); 3-(3,4-dichlorophenyl)-1-methoxy-1methylurea Lead compounds Lead hexafluorosilicate Lead alkyls Lead diazide; lead azide Lead chromate Lead di(acetate) Trilead bis(orthophosphate) Lead acetate, basic; lead acetate Lead (II) methanesulphonate Lead sulfochromate yellow; C.I. Pigment Yellow 34 [this substance is identified in the Colour Index by Colour Index Constitution Number, C.I. 77603] Lead chromate molybdate sulfate red; C.I. Pigment Red 104 [this substance is identified in the Colour Index by Colour Index Constitution Number, C.I. 77605] Lead hydrogen arsenate Lead 2,4,6-trinitro-m-phenylene dioxide; lead 2,4,6trinitroresorcinoxide; lead styphnate Octamethylcyclosiloxane (D4)

330-55-2

—

3

2

—

No (EPA, 1995)

Testes: disorganised seminiferous tubules, reduction in number of spermatozoa; more advanced regressive changes at highest dose and vacuolation of Sertoli and Leydig cells —

25808-74-6 — 13424-46-9 7758-97-6 301-04-2 7446-27-7 1335-32-6 17570-76-2 1344-37-2

— — — — — — — — —

3 3 3 3 3 3 3 3 3

1 1 1 1 1 1 1 1 1

— — — — — — — — —

Yes (Saxena et al., 1984) See above See above See above See above See above See above See above See above See above

Testes: tubular degeneration, absence of spermatogenesis See above See above See above See above See above See above See above See above See above

12656-85-8

—

3

1

—

See above

See above

7784-40-9 15245-44-0

— —

3 3

1 1

— —

See above See above

See above See above

556-67-2

—

3

—

—

Yes (Burns-Naas et al., 2002)

Ovaries: hypoactive; vagina: mucification

Yes (Elbetieha and AlHamood, 1997)

Seminal vesicles: reduced weight; preputial gland: reduced weight; ovaries: increased weight (no histology). Also increased oestrous cycle length (Kanojia et al., 1996) See above See above See above See above See above See above

Chromium (VI) compounds

Chromium (VI) trioxide Potassium dichromate Ammonium dichromate Sodium dichromate anhydrate Sodium dichromate, dihydrate Sodium chromate Cadmium compounds Cadmium (non-pyrophoric); [1] Cadmium oxide (non-pyrophoric) [2] Cadmium fluoride Cadmium chloride Cadmium sulphate Cadmium sulphide Cadmium (pyrophoric) Benzo[a]pyrene

1333-82-0 7778-50-9 7789-09-5 10588-01-9 7789-12-0 7775-11-3

— 2 2 2 2 2

3 — — — — —

— 2 2 2 2 2

— — — — — —

See See See See See See

above above above above above above

7440-43-9 [1]

—

3

—

3

Yes (Timbrell, 1991) See above

Testes: necrosis, degeneration, complete loss of spermatozoa See above

1306-19-0 [2] 7790-79-6 10108-64-2 10124-36-4 1306-23-6 7440-43-9 50-32-8

2 2 2 — — 2

— — — 3 3 —

2 2 2 — — 2

— — — 3 3 —

See above See above See above See above See above Yes (Archibong et al., 2003)

See above See above See above See above See above Testis: reduced weight; sperm analysis: reduced sperm motility and density (no testicular histology quoted). Also raised LH levels

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

Material name

Benzo[def]chrysene n-Hexane Nonylphenol; [1]

110-54-3 25154-52-3 [1]

— —

3 3

— —

— 3

Yes (Hansen, 1991) No (Cunny et al., 1997)

Testes: atrophy (Rat ‘functional’ effects restricted to evidence of oestrogenicity in F1 and F2 animals, Chapin et al., 1999)

1-Bromopane; n-propyl bromide

84852-15-3 [2] 106-94-5

2

—

—

3

Yes (NTP, 2003)

1,2-Dibromo-3-chloropropane

96-12-8

1

—

—

—

Yes (Meistrich et al., 2003)

1,2,3-Trichloropropane

96-18-4

2

—

—

—

Yes (NTP, 1990a; Johannsen et al., 1988)

Dodecachloropentacyclo[5.2.1.02,6.03,9.05,8]decane; mirex 2-Brompropane

2385-85-5

—

3

—

3

Yes (Yarbrough et al., 1981)

Prostate: reduced weight; seminal vesicles: reduced weight; sperm analysis: reduced sperm quality (no details); testes: inhibition of spermiation; ovaries: follicular cysts. Also increased oestrous cycle length Testes: loss of differentiating germ cells; morphological changes to Sertoli cells; tubular atrophy Ovary: reduced weight; epididymides: reduced weight; testes: reduced weight. However, a number of other studies do not describe effects on these organs (NTP, 1993) Testes: tubular degeneration, reduction of spermatogonia

75-26-3

1

—

—

—

Yes (Li et al., 2001; Omura et al., 1997; Lim et al., 1997)

a,a,a,4-Tetrachlorotoluene; p-chlorobenzotrichloride 2-Methoxyethanol; ethylene glycol monomethyl ether

5216-25-1 109-86-4

— 2

3 —

— 2

— —

2-Ethoxyethanol; ethylene glycol monoethyl ether 1,2-Dimethoxyethane; ethylene glycol dimethyl ether; EGDME Allyl glycidyl ether; allyl 2,3-epoxypropyl ether; prop-2-en-1-yl 2,3-epoxypropyl ether 2,3-Epoxypropan-1-ol; glycidol; oxiranemethanol R-2,3-Epoxy-1-propanol

110-80-5 110-71-4

2 2

— —

2 2

— —

Yes (ECB, 2000) Yes (Foster et al., 1984; Lee et al., 1989) Yes (Foster et al., 1984) Yes (Nagano et al., 1984)

106-92-3

—

3

—

—

Yes (Kodama et al., 1961)

Testes: necrosis

556-52-5 57044-25-4

2 2

— —

— —

— —

Testes: atrophy See above

Fenarimol (ISO); 2,4 0 -dichloro-a-(pyrimidin5-yl)benzhydryl alcohol 1,2-Bis(2-methoxyethoxy)ethane; TEGDME; triethylene glycol dimethyl ether; triglyme Bis(2-methoxyethyl) ether Hexan-2-one; methyl butyl ketone; butyl methyl ketone; methyl-n-butyl ketone

60168-88-9

—

3

—

3

Yes (NTP, 1990b) Yes (read-across from Glycidol) (NTP, 1990b) No (Hirsh et al., 1986)

—

112-49-2

—

3

2

—

Yes (HSE, 2003)

Testes: toxicity

111-96-6 591-78-6

2 —

— 3

2 —

— —

Testes: atrophy Testes: atrophy; sperm analysis: reduced number, motility, and number morphologically normal

Quinomethionate; chinomethionat (ISO); 6-methyl-1, 3-dithiolo(4,5-b)quinoxalin-2-one 2-Methoxyethyl acetate; methylglycol acetate

2439-01-2

—

3

—

—

110-49-6

2

—

2

—

2-Ethoxyethyl acetate; ethylglycol acetate

111-15-9

2

—

2

—

Vinclozolin (ISO); N-3,5-dichlorophenyl-5-methyl5-vinyl-1,3-oxazolidine-2,4-dione

50471-44-8

2

—

2

—

Yes (Lee et al., 1989) Yes (Read-across from 2,5hexanedione, Isobe et al., 1998; Abdel-Rahman et al., 1976) No (FAO/WHO, 1987; HSE, 2003) Yes (read-across from 2-methoxyethanol) (WHO Working Group, 1990) Yes (read-across from 2-ethoxyethanol) (WHO Working Group, 1990) Yes (Wong et al., 1995)

4-Nonylphenol, branched [2]

Testes: degeneration of pachytene spermatocytes Testes: tubular atrophy

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

Testes: reduced weight, tubular damage, reduction of spermatogonia; ovary: reduced weight. Also lengthened oestrous cycles Testes: atrophy Testes: degeneration of pachytene spermatocytes, atrophy

— Testes: degenerative changes in germinal epithelium of seminiferous tubules. Sperm analysis: reduced number, motility, and number morphologically normal. Similar effects to above

Testes: Leydig cell hyperplasia; prostate: atrophy (continued on next page) 245

246

Table 2 (continued) Classification

Could rodent functional effects be predicted in a sub-chronic R60 R62 R61 R63 study? (reference)

If yes, nature of effect

Methoxyacetic acid

625-45-6

2

—

2

—

Yes (NTP, 1986)

Bis(2-ethylhexyl) phthalate; di(2-ethylhexyl) phthalate; DEHP Dibutyl phthalate; DBP

117-81-7

2

—

2

—

Yes (NTP, 1982)

Testes: atrophy; epididymides, and seminal vesicles: reduced weight; sperm analysis: reduced number, motility, and number morphologically normal Testes: degeneration of seminiferous tubules

84-74-2

—

3

2

—

Yes (NTP, 1995)

84777-06-0 [1]

2

—

2

—

Yes (NTP, 1985)

Testis and epidymides: reduced weight and degeneration of germinal epithelium; epididymides: reduced weight Testes, epididymides, and seminal vesicles: reduced weight

- [2] 131-18-0 [3] 605-50-5 [4] 85-68-7 68515-42-4

— —

3 3

2 2

— —

Yes (Piersma et al., 1995) Yes (ECB, 2000a)

Testes: reduced weight, degeneration, Leydig cell hyperplasia Testes: atrophy

98-95-3 121-14-2 [1]

— —

3 3

— —

— —

Yes (Mitsumori et al., 1994) Yes (BIBRA Working Group, 1994)

Testes: atrophy Testes: atrophy; ovaries: atrophy

25321-14-6 [2] 606-20-2 602-01-7 610-39-9 618-85-9 619-15-8 88-85-7

— — — — — —

3 3 3 3 3 3

— — — — — 2

— — — — — —

See above See above See above See above See above Yes (Takahashi et al., 2003)

2-Nitrotoluene

88-72-2

—

3

—

—

Yes (Dunnick, 1992; HSE, 2003)

Azafenidin 1,3-Diphenylguanidine

68049-83-2 102-06-7

— —

3 3

2 —

— —

No (EPA, 1999) Yes (Bempong and Hall, 1983; Koeter et al., 1992)

4,4 0 -Oxydianiline and its salts; p-aminophenyl ether Carbendazim (ISO); methyl benzimidazol-2-ylcarbamate Benomyl (ISO); methyl 1-(butylcarbamoyl)benzimidazol2-ylcarbamate Molinate (ISO); S-ethyl 1perhydroazepinecarbothioate; S-ethyl perhydroazepine-1-carbothioate Epoxiconazole; (2RS,3SR)-3-(2chlorophenyl)-2-(4-fluorophenyl)-[(1H1,2,4-triazol-1- yl)methyl]oxirane Acrylamide; prop-2-enamide 2-Chloracetamide

101-80-4

—

3

—

—

10605-21-7

2

—

2

—

17804-35-2

2

—

2

—

Insufficient data to determine a selective rodent functional effect Yes (WHO Working Group, 1993a) Yes (WHO Working Group, 1993b)

See above See above See above See above See above Prostate: reduced weight; seminal vesicles: reduced weight; sperm analysis: reduced motility and number morphologically normal Testes: degeneration; sperm analysis: reduced number and motility; altered oestrous cylce — Testes: reduced weight, irregularly shaped seminiferous tubules, loss of Leydig cells, loss of spermatids, and spermatozoa in tubules; sperm analysis: reduced sperm count and number of morphologically normal sperm Although mutacat 2

2212-67-1

—

3

—

—

Yes (Ellis et al., 1998)

Testes: tubular atropy; sperm analysis: reduced number of morphologically normal sperm

133855-98-8

—

3

—

3

—

79-06-1 79-07-2

— —

3 3

— —

— —

Likely based on mode of action (inhibition of steroidogenesis, Zarn et al., 2003) Yes (EPA, 1994) Yes (Anon, 1991)

1,2-Benzenedicarboxylic acid, dipentylester, branched, and linear; [1] n-Pentyl-isopentylphthalate; [2] Di-n-pentyl phthalate; [3] Diisopentylphthalate [4] BBP; benzyl butyl phtalate 1,2-Benzenedicarboxylic acid; di-C7-11branched and linear alkylesters Nitrobenzene 2,4-Dinitrotoluene; dinitrotoluene, technical grade; [1] Dinitrotoluene [2] 2,6-Dinitrotoluene 2,3-Dinitrotoluene 3,4-Dinitrotoluene 3,5-Dinitrotoluene 2,5-Dinitrotoluene Dinoseb; 6-sec-butyl-2,4-dinitrophenol

Testes: diffuse atrophy; prostate: prostatitis; sperm analysis: reduced count Testes: reduced weight; epididymides: reduced weight, reduced epididymal sperm count

Testes: atrophy Testes: atrophy

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

CAS#

Material name

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

2. Results One-hundred-and-two chemicals have been classified as R60 or R62 up to and including the 29th ATP. Data sets that were considered relevant to 73 of these 102 chemicals were analysed. As some chemicals were grouped together, a total of 44 data sets were analysed. The chemicals studied and references used are listed in Table 2. Pathologies that gave alerts for potential effects on fertility occurred in both the male and female sex and accessory organs, but most chemicals caused testicular toxicity. It was considered that a well-performed sub-chronic toxicity study would demonstrate target organ toxicity to the reproductive system for 68 of these 73 chemicals. The observations that would contribute to this prediction include reduced weight of the testes or epididymides, effects upon sperm morphology/motility/number, disorganisation or degeneration in the seminiferous tubules, changes to or loss of Leydig or Sertoli cells. Also, Leydig cell hyperplasia could indicate a possible antiandrogen mode of action, as could prostatic atrophy. Effects in the female reproductive tract included changes in ovary or uterus weight, ovarian atrophy, or changes in the extent of vaginal mucification. Only 4 of the chemicals studied resulted in pathology of the female reproductive tract. Some chemicals caused changes to the regularity and length of the oestrous cycle, but as vaginal smears are not included in a standard OECD 407 or 408 study, these functional changes would not usually be detected in this type of test. As most chemicals caused pathology of the male reproductive tract, it would be easy to assume that most chemical effects upon fertility are male-mediated. Whilst there may be some truth in this, it should also be remembered that insults which lead to a morphological change in the male reproductive tract are also likely to have a deleterious effect on the female organism, especially those which are hormonally mediated. It could be argued that the reason pathological effects are more commonly seen in the male is that histology of the testes is more likely to detect effects upon fertility than histology of the female reproductive organs, since degeneration of epithelial cells in the uterus and other hormonally sensitive organs are part of the normal histology in hormonal cycling rodents. The inclusion of sperm analyses (counts, motility, and morphology) may increase the sensitivity of the repeat dose study in detecting effects in the male. The testicular effects that were predictive of functional effects could be the result of a number of toxic modes of action, some of which may also give rise to functional effects in females, including: • • • •

Direct damage to DNA. Direct damage to Sertoli or Leydig cells. Secondary to a physiological change. Hormonal modulation.

247

Examples of materials thought to act in these ways and the effects they have on the male (and where applicable the female) reproductive system are reviewed below. 2.1. Direct damage to DNA As the adult testis contains many rapidly dividing cells, and successful mating relies on the integrity of this process, materials that directly interfere with DNA replication clearly have the potential to adversely affect fertility. Amongst the most damaging chemicals in this regard are the alkylating agents. These materials act by transferring an alkyl group to critical cellular constituents, and particularly active sites are the N7 and O6 of guanine and the N3 and N7 of adenine (Timbrell, 1991). Alkylating agents can therefore lead to miscoding, cross linking, and DNA breakage, and in germ cells these effects have great implications for fertility. Not only may DNA damage lead to the formation of compromised sperm, DNA-damaged germ cells may also undergo apoptosis as the cell attempts to eliminate the aberrant DNA, leading to reduced sperm output. For example, the alkylating agent cyclophosphamide induces apoptosis in male germ cells, resulting in oligozoospermia or azoospermia (Cai et al., 1997). In addition to testicular damage, this chemotherapeutic agent has also been shown to cause ovarian atrophy, resulting in a marked loss of primordial follicles (Meirow et al., 1999). Sixteen of the materials that were reviewed were given the risk phrase R46: May cause heritable genetic damage. These materials are therefore likely to exert all or some of their effects on fertility via genetic damage to germ cells: the chromium compounds, some of the cadmium compounds (cadmium fluoride, cadmium chloride, and cadmium sulphate), benzo[a]pyrene, 1,2-dibromo-3-chloropropane, 2-nitrotoluene, carbendazim (ISO), benomyl (ISO), and acrylamide. In addition, although insufficient information could be obtained in the literature to determine a selective effect on rodent fertility for 4,4 0 -oxydianiline and its salts, the fact that this material has been classified a germ cell mutagen clearly indicates a threat to fertility. 2.2. Direct damage to Sertoli or Leydig cells Sertoli cells anchor and nourish developing germ cells, and are thought to play a critical role in the coordination of spermatogenesis. Damage to Sertoli cells can be histologically apparent as vacuolation and/or reduction in cell number, and as it is thought that Sertoli cells play a major role in spermiation, retention of stage VIII spermatids (Lee and Hendel, 1998). Chemicals that have been implicated as Sertoli cell toxicants include certain phthalate esters (DEHP, DBP, and BBP), and 2,5-hexanedione. However, it is important to note that due to the interdependence of the cells involved in spermatogenesis (germ, Sertoli, and

248

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

Leydig cells) and the hormonal support upon which each relies, it is not generally possible to characterise the primary target cell of testicular toxicants from standard regulatory toxicity data. 2.3. Physiological change The process of spermatogenesis is very sensitive to changes in oxygen tension and body temperature. Therefore, toxic insults that result in changes in lung function, erythropoesis, or raised body temperature can be deleterious to normal sperm production (Sharpe, 1998). The testicular toxicity of cadmium compounds is thought to be mainly due to a change in a physiological parameter, namely alteration of blood flow through the testis, resulting in ischaemic necrosis and complete loss of spermatozoa (Timbrell, 1991). Furthermore, chemicals that cause a decrease in blood pressure could lead to erectile dysfunction, thus affecting mating behaviour without affecting the histological appearance of the gonads. 2.4. Hormonal modulation The entire reproductive cycle of males and females is under hormonal control from the early stages of embryonic development all the way to successful progeneration. It is therefore not surprising that alteration of the level of androgens or oestrogens in an organism or the ability of endocrine tissues to respond to them may affect reproductive success. There has been increasing awareness and interest in recent years in the ability of xenobiotics to interfere with the hormonal status of animals. Such ‘endocrine disrupting chemicals’ (EDCs) may act via a number of mechanisms, including: • Direct interaction with oestrogen or androgen receptors. • Perturbing hypothalamic-pituitary control of sex steroid secretion. • Enzyme activation/inactivation in hormone synthesis pathways in particular e.g. aromatase.

2.4.1. Direct interaction with oestrogen or androgen receptors Androgens and oestrogens are often thought of as the main male and female sex hormones, respectively. It should however be remembered that both hormones are critical in regulating development and function of tissues in both sexes (Hess, 2003), and sex steroid receptors are not only expressed in sex and accessory organs, but also in many other tissues (White and Parker, 1998). Although this brief review separates receptor agonists and antagonists into distinct categorises, it should be borne in mind that it is not uncommon for hormonally active xenobiotics to exhibit mixed agonist/antagonist activity. This could be due to the different receptor subtypes responding differently to a chemical, the distribution of receptors in different tissues,

the dose applied, or the level of endogenous hormone levels. 2.4.1.1. Oestrogen agonists. A number of chemicals are able to bind to ERs and initiate transcriptional activity due to their structural similarity to endogenous hormones. Among these are diethylstilboestrol (DES), the first material that was acknowledged to elicit toxicity by mimicking an endogenous hormone. DES, a synthetic oestrogen, was used therapeutically in women that suffered from recurrent miscarriage, and caused reproductive tract abnormalities in both male and female offspring that were exposed to this chemical in-utero. The most notable lesions reported were vaginal adenosis and clear cell carcinomas in young women whose mothers had been exposed to high levels of DES during pregnancy (Herbst et al., 1971). Plants are another source of xenoestrogens. Although only weakly oestrogenic, the isoflavones genistein and diazein are present in significant quantities in soy products. Due to the ubiquitous nature of ERs, the biological effects of oestrogens are many and varied. For example, 17-b oestradiol has been shown to cause Leydig cell atrophy, seminiferous tubule degeneration, and reduced sperm production when administered to adult male rats (Cook et al., 1998). This is presumably due to disruption of the normal feedback mechanisms that exist to control oestrogen secretion from Leydig cells. Other potential effects of oestrogen agonists that are associated with fertility and reproduction in male and female animals are listed below. It is however important to note that due to the different characteristics of xenoestrogens (e.g. potency, metabolism) not every chemical that activates the ER in-vitro is active in-vivo, or elicits all these effects: • Increases in oestrogen-dependent tissue weights such as the uterus. • Precocious puberty in females. • Delayed puberty in males. • Altered oestrous/menstrual cycles. • Testicular atrophy. • Effects on female fertility. • Effects on mammary gland cell growth leading to neoplasia. • Developmental abnormalities (DES). (Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment, 2003; Hess, 2003; Sharpe and Skakkebaek, 1993). 2.4.1.2. Oestrogen antagonists. Oestrogen antagonists bind with ERs and displace the endogenous ligand, but as they do not initiate transcription they have antioestrogenic activity. Oestrogen antagonists are much less common than oestrogen agonists, but this class of materials includes the breast cancer treatment Tamoxifen and the industrial chemicals polychlorinated dibenzo-p-dioxin (PCDD) and

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

2,3,7,8-tetrachloro-dibenzo-p-dioxin. The antioestrogenic activity of oestrogen antagonists can result in: • Decreases in oestrogen-dependent tissue weights such as the uterus. • Altered oestrous/menstrual cycles. • Testicular atrophy. • Effects on male and female fertility. • Developmental abnormalities. (Silvers and Rorke, 1998; Oliveira et al., 2001; MacGregor and Jordan, 1998). 2.4.1.3. Androgen agonists. Chemicals that activate the androgen receptor are much less common than those that activate the oestrogen receptor. In a study which tested the oestrogen and androgen agonism and antagonism of 200 pesticides in an in-vitro reporter gene assay using Chinese hamster ovary cells, 44 were identified as oestrogenic, 5 were identified as antioestrogenic, 66 were antiandrogenic, and none was androgenic (Kojiama et al., 2004). It should however be noted that androgen antagonists can act as agonists, depending on ligand binding affinity, concentration, and the presence of competing natural ligands. There are therefore a small number of chemicals that have been shown to bind and activate the AR, generally at high concentrations and in the absence of testosterone. These include hydroxyflutamide and metabolites of the fungicide Vinclozolin (Maassaad et al., 2002; Wong et al., 1995). As the AR regulates the expression of genes involved in the proliferation and differentiation of prostate cancer cells, exposure to an androgen agonist could result in an increased risk of prostate cancers (Culig et al., 2003). 2.4.1.4. Androgen antagonists. Antiandrogens reduce the ability of androgen-dependent tissues and feedback mechanisms to respond to endogenous stimuli, leading to hypersecretion of LH from the pituitary. Therefore, when the fungicide Vinclozolin or some of its metabolites are administered to adult rats they can cause Leydig cell hyperplasia, atrophy of the prostate, and seminal vesicles and loss of spermatogenesis. However, the developing organism is generally more sensitive to interference by endocrine modulators than is the adult (Gray and Kelce, 1996), and the two developmental stages when rodents are unsurprisingly particularly sensitive to androgen antagonists are the pre-natal period (during differentiation of the reproductive tract) and the peri-pubertal period (around the time of sexual maturation) (Kavlock and Cummings, 2005). These periods coincide with respective surges in testosterone levels. Classic findings related to the administration of androgen antagonists during gestation and the post-natal period are hypospadias, cleft phallus, decreased anogenital distance and retention of nipples in male offspring, delayed sexual maturation and long-term effects on sex and accessory organs (Wong et al., 1995; Kavlock and Cummings, 2005).

249

2.4.2. Perturbing hypothalamic-pituitary control of sex steroid secretion Sex steroid secretion is controlled by the release of gonadotropin-releasing hormone (GnRH) from the hypothalamus and subsequent release of the gonadotropins follicle stimulating hormone (FSH) and luteinising hormone (LH) from the anterior pituitary gland. These chemicals in turn are regulated by several types of feedback signals and multiple neurochemicals. As a result, many different chemicals can disrupt hormonal status indirectly, that is without directly interacting with the ER or AR. For example, administration of general anaesthetics such as sodium pentobarbital or phenobarbital, chemicals that disrupt the synthesis of noradrenaline or agents that interfere with a-noradrenergic receptor stimulation to rats just before the LH surge which occurs on the afternoon of proestrus can delay ovulation. Also, while d-9-tetrahydrocannabinol (the primary psychoactive in marijuana) is not oestrogenic, it can delay the sexual development of female rats. Although the mechanism of this disruption is currently unknown, it is generally regarded to be mediated by altered hypothalamic regulation of pituitary function (Cooper et al., 1998). Interference with male fertility by this mode of action is less documented. This is perhaps due to the non-cyclical nature of male reproductive hormone stimuli, and the fact that basal levels can vary significantly without apparently affecting sperm production (Cooper et al., 1998). 2.4.3. Enzyme activation/inactivation Another physiological process that could be disturbed by xenobiotics and lead to a hormonal imbalance and subsequent effects upon fertility is the synthesis or clearance of the many hormones involved in reproduction. For example, azole fungicides exert effects upon reproductive organs, fertility, and development in a number of species, an effect thought to be due to inhibition of sterol 14a-demethylase and aromatase (Zarn et al., 2003). Aromatase is utilised in the conversion of androgens to oestrogens and 14ademethylase catalyses the formation of cholesterol (which in turn is metabolised to form sex steroids). Aromatase is necessary for sexual differentiation of the brain, and the activity and expression of this enzyme are high during the late embryonic and early neonatal period (Tsuruo, 2005), so during this period animals will be especially vulnerable to effects of aromatase inhibitors. Therefore, while a reproductive or repeat-dose toxicology study was not available for epoxiconazole, the toxicity of azole fungicides is well known, thus giving an alert for potential class effects of this material. Pesticides such as Mirex induce hepatic enzymes such as UDP-glucoronosyltransferase and monooxygenases, which eliminate testosterone (Maassaad et al., 2002). This can therefore lead to retarded sexual maturation in young males and impaired spermatogenesis in adults. A total of five materials caused reproductive effects that may perturb mating performance and/or fertility at doses that did not lead to a pathological correlate in a sub-chronic

250

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

repeat dose study. These materials were linuron, quinomethionate, nonylphenol, azafenidin, and fenarimol. 2.5. Linuron Linuron (CAS 330-55-2) is classified as R61 Rep Cat 2: ‘may cause harm to the unborn child’ and R62 Rep Cat 3: ‘possible risk of impaired fertility’. This herbicide is thought to exert all or some of its effects by androgen antagonism (McIntyre et al., 2000). The rat sub-chronic toxicity data quoted as part of the EPA’s re-registration eligibility decision (RED) (EPA, 1995) quotes a NOEL of 80 ppm (quoted as equivalent to 4 mg/kg/day) based on haematological findings at levels of 400 ppm (quoted as 20 mg/kg/day) and above. These changes were decreased red blood cell counts and increased white blood cell counts. In addition, at the high level of 3000 ppm (quoted as 150 mg/kg/day) growth was retarded. However, although there was no evidence of toxicity to the reproductive system in the sub-chronic study. In a 2-year carcinogenicity study, administration of linuron at dietary levels of 125 or 625 ppm (quoted as 6.25 or 31.25 mg/kg/day) was associated with an increased incidence of interstitial (Leydig) cell adenomas. Haematological changes confirming increased erythrocyte destruction and turnover were also apparent at these dosages (increased mean cell volume, decreased erythrocyte counts, and possible reticulocytosis). The overall NOEL in that study was 50 ppm (2.5 mg/kg/day). Leydig cell hyperplasia and adenoma formation can be caused by a number of modes of action, including androgen antagonism (Clegg et al., 1997). The failure to detect any changes in the reproductive system in the sub-chronic study may have been due to a number of factors. Firstly, the study may not be as reliable as a contemporary guideline study performed to good laboratory practice (GLP). Since the advent of GLP regulations and guidelines in the 1970s, it has become easier to distinguish between studies of high quality and studies of low quality. Secondly, as the study was performed in 1963, it may not be as sensitive as a more contemporary study. For example, it is possible that the testes would have been preserved in formalin rather than Davidson’s or Bouin’s fixatives that are recommended by modern regulatory guideline. As formalin does not preserve the morphologic detail of the testes as well as these alternative fixatives (Latendresse et al., 2002), slides prepared from formalinfixed testes yield slides of a lower quality, thus reducing the sensitivity of the assay. Furthermore, different staining techniques (e.g. haematoxylin and eosin, periodic acidSchiff, or immunohistochemical staining) may result in inter-study variation in interpretation of histological findings. However, the differences that are noticed between these techniques are generally within the seminiferous tubule, such as differential tubular shrinkage with different fixatives, or difficulty in differentiating between Sertoli cells from early spermatogonia or late spermatogonia from spermatocytes with different staining techniques (Laten-

dresse et al., 2002). Overt changes in the interstitial space such as Leydig cell adenoma would likely be identified if the tissues were fixed in formalin and stained with H&E. Thirdly, Leydig cell adenomas are relatively common lesions in chronic studies, but much less so in sub-chronic tests. In all but the most potent antiandrogens it is likely that a treatment period in excess of 3 months would be necessary to cause the chronic LH over stimulation that would lead to Leydig cell hyperplasia/adenoma. This is considered the most likely explanation, and is supported by two 2-generation studies, where parental generation males would have started treatment from young adulthood and terminated after much less than a year of treatment. In these parental males, no Leydig cell hyperplasia or adenomas, and no other testicular or sperm changes were observed at dosages up to around 50 mg/kg/day. However, F1 generation males that were examined in adulthood (i.e. males potentially exposed in-utero and throughout independent life) showed testicular atrophy and intratubular fibrosis, epididymal inflammatory response, or oligospermia at that dose. F1 generation and F2 generation litters also contained fewer offspring than their control counterparts. In-vitro analysis showed that linuron had a weak affinity for the androgen receptor (EPA, 1995). These results are therefore consistent with the hypothesis that the developing reproductive tract is more sensitive to androgen antagonists than are the adult sex organs (Kavlock and Cummings, 2005). However, there was no evidence in a rat pre-natal developmental toxicity study that linuron was antiandrogenic (EPA, 1995). In that study the NOAEL was 12.1 mg/kg/day, as at the higher dose of 49.8 mg/kg/day increased post-implantation loss was observed. Increased fetal death was also a feature of linuron toxicity in a rabbit developmental toxicity study (NOAEL 25 mg/kg/day, LOAEL 100 mg/kg/day). Therefore, while a 90-day repeat-dose toxicity study may not identify possible effects on fertility at a dosage that caused a bodyweight effect (150 mg/kg/day), in-vitro assays show that linuron competes with androgens for binding to the androgen receptor (McIntyre et al., 2000). Furthermore, rat (and also rabbit) developmental toxicity studies showed that linuron has the potential to cause in-utero death. In conclusion, in the case of Linuron, an in-vitro receptor binding or transcriptional assay would have determined that this material has the potential to interact with the androgen receptor. This would logically lead to the question of whether animals in the embryonic/fetal/neonatal period would be more sensitive to this aspect of linuron’s toxicity than would adults. Such an alert could then be addressed in a littering (e.g. 2-generation) study.

2.6. Quinomethionate Quinomethionate (also known as Chinomethionat and Morestan) (CAS 2439-01-2) is a quinoxaline fungicide,

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

251

and has been classified as R62 Rep Cat 3: ‘possible risk of impaired fertility’. Unless otherwise stated, the following information was taken from WHO Pesticides Residues in Food, 1987 (FAO/WHO, 1987). In a sub-chronic study performed by Bayer and reported in 1983 groups of 20 male and 20 female rats were administered test diets containing 0 (control), 10, 25, 60, 150, or 500 ppm quinomethionate for 3 months. These doses approximately correspond to 1, 2.5, 6, 15, or 50 mg/kg/ day for young adult rats.1 The study design appears to contain most of the requirements of the current OECD 408 guideline: daily clinical observations, weekly bodyweight, and food consumption measurement, haematology, blood chemistry, and urinalysis after 4 and 13 weeks’ treatment, a full necropsy and histological examination of selected tissues. Treatment-related findings included yellow coats in male rats treated at 50 mg/kg/day and reduced bodyweight and food consumption in males and females at 15 and 50 mg/kg/day. Haemoglobin, haematocrit, red blood cell number, and mean cell haemoglobin were all decreased in male and female rats at 50 mg/kg/day and females also showed reduced white blood cell counts and males reduced platelet counts at this dose. Haemoglobin, haematocrit, red blood cell number were reduced in females at 15 mg/kg/ day, and females at 6 mg/kg/day showed decreased mean cell haemoglobin at 3 months. Urinary protein was decreased in both sexes at the end of the treatment period at 50 mg/kg/day. At necropsy, relative liver and kidney weights were increased in females at 15 and 50 mg/kg/day and relative lung and testes weights were increased in males at the same doses. Relative brain weights were increased at 6 (males only) and at 15 and 50 mg/kg/day in both sexes. However, since the brain and the testes are highly conserved organs these changes in relative weights may be due to the decreased bodyweights at these doses. In the absence of any reported histological changes in these organs this increased testicular weight has been taken as artefactual rather than an indication of toxicity. The NOAEL for the study was taken as 6 mg/kg/day. Like linuron, there were no clear indications of target organ toxicity to the reproductive system in this subchronic test. In a similar way, a longer term study did show changes, namely epididymal and testicular toxicity (HSE, 2003). However, unlike linuron, quinomethionate is not an androgen antagonist, and it does not interact with the oestrogen receptor (Kojiama et al., 2004). It therefore appears that quinomethionate acts by an indirect mechanism to cause these effects.

A two-generation (two litters per generation) study, reported in 1984 was available. In that study, groups of 10 male and 20 female rats were exposed to diets containing 0 (control), 15, 60, or 240 ppm. These doses are approximately equal to 0.75–1.5, 3.0–6.0, or 12–24 mg/kg/day for adult animals and 4.5, 18, and 72 mg/kg/day for weanlings. Standard indices of mating performance and fertility as well as general observations such as clinical signs and bodyweight were recorded. This study showed that at 240 ppm (12–24 mg/kg/day for adults) litter size was significantly reduced in the F2a generation (second generation first litter). There was also increased post-natal mortality at this dose in the F1a, F2a, and F2b offspring (72 mg/ kg/day). The NOAEL for this study was 15 ppm (0.75– 4.5 mg/kg/day) based on maternal toxicity, and there were apparently no effects on fertility or reproduction at 60 ppm (3–18 mg/kg/day). While one rat pre-natal developmental toxicity study showed no evidence of in-utero mortality at doses up to 62.5 mg/kg/day, a number of other studies show that this chemical has the potential to cause post-implantation loss at doses that do not cause marked maternal toxicity (EPA, 1987). In one study groups of 25 female rats were dosed by gavage at 0 (control), 50, 75, 90, or 110 mg/kg/ day between days 6 and 15 of gestation. The maternal NOEL was set at 90 mg/kg/day based on an equivocal decrease in bodyweight gain and food consumption, while the developmental NOEL was set at 50 mg/kg/day due to increased post-implantation loss and malformations (unspecified). Although the results of studies for this chemical are conflicting, the weight of evidence is suggestive of a possible effect on fertility and development, in the absence of clear effects on reproductive organs in a sub-chronic test. It is possible that these effects could be related to quinomethionate’s ability to inhibit sulphydryl enzymes including pyruvate dehydrogenase, succinate dehydrogenase, malate dehydrogenase, and a-ketoglutarate oxidase (Carlson and DuBois, 1970). Inhibition of enzymes required for cellular respiration can impact fertility (Miki et al., 2004), and there are also several examples of developmental toxins that act on the conceptus by inhibiting key enzymes (6-aminonicotinamide and 5-fluorouracil inhibiting glucose-6-phosphate dehydrogenase and thymidylate synthetase for example). Although studies performed to current OECD guidelines and to GLP may provide more confidence in the presence or absence of effects, most of the studies described above were performed after the advent of GLP in a competent laboratory. It is therefore probable that if such a negative sub-chronic test were available, unless it was considered of low quality, it would not be repeated for animal welfare and economic reasons.

1 Where mg/kg/day equivalents based on food consumption were not quoted in source data, the following calculation was used to derive approximate equivalents: weanling rats: dietary level in ppm · 0.3 = dose in mg/kg/day; young adult rats: dietary level in ppm · 0.1 = dose in mg/ kg/day; rats >90 days of age dietary level in ppm · 0.05 = dose in mg/kg/ day.

2.7. Nonylphenol Nonylphenol (including branched 4-nonylphenol) (CAS 25154-52-3 and 84852-15-3, respectively) is widely used in the production of many commercial materials, and has

252

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

been classified as R62 Rep Cat 3: ‘possible risk of impaired fertility’ and R63 Rep Cat 3: ‘possible risk of harm to the unborn child’. It has long been known that nonylphenol binds to the ER and exhibits weak oestrogenic activity (White et al., 1994; Shelby et al., 1996). Nonylphenol is considered an endocrine disrupting chemical (EDC) of importance, due to its use in many processes and its well known presence in the environment. However, it is well known that doses of an EDC that do not exhibit oestrogenic activity in sexually mature animals may affect developing organisms (Gray and Kelce, 1996). In a GLP rat 90-day sub-chronic toxicity study (Cunny et al., 1997) branched 4-nonylphenol was administered to 4 groups of rats at dietary concentrations of 0 (control), 200, 650, or 2000 ppm (doses quoted as approximately 15, 50, or 150 mg/kg/day, respectively). In addition to the standard measurements taken in an EPA sub-chronic toxicity study, vaginal smears were taken during week 8 of treatment for assessment of the regularity of the oestrous cycle and sperm counts; motility and morphology were assessed at termination. The only toxicologically significant findings in this study were a small reduction in food consumption and bodyweight gain at 150 mg/kg/day. There were therefore no effects indicative of oestrogen agonism at a dosage that elicited slight general toxicity, and the NOAEL for the study was set at 50 mg/kg/day. These results appear to conflict with results of uterotrophic assays in the immature rat where the LOEL for nonylphenol was 48 mg/kg/day (Moffat (1996) cited in Cunny et al. (1997)). This difference is likely to be due to the very weak oestrogenicity of nonylphenol. In the neonatal rat model, there is effectively an absence of background oestrogen, meaning that very weak oestrogens can be identified. However, in the adult rat model (as used in the sub-chronic study) any additional oestrogenic effect of nonylphenol is likely to be biologically inconsequential compared to the normal effects of circulating endogenous oestrogens. In a 3-generation study that appeared to meet or exceed the requirements of OECD test guideline 414 (Chapin et al., 1999), groups of Sprague Dawley rats were exposed to nonylphenol at dietary levels of 0 (control), 200, 650, or 2000 ppm (doses quoted as approximately 9–35, 30–100, or 100–350 mg/kg/day, respectively). Treatment was throughout 3 successive generations, and the study was terminated when the F3 animals were weaned. Unlike the other mate-

rials included in Table 3 there was no effect upon the fertility of these rats. This material was included, however, because while there were no effects on the reproductive system of parental animals, the reproductive systems of their offspring were affected. These effects were precocious vaginal opening at 30–100 and 100–350 mg/kg/day and protracted oestrous cycles in F1 and F2 animals at 100–350 mg/kg/day. These are all findings that can be linked to an oestrogenic mode of action. In addition, there were equivocal effects on F2 generation sperm numbers at 30–100 and 100–350 mg/kg/day. However, the difference was slight and may have been a chance finding unrelated to treatment. 2.8. Azafenidin The herbicide azafenidin (CAS 68049-83-2) has been classified as R61 Rep Cat 2: ‘may cause harm to the unborn child’ and R62 Rep Cat 3: ‘possible risk of impaired fertility’. Azafenidin acts by protoporphyrinogen oxidase inhibition, and other herbicides with this mode of action are known to be developmental toxins (Kawamura et al., 1996). The following in-vivo data are all taken from California Environmental Protection Agency Department of Pesticide Regulation Medical Toxicology Branch Summary of Toxicological Data for Azafenidin (EPA, 2000). Groups of 10 male and 10 female rats were administered Azafenidin in the feed at dietary concentrations of 0 (control), 50, 300, 900, or 1500 ppm for 90-days in a subchronic toxicity study reported in the 1990s. Three females at 1500 ppm and one control female died during the study. Treatment at 1500 ppm was associated with impaired bodyweight gain and pallour in both sexes. Decreased haemoglobin and haematocrit and increased reticulocyte number and methaemoglobin concentrations were seen at 900 and 1500 ppm (approximately 90 and 150 mg/kg/day) The numbers of red blood cells and white blood cells were decreased and increased, respectively, at 1500 ppm. Microscopic examination revealed a dose-related increased incidence of pigmented hepatic Kupffer cells, increased extramedullary erythropoiesis in the spleen and increased erythropoiesis in the bone marrow at 900 and 1500 ppm. The NOAEL was set at 50 ppm (reported to be 4.02– 4.73 mg/kg/day) because at doses of 300 ppm and above

Table 3 potential alerts in the toxicological database for fertility Chemical

Effects in repeat-dose toxicity study

Effects in pre-natal developmental toxicity study

ER or AR (ant)agonist?

Azafenidin

Increased red blood cell turnover

No

Quinomethionate Fenarimol

Increased red blood cell turnover Enzyme induction

Increased resorptions, reduced fetal weights, sternebral variations Increased resorptions, malformations Increased incidence of hydronephrosis

Linuron Nonylphenol

Haematological changes Bodyweight and food effects only

increased post-implantation loss No developmental toxicity

No Yes: AR antagonist and ER agonist Yes: AR antagonist Yes: ER agonist

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

a dose-related decrease in cytochrome P-450 activity was apparent. It should however be noted that although this test was termed a 90-day study, it appears that the treatment period was longer than 90-days, because it was reported that after 90 days’ treatment these animals went on to be used in a supplemental investigation of reproductive performance. In this supplementary investigation animals were mated on a 1:1 basis within treatment groups until evidence of copulation was observed. Exophthalmus was apparent in three animals at 900 ppm (approximately 45 mg/kg/day), and bodyweight gain during gestation was impaired at 300 and 900 ppm (approximately 17.1 and 51.3 mg/kg/day). This would have been at least in part due to the failure of females in these groups to produce a litter. The NOAEL for this portion of the study was therefore 50 ppm (reported to be 5.4 mg/kg/day during this phase). A full two-generation study was also performed on azafenidin and reported in 1995. Groups of 30 male and 30 female CD rats received azafenidin at dietary levels of 0 (control), 5, 30, 180, or 1080 ppm through two successive generations. The pre-mating treatment period for the parental (F0) animals was 70 days, in-line with OECD test guideline 416. The parental NOEL was 30 ppm (approximately 1.7–2.3 mg/kg/day for F0 animals and 2.3–2.8 mg/ kg/day for F1 animals) due to diarrhoea in F1 males at 180 ppm and impaired bodyweight performance in F1 males and females during their pre-mating period at 180 ppm. Reproductive effects included decreased implantation efficiency, increased gestation length, decreased litter size and offspring weights at 180 ppm and above. The reproductive NOAEL was therefore also 30 ppm (1.7– 2.8 mg/kg/day). It is not clear from the publicly available data whether other standard investigations such as necropsy, organ weights and histopathological examination were performed. A number of developmental toxicity studies have been performed on this material. In one test reported in 1995, groups of 25 mated female CD rats were dosed by gavage with corn oil suspensions containing 0 (control), 3, 8, 16, or 24 mg/kg/day azafenidin from gestation days 7 to 16. Increased resorptions and reduced litter size were noted at 16 and 24 mg/kg/day, and fetal weights were reduced at 24 mg/kg/day. The NOAEL for developmental toxicity was therefore 8 mg/kg/day. It is assumed that standard fetal evaluations were performed on this study but there is no specific mention that these were performed in the summary information. A dermal study performed in the rat also showed clear evidence of developmental toxicity at doses of 25 mg/kg/day and above. In that test, dosages of 0 (control), 5, 25, 50, or 100 mg/kg/day were administered at a dose volume of 2 ml/kg for a period of 6 h/day. There were no signs of maternal toxicity in any group. Mean gravid uterine weights and fetal body weights were reduced at 50 and 100 mg/kg/day, and fetal weights and total resorptions were increased at 25 mg/kg/day. A dose-related increase in bent long bones was also reported

253

at doses of 25 mg/kg/day and above. The NOAEL for developmental toxicity was therefore 5 mg/kg/day. Along with quinomethionate, azafenidin serves to illustrate that information other than pathology of the reproductive tract may be indicative of an increased likelihood of effects upon fertility/developmental toxicity. Protoporphyrinogen oxidase is one of the key enzymes in porphyrin synthesis as part of haem biosynthesis (Kawamura et al., 1996), hence the effects on red blood cell indices in the repeat-dose toxicity studies. Other inhibitors of haem biosynthesis are known to be developmentally toxic, such as hexachlorobenzene, griseofulvin, diazinon, Phenobarbital, lead, and SLA 3992 (Kawamura et al., 1996), and, as with the case of quinomethionate it is unsurprising that the potent inhibition of such a critical biosynthetic pathway would affect the complex series of interactions that occur around the time of conception and implantation. Therefore, from the sub-chronic data alone it may not be immediately apparent that azafenidin may pose a risk to reproduction, as sex and accessory organs were apparently unaffected by treatment. However, the inclusion of data from developmental toxicity studies and knowledge of the mode of action of the material make it quite clear that this material has the potential to affect fertility and development. 2.9. Fenarimol Fenarimol (CAS 60168-88-9) is a fungicide that acts by inhibiting ergosterol synthesis. It is classified as R62 Rep Cat 3: ‘possible risk of impaired fertility’ and R63 Rep Cat 3 ‘possible risk of harm to the unborn child’. Unless otherwise stated, the following information was taken from the Food and Agriculture and World Health Organisation’s evaluation of pesticides in food (FAO/WHO, 1995). This evaluation commented that many of the original studies were poorly reported, however as much information as was obtainable is reported here. A sub-chronic toxicity study was performed on fenarimol in 1975, which while deficient by the standards of today’s OECD 408 guideline gives some indication of the effects of repeated administration of fenarimol. Groups of 20 male and 20 female Wistar rats were exposed to diets containing 500, 200, or 800 ppm fenarimol for a period of 3 months. These resulted in exposure levels of 2.5, 10, or 40 mg/kg/day. A control group containing 25 male and 25 female rats were exposed to untreated diet throughout the same period. Five animals of each sex per group were retained at the end of treatment for a two-week recovery period. Aside from clinical signs, food consumption, and bodyweight measurements, investigations carried out during the study included ophthalmoscopy, haematology, clinical chemistry (although only blood urea nitrogen, glucose, and alanine aminotransferase were measured), organ weights, histopathology (no nervous tissue examined), and determination of hepatic para-nitroanisole metabolism in

254

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

five animals per sex per group. Males at 40 mg/kg/day showed a slight decrease in bodyweight gain, and there was a dose-related increase in liver weight for both sexes, which became statistically significant at 40 mg/kg/day. Relative kidney weights of males and females and thyroid weights of females only were also increased at 40 mg/kg/ day. Hepatic para-nitroanisole metabolism showed a significant dose-related increase in males at 10 and 40 mg/kg/day and was increased in females at 40 mg/kg/day only. Histological examination revealed hepatic centrilobular hypertrophy in males at 40 mg/kg/day and a shift from slight to moderate fatty metamorphosis in females at the same dose. The only effect apparent at the end of the two-week recovery period was increased liver weights at 40 mg/kg/ day. The NOAEL was therefore set at 50 ppm, or 2.5 mg/kg/day. An 18-month test was also performed on fenarimol. The results of this study correlated well with those of the 90-day test, with similar findings and a similar NOAEL. One difference was an increase in ovarian weights that was apparent at the end of the treatment period at 16– 22.9 mg/kg/day. As previously mentioned, as very few 18month studies will be performed under REACH it was decided not to take this alert for effects on reproduction into account as it was not expressed in the 90-day study. In one rat developmental toxicity study, groups of 25 mated female Wistar rats were dosed with fenarimol by gavage at doses of 0 (control), 5, 13, or 35 mg/kg/day between days 6 and 15 of gestation. There was no maternal toxicity evident, and the only reported finding in fetuses was an increased incidence of hydronephrosis at 35 mg/ kg/day (30% of fetuses in 62% of litters compared with 9% of fetuses in 25% of litters in the control group). A follow-up study was therefore performed to investigate the reversibility of this effect. That study involved dosing groups of mated female Wistar rats at 0 (control) or 35 mg/kg/day between days 6 and 15 of gestation. Two groups containing 25 and 15 rats were terminated on day 20 or 21 of gestation, respectively. Other rats were allowed to litter and their progeny were examined on days 1, 7, 21, 42, or 63 of age. It is reported that 15 control and 20 treated litters were examined post-natally, however it is unclear whether 15 and 20 litters were examined on each of the above-mentioned days or whether the sacrifice of 15 and 20 litters was spread between these days. Nevertheless, as the results of the pre-natal phase of the study are those of interest, the power of the post-natal phase of the study is of limited consequence in this evaluation. The results of the pre-natal phase do conflict somewhat with the reported outcome of the previous pre-natal development study. It is reported that at day 20 of gestation, fetal weights were slightly lower than control. As effect was no longer evident by day 21, it is possible that the previous study was terminated on this later gestation day. However, it is also reported that there was a slight increase in the number of early resorptions, an effect that would be evident at either day 20 or day 21 of gestation. A slight increase in skeletal variants such as cervical ribs and 14 thoracic ribs

was also observed, although only day 20 fetuses were examined skeletally. Some (but not all) the effects seen in these studies appear to be indicative of fetal retardation rather than frank teratogenicity (a conclusion confirmed by the post-natal investigations). Given the differences between fetal weights at day 20 or day 21 of gestation, this retardation appeared to reverse following cessation of maternal treatment on gestation day 15. Therefore, an OECD 414 guideline study would likely result in more profound fetal effects, as treatment continues up to gestation day 19 or 20 (the day before sacrifice). A number of rat multigeneration studies, including a cross-over mating trial to identify the affected sex have been reported in the literature (Hirsh et al., 1986). In these tests, parental (F0) weanling animals were fed diets containing 0 (control), 50, 130, or 350 ppm fenarimol until they were approximately 90 days of age. These doses equated to approximately 2.5, 6.5, or 17.5 mg/kg/day. The group size was 30 males and 30 females in the control group and 20 males and 20 females in the treated groups. After 90-days of age the animals were mated on a 1:1 basis within treatment groups for three matings. Offspring from the second mating were allocated to the next phase of the study, and treated with the same test diets as their parents for 58 days. These animals were then mated within treatment groups and their fertility assessed during exposure to the test diets and after a recovery period of 63 days. Treatment with fenarimol was associated with time and dose-dependent infertility. At the initial pairing, there was a suggestion of impaired fertility at 17.5 mg/kg/day (75% of females pregnant compared with 93% controls), although statistical significance was not attained. By the 3rd mating there was a clear effect, and the percentage of pregnant females throughout the groups was 93%, 89%, 63%, and 59% from control to high dose, respectively, with values at the top two doses attaining statistical significance. It should be noted that by the time of the third mating trial the animals had received 7.5 month’s treatment, far in excess of the OECD 416 guideline requirement of 10 weeks (2.5 months). The F1 generation animals derived from the second pairing showed a clear treatment-related infertility at 6.5 or 17.5 mg/kg/day, but not at 2.5 mg/kg/day. The reversibility phase demonstrated that these effects were only partly reversible. Importantly, there was no evidence of an effect of fenarimol on the macro or microscopic appearance of the male or female sex organs, and spermatogenesis, as judged by evaluation of testis sections, was apparently unaffected. A cross-over mating trial demonstrated that the male was the affected sex. Further studies have shown that fenarimol is an aromatase inhibitor, and while it does not readily cross the placenta it is excreted in milk and concentrates in the hypothalamus of neonates (Hirsh et al., 1987). This led to the hypothesis that fenarimol may exert some or all its effects on the male via a central inhibition of aromatase activity, thus depressing male sexual behaviour. Later studies have confirmed that fenarimol is a potent aroma-

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

tase inhibitor and that it also has the ability to act as an oestrogen agonist and an androgen antagonist (Andersen et al., 2002; Kojiama et al., 2004). There are therefore several modes of action by which this material may affect reproduction and development. It is important to note that relatively simple and cheap in-vitro receptor binding assays or enzyme activity assays could have provided an alert for these effects before the commencement of the animal tests. 3. Discussion The upcoming European chemicals legislation REACH (Registration, Evaluation, and Authorisation of Chemicals) will require the risk assessment of many thousands of chemicals. It is therefore necessary to develop intelligent testing strategies to ensure that chemicals of potential concern are prioritised for appropriate testing to inform a robust risk assessment. This will avoid the unnecessary testing of chemicals that are of low concern. One of the endpoints assessed under the REACH requirements is mating performance and fertility. Animal tests that address this endpoint use a relatively large number of animals and are also costly in terms of resource, time, and money. Intelligent testing strategies therefore need to be developed which can take into account data from a wide variety of sources, avoiding the unnecessary use of animals, wherever possible. The level of concern for each chemical needs to be assessed, taking into account likely human exposure to the material (since a material which shows no human exposure cannot be a risk), and ‘alerts’ in each chemical’s available toxicological database. To determine the level of concern represented for any chemical it is necessary to have a good understanding of the relevance and importance of different types of data. This becomes especially important where the toxicological database that a risk assessor is basing a judgement on contains a prenatal developmental toxicity study and a repeat-dose toxicity study in a relevant species, but does not contain a study directly addressing reproductive performance and fertility. It is commonly thought that in such cases the absence of effects upon the gonads or accessory sex organs in the sub-chronic tests, coupled with lack of adverse effects in the prenatal development study provide sufficient reassurance that fertility is unlikely to be affected. However, this is a contentious issue and there appear to be no published reviews that substantiate this claim. It should be noted that this type of review cannot hope to predict the certainty that a particular chemical that has been the subject of a sub-chronic study and a pre-natal development study to disrupt fertility. There are of course several limitations of this approach, such as the fact that the severity of classification depends on robustness and quality of supporting data. This means that materials that show clear pathology are more likely to be classified for the fertility endpoint than materials that show more subtle functional effects. Therefore, the list of chemicals that has been chosen may contain a bias towards chemicals that show effects in repeat-dose studies, and any hypothesis

255

apparently supported by this set of chemicals needs to be tested using a much larger data set to test the ‘failure rate’ of this approach. However, useful principles may be highlighted regarding the modes of action that may lead to the generation of alerts, and the types of information that may increase a risk assessor’s confidence. Equally or more important is learning from the modes of action of chemicals that do not provide alerts in these data types but that adversely affect fertility. Available rat general and reproductive toxicity data for 44 chemicals that have been classified as toxic to fertility (R60 or R62) were analysed for evidence that effects on rodent fertility could reasonably expected to be predicted from a sub-chronic toxicity study. These data sets were considered relevant for 73 of the 102 chemicals currently classified as toxic to reproduction (fertility) under this system. Appropriate data for the remaining 29 chemicals were not found. For all but 5 of these chemicals it was considered that a well-performed sub-chronic toxicity study would have resulted in pathology in the male and in some cases the female reproductive tract. As with all toxicology studies, this is dose dependent, especially for those materials which are endocrine actives and require much higher doses to elicit effects in adult than in developing organisms. Of those five materials that did not cause effects in repeat-dose toxicity studies, but did affect rat fertility, three showed evidence of direct interaction with oestrogen or androgen receptors (linuron, nonylphenol, and fenarimol). The remaining chemicals (quinomethionate and azafenidin) act by modes of action that do not require direct interaction with steroid receptors. However, both these materials caused in-utero deaths in pre-natal developmental toxicity studies, and the relatively low NOAELs and the nature of the hazard identified in the sub-chronic tests provides an alert for possible effects on fertility (or early embryonic development). It is clear from this review that littering studies do sometimes give lower NOAELs than their corresponding pre-natal developmental or sub-chronic studies. It is therefore important that when an alert does exist and human exposure to a toxicant is expected that a definitive test is performed. In this way, an alert for effects on fertility would have been generated for all 73 chemicals if the following information were available: • A repeat-dose toxicity study. • A pre-natal developmental toxicity study. • In-vitro ER and AR receptor binding/transactivation studies. For most chemicals the alert would arise from the repeat-dose toxicity study. For those 5 chemicals that showed no evidence of target organ toxicity in the sex or accessory organs, the alerts would have come from the pre-natal development study where increased in-utero deaths were observed or due to their affinity for sex steroid receptors. Furthermore, it is clear that sub-chronic toxicol-

256

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

ogy alerts for potential effects on reproduction are not solely derived from effects on the reproductive or sex organs, as demonstrated by the effects on red blood cell turnover seen with azafenidin and quinomethionate (Table 3). Further alerts could have been generated from the knowledge of the mode of action of these materials. For example, while in the above table the ability of fenarimol to block and activate the AR and ER, respectively, the primary mode of action of fenarimol is thought to be inhibition of aromatase. This clearly indicates that the inherent biological reactivity of materials, such as their ability to inhibit key enzymes as well as affinity with the ER or AR needs to be taken into account during an initial evaluation. The use of structure activity relationships (SAR) to predict such inherent biological reactivity of chemicals is becoming more common (Bradbury et al., 1998). This could enable a further screen, even before in-vitro testing that would add further information to the database. 4. Conclusion From the chemicals reviewed it would appear that where there are no alerts from a repeat-dose toxicity study, a prenatal developmental toxicity study and sex steroid receptor binding assays, there exists a low priority for animal studies to address the fertility endpoint. Conversely, alerts in these data types could lead to priority testing. Furthermore, as biological reactivity such as the inhibition of key enzymes or the ability to bind to the ER and/or AR is a function of the chemical structure of materials, there is clearly a role for SARs in identifying those materials with the potential to have these properties. It should be remembered that fertility is just one segment of the reproductive cycle that is assessed in reproductive screening studies (OECD 421 or 422) that are recommended under REACH. In determining the true value of these screening studies it is necessary to not only repeat the present exercise with a larger data set, but also to find out how reliably these test protocols predict effects on littering and early post-natal survival of offspring. This information is critical in designing an integrated testing strategy and should be gathered as a matter of priority. Acknowledgment Thanks to Philip Carthew for advice on the direction and scope of this review. References Abdel-Rahman, M.S., Hetland, L.B., Couri, D., 1976. Toxicity and metabolism of methyl n-butyl ketone. American Industrial Hygiene Association Journal 37 (2), 95–102. Andersen, H.R., Vinggard, A.M., Rasmussen, T.H., Gjermandsen, I.M., Bonefeld Jørgensen, E.C., 2002. Effects of currently used pesticides in assays for estrogenicity, androgenicity, and aromatase activity in-vitro. Toxicology and Applied Pharmacology 179, 1–12.

Anon, 1991. Final report on the safety assessment of chloroacetamide. Journal of the American College of Toxicology 10 (1), 21–32. Archibong, A.E., Ramesh, A., Niaz, M.S., Inyang, F., Kopsombut, P.M., 2003. Pertubation of testicular and epididymal function of adult F-344 rats by sub-chronic inhaled benzo[a]pyrene exposure. Biology of Reproduction 68 (Suppl. 1), 182. Bempong, M.A., Hall, E.V., 1983. Reproductive toxicology of 1,3diphenylguanidine: analysis of induced sperm abnormalities in mice and hamsters and reproductive consequences in mice. Journal of Toxicology and Environmental Health 11 (4–6), 869–878. BIBRA Working Group, 1994. 2,4-Dinitrotoluene. Toxicity Profile. BIBRA Toxicology International. Available from: . Bradbury, S.P., Mekenyan, O.G., Ankley, G.T., 1998. The role of ligand flexibility in predicting biological activity: structure–activity relationships for aryl hydrocarbon, estrogen and androgen receptor binding affinity. Environmental Science and Technology 17, 15–25. Burns-Naas, L.A., Meeks, R.G., Kolesar, G.B., Mast, R.W., Elwell, M.R., Hardisty, J.F., Thevenaz, P., 2002. Inhalation toxicology of octamethylcyclotetrasiloxane (D4) following a 3-month nose-only exposure in Fischer 344 rats. International Journal of Toxicology 21 (1), 39–53. Cai, L., Hales, B.F., Robaire, B., 1997. Induction of apoptosis in the germ cells of adult male rats after exposure to cyclophosphamide. Biology of Reproduction 56, 1490–1497. Carlson, G.P., DuBois, K.P., 1970. Studies on the toxicity and biochemical mechanism of action of 6-methyl-2,3-quino-oxalinedithiol cyclic carbonate (Morestan). Journal of Pharmacology and Experimental Therapeutics 173, 60–70. Chapin, R.E., Delaney, J., Yefan, W., Lanning, L., Davis, B., Collins, B., Mintz, N., Wolfe, G., 1999. The effects of 4-nonylphenol in rats: a multigeneration reproduction study. Toxicogical Sciences 52, 80–91. Cunny, H.C., Mayes, B.A., Rosica, K.A., Trutter, J.A., Van Miller, J.P., 1997. Sub-chronic toxicity (90-day) study with para-nonylphenol in rats. Regulatory Toxicology and Pharmacology 26 (2), 172–178. Clegg, E.D., Cook, J.C., Chapin, R.E., Foster, P.M., Daston, G.P., 1997. Leydig cell hyperplasia and adenoma formation: mechanisms and relevance to humans. Reproductive Toxicology 11 (1), 107–121. Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment, 2003. Phytoestrogens and Health. Food Standards Agency. Available from: . Cook, J.C., Johnson, L., O’Connor, L.B., Krams, C.H., Frame, S.R., Hurtt, M.E., 1998. Effects of dietary 17 beta-estradiol exposure on serum hormone concentrations and testicular parameters in male Crl:CD BR rats. Toxicological Sciences 44 (2), 155–168. Cooper, R.L., Goldman, J.M., Tyrey, L., 1998. The hypothalamus and pituitary as targets for reproductive toxicants. In: Korach, K.S. (Ed.), Reproductive and Developmental Toxicology. Marcel Dekker Inc., New York. Culig, Z., Klocker, H., Bartsch, G., Steiner, H., Hobisch, A., 2003. Androgen receptors in prostate cancer. Journal of Urology 170 (4, Part 1 of 2), 1363–1369. Dunnick, J.K., 1992. Toxicity studies of o-, m- and p-nitrotoluenes. National Toxicology Program Toxicity Report Series, vol. 23. ECB, 2000. IUCLID Dataset for a,a,a,4-Tetrachlorotoluene. European Chemicals Bureau. Available from: . ECB, 2000a. IUCLID Dataset for 1,2-Benzenedicarboxylic Acid, di-C7-11branched and Linear Alkylesters. European Chemicals Bureau. Available from: . Elbetieha, A., Al-Hamood, M.H., 1997. Long-term exposure of male and female mice to trivalent and hexavalent chromium compounds: effect on fertility. Toxicology 116, 39–47. Ellis, M.K., Richardson, A.G., Foster, J.R., Smith, F.M., Widdowson, P.S., Farnworth, M.J., Moore, R.B., Pitts, M.R., Wickramartne, G.A.D.S., 1998. The reproductive toxicity of Molinate and metabolites

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258 to the male rat: effects on testosterone and sperm morphology. Toxiciology and Applied Pharmacology 151 (1), 22–32. EPA, 1994. Chemical Summary for Acrylamide prepared by Office of Pollution Prevention and Toxics, US Environmental Protection Agency. EPA Ref: 749-F-94-005a. Available from: . EPA, 1995. Reregistration Eligibility Decision—Linuron. United States Environmental Protection Agency: EPA 738-R-95-003. EPA, 1987. California Environmental Protection Agency Department of Pesticide Regulation Medical Toxicology Branch: Summary of Toxicology Data for Oxythioquinox. Available from: . EPA, 2000. California Environmental Protection Agency Department of Pesticide Regulation Medical Toxicology Branch Summary of Toxicological Data for Azafenidin Revised 9/7/2000. Available from: . EU, 2006. Regulation (EC) No. 1907/2006 of the European Parliament and of the Council of 18 December 2006 Concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals. Official Journal of the European Union L396/1. FAO/WHO, 1987. Chinomethionat (Pesticide residues in food: 1987 evaluations part II toxicology). International Program on Chemical Safety. Available from: . FAO/WHO, 1995. Fenarimol (pesticide residues in food: 1995 evaluations part II toxicology). International Program on Chemical Safety. Available from: . Foster, P.M.D., Creasey, D.M., Foster, J.R., Gray, T.J.B., 1984. Testicular toxicity produced by EGME and EGEE in the rat. Environmental Health Perspectives 57, 207–217. Gondzik, M., 1970. Histology and histochemistry of rat testicles as affected by carbon disulphide. Patologia Polska 21 (2), 133–139. Gray, L.E., Kelce, W.R., 1996. Latent effects of pesticides and toxic substances on sexual differentiation of rodents. Toxicology and Industrial Health 12, 515–531. Hansen, E., 1991. Effects on reproduction of n-hexane. Nordic Chemicals Control Group. KemI Report, vol. 10/91, 31–47. Herbst, A.L., Ulfelder, H., Poskanzer, D.C., 1971. Adenocarcinoma of the vagina. Association of maternal stilbestrol therapy with tumour appearance in young women. New England Journal of Medicine 284, 878–881. Hess, R.A., 2003. Estrogen in the adult male reproductive tract: a review. Reproductive Biology and Endocrinology 1, 52–65. Hirsh, K.S., Adams, E.R., Hoffman, D.G., Markham, J.K., Owen, N.V., 1986. Studies to elucidate the mechanism of fenarimol-induced infertility in male rats. Toxicology and Applied Pharmacology 86, 391–399. Hirsh, K.S., Weaver, D.E., Black, L.J., Falcone, J.F., MacLusky, N.J., 1987. Inhibition of central nervous system aromatase activity: a mechanism for fenarimol-induced infertility in the male rat. Toxicology and Applied Pharmacology 91, 235–245. HSE, 2003. Classification and labelling for adverse effects on fertility CMR and pesticides working groups. Document ECBI/56/02 Add. 1 Rev. 1. Available from: . Isobe, Y., Horimoto, M., Tachibana, M., Kawashima, K., 1998. Effects of 2,5-hexanedione on rat male fertility: 1. Studies on rat sperm motility. Teratology 57 (3), 29A–30A. Johannsen, F.R., Levinskas, G.J., Rusch, G.M., Terrill, J.B., Schroeder, R.E., 1988. Evaluation of the sub-chronic and reproductive effects of a series of chlorinated propanes in the rat. I. Toxicity of 1,2,3trichloropropane. Journal of Toxicology and Environmental Health 25 (3), 299–315. Kanojia, R.K., Junaid, M., Murthy, R.C., 1996. Chromium induced teratogenicity in female rat. Toxicology Letters 89 (3), 207–213. Kavlock, R., Cummings, A., 2005. Mode of action: inhibition of androgen receptor function—vinclozolin-induced malformations in reproductive development. Critical Reviews in Toxicology 35, 721–726.

257

Kawamura, S., Kato, T., Matsuo, M., Katsuda, Y., Yasuda, M., 1996. Species difference in protoporphyrin IX accumulation produced by an N-phenylimide hermicide in embryos between rats and rabbits. Toxicology and Applied Pharmacology 141, 520–525. Kodama, J.K., Guzman, R.J., Dunlap, M.K., Loquvam, G.S., Lima, R., Hine, C.H., 1961. Some effects of epoxy compounds on the blood. Archives of Environmental Health 2, 56–67. Koeter, H.B., Regnier, J.F., van Marwijk, M.W., 1992. Effect of oral administration of 1,3-diphenylguanidine on sperm morphology and male fertility in mice. Toxicology 71 (1–2), 173–179. Kojiama, H., Katsura, E., Takeuchi, S., Niiyama, K., Kobayashi, K., 2004. Screening for estrogen and androgen receptor activities in 200 pesticides by in-vitro reporter gene assays using Chinese hamster ovary cells. Environmental Health Perspectives 112 (5), 524–530. Latendresse, J.R., Warbrittion, A.R., Jonassen, H., Creasy, D.M., 2002. Fixation of testes and eyes using a modified davidson’s fluid: comparison with Bouin’s fluid and conventional Davidson’s fluid. Toxicologic Pathology 30 (4), 524–533. Lee, K.P., Kinney, L.A., Valentine, R., 1989. Comparitive testicular toxicity of bis(2-methoxyethyl) ether and 2-methoxyethanol in rats. Toxicology 59 (3), 239–258. Lee, L.-H., Hendel, J.J., 1998. Sertoli cell toxicants. In: Korach, K.S. (Ed.), Reproductive and Developmental Toxicology. Marcel Dekker Inc., New York. Li, W.H., Fan, Q.Y., Shen, L.J., Zhang, Y.H., Wu, X.Y., Ding, X.C., 2001. Study on testicular toxicity of 2-bromopropane, an environmental endocrine disrupter. Reproduction and Contraception 21 (3), 157– 160. Lim, C.H., Maeng, S.H., Lee, J.Y., Chung, Y.H., Kim, T.G., Park, J.H., Moon, Y.H., Yu, I.J., 1997. Effects of 2-bromopropane on the female reproductive function in Sprague–Dawley rats. Industrial Health 35 (2), 278–284. Maassaad, C., Entezami, F., Massade, L., Benahman, M., Olivennes, F., Barouki, R., Hamamah, S., 2002. How can chemical compounds alter human fertility?. European Journal of Obstetrics and Gynaecology and Reproductive Biology 100 127–137. MacGregor, J.I., Jordan, V.C., 1998. Basic guide to the mechanisms of antiestrogen action. Pharmacological Reviews 50 (2), 152–196. McIntyre, B.S., Barlow, N.J., Wallace, D.G., Maness, S.C., Gaido, K.W., Foster, P.M.D., 2000. Effects of in utero exposure to linuron on androgen-dependent reproductive development in the male Crl:CD(SD)BR rat. Toxicology and Applied Pharmacology 167, 87– 99. Meirow, D., Lewis, H., Nugent, D., Epstein, M., 1999. Subclinical depletion of primordial follicular reserve in mice treated with cyclophosphamide: clinical importance and proposed investigative tool. Human Reproduction 14 (7), 1903–1907. Meistrich, M.L., Wilson, G., Shuttlesworth, G.A., Porter, K.L., 2003. Cibromochloropropane inhibits spermatogonial development in rats. Reproductive Toxicology 17 (3), 263–271. Miki, K., Qu, W., Goulding, E.H., Willis, W.D., Bunch, D.O., Strader, L.F., Perreault, S.D., Eddy, E.M., O’Brien, D.A., 2004. Glyceraldrhyde 3-phosphate dehydrogenase-S, a sperm specific glycolytic enzyme, is required for sperm motility and male fertility. Proceedings of the National Academy of Sciences 101 (47), 16501–16506. Mitsumori, K., Kodama, Y., Uchida, O., Takada, K., Saito, M., Naito, K., Tanaka, S., Kurokawa, Y., Usami, M., Kawashima, K., 1994. Confirmation study, using nitrobenzene, of the combined repeat dose and reproductive/developmental toxicity test protocol proposed by the Organisation for Economic Cooperation and Development (OECD). Journal of Toxicological Sciences 19 (3), 141–149. Nagano, K., Nakayama, E., Oobayashi, H., Nishizawa, T., Okuda, H., Yamazaki, K., 1984. Experimental studies on toxicity of ethylene glycol alkyl ethers in Japan. Environmental Health Perspectives 57, 75–84. NTP, 1982. Carcinogenesis bioassay of Di(2-ethylhexyl)phthalate (CAS No. 117-81-7) in F344 rats and B6C3F1 mice (feed studies). NTP Technical Report #217.

258

M.P. Dent / Regulatory Toxicology and Pharmacology 48 (2007) 241–258

NTP, 1985. Di-N-pentylphthalate (CAS No. 131-18-0): Reproduction and fertility assessment in CD-1 mice when administered in feed. NTP report #RACB84048. NTP, 1986. Methoxyacetic acid (CAS No. 625-45-6): reproductive and fertility assessment in CD-1 mice when administered in the drinking water. NTP report #RACB84103. NTP, 1990a. Reproductive toxicology of 1,2,3-trichloropropane (CAS No. 96-18-4) administered by gavage in CD-1 Swiss mice. NTP report #RACB91039. NTP, 1990b. Toxicology and carcinogenesis studies of glycidol (CAS No. 556-52-5) in F344/N rats and B6C3F1 mice (gavage studies). NTP Technical Report #374. NTP, 1993. Toxicology and carcinogenesis of 1,2,3-trichloropropane (CAS No. 96-18-4) in F344/N rats and B6C3F1 mice (gavage studies). NTP Technical Report #384. NTP, 1995. Toxicity studies of dibutyl phthalate (CAS No. 84-74-2) administered in feed to F344/N rats and B6C3F1 mice. NTP Report #TOX 30. NTP, 2003. NTP-CERHR monograph on the potential human reproductive and developmental effects of 1-bromopane (1-BP). National Toxicology Program. Center for the Evaluation of Risks to Human Reproduction, vol. NIH 04-4479.

OECD guidelines for the testing of chemicals: OECD 407: repeated dose 28-day oral toxicity study in rodents, adopted by the Council on 27th July 1995. OECD 408: repeated dose 90-day oral toxicity study in rodents, adopted by the Council on 21st September 1998. OECD 414: pre-natal developmental toxicity study, adopted by the Council on 22nd January 2001. OECD 416: two-generation reproduction toxicity study, adopted by the Council on 22nd January 2001. OECD 421: reproduction/developmental toxicity screening test, adopted by the Council on 27th July 1995. OECD 422: combined repeat dose study with the reproduction/developmental toxicity screening test, adopted by the Council on 22 March 1996. Oliveira, C.A., Carnes, K., Franca, L.R., Hess, R.A., 2001. Infertility and testicular atrophy in the antiestrogen-treated adult male rat. Biology of Reproduction 65, 913–920. Omura, M., Romero, Y., Zhao, M., Inoue, N., 1997. Histopathological changes of the testis in rats caused by subcutaneous injection of 2bromopropane. Journal of Occupational Health 39 (3), 234–239. Piersma, A.H., Verhoef, A., Dortant, P.M., 1995. Evaluation of the OECD 421 reproductive toxicity screening test protocol using butyl benzyl phthalate. Toxicology 99 (3), 191–197.

Saxena, D.K., Lal, B., Murthy, R.C., Chandra, S.V., 1984. Lead induced histochemical changes in the testes of rats. Industrial Health 22 (4), 255–260. Sharpe, R.M., Skakkebaek, N.E., 1993. Are oestrogens involved in falling sperm counts and disorders of the male reproductive tract? Lancet 341, 1392–1395. Sharpe, R.M., 1998. Toxicity of spermatogenesis and its detection. In: Korach, K.S. (Ed.), Reproductive and Developmental Toxicology. Marcel Dekker Inc., New York. Shelby, M.D., Newbold, R.R., Tully, D.B., Chae, K., Davis, V.L., 1996. Assessing environmental chemicals for estrogenicity using a combination of in vitro and in vivo analysis. Environmental Health Perspective 104, 1296–1300. Silvers, K.J., Rorke, E.A., 1998. TCDD and uterine function. In: Korach, K.S. (Ed.), Reproductive and Developmental Toxicology. Marcel Dekker Inc., New York. Takahashi, K.L., Hojo, H., Aoyama, H., Teramoto, S., 2003. Acute testicular toxicity of 2,4-dinitrophenol and 2-sec-butyl-2,6-dinitrophenol in rats. Journal of Toxicological Sciences 28 (4), 336. Timbrell, J.A., 1991. Principles of Biochemical Toxicology, second ed. Taylor and Francis, London. Tsuruo, Y., 2005. Topography and function of androgen-metabolizing enzymes in the central nervous system. Anatomical Science International 80, 1–11. WHO Working Group, 1990. 2-Methoxyethanol, 2-ethoxyethanol, and their acetates. Environmental Health Criteria 115, 126. WHO Working Group, 1993a. Carbendazim. Environmental Health Criteria 149, 132. WHO Working Group, 1993b. Benomyl. Environmental Health Criteria 148, 135. White, R., Parker, M.G., 1998. Molecular mechanisms of steroid hormone action. Endocrine-Related Cancer 5, 1–14. White, R., Jobling, S., Hoare, S.A., Sumpter, J.P., Parker, M.G., 1994. Environmentally persistent alkylphenolic compounds are estrogenic. Endocrinology 135 (1), 175–182. Wong, C., Kelce, W.R., Sar, M., Wilson, E.M., 1995. Androgen receptor antagonist versus agonist activities of the fungicide vinclozolin relative to hydroxyflutamide. The Journal of Biological Chemistry 270 (34), 19998–20003. Yarbrough, J.D., Chambers, J.E., Grimley, J.M., Alley, E.G., Fang, M.M., Morrow, J.T., Ward, B.C., Conroy, J.D., 1981. Comparative study of 8-monohydromirex and mirex toxicity in male rats. Toxicology and Applied Pharmacology 58, 105–117. Zarn, J.A., Bruschweiler, B.J., Schlatter, J.R., 2003. Azole fungicides affect mammalian steriodogenesis by inhibiting sterol 14a-demethylase and aromatase. Environmental Health Perspectives 111 (3), 255–261.