Bis(2-ethylhexyl) phthalate

Specific investigations: other studies

Additional information

Proposed mechanisms of hepatocarcinogenicity:

This part has been cited in the EU Risk Assessment 2008 (p 403-408)

The mechanisms through which peroxisome proliferators (PPs) such as DEHP induce liver tumours in rodents have been extensively studied and discussed in the last years.

A brief overview of the current opinion on mechanisms and the significance for humans will be outlined here, for further details are referred to relevant reviews and criteria documents (Doull et al, 1999; Cattley et al., 1998; Youssef and Badr, 1998; Wolfgang et al., 1996; IARC, 1995; ATSDR, 1993; Bentley et al., 1993; ECETOC, 1992; WHO, 1992).

Negative results have been obtained in the majority of the genotoxicity studies on DEHP, MEHP and 2-EH. More conclusive positive results were obtained on cell transformation, induction of aneuploidy, and cell proliferation. These test systems are, however, also sensitive to several non-genotoxic substances such as tumour promoters and peroxisome proliferators.

Taken together all the results, both negative and positive, DEHP and its major metabolites are considered to be non-genotoxic substances. The results of tumour initiating and/or promoting studies indicate that DEHP have no tumour initiating activity, positive promoting activity in mice liver and a weak or no promoting activity in rat liver.

In the past, generally, three mechanisms have been proposed to account for liver carcinogenesis induced by DEHP and other PPs in rodents:

1) Induction of peroxisome proliferation leading to oxidative stress and generation of electrophilic free radicals, and

2) Increased hepatocyte proliferation/ suppression of hepatocellular apoptosis which could lead to fixation of a previously existing DNA-damage, enhancing the conversion rate of initiated cells to tumor cells, as well as increasing the susceptiblity of hepatocytes to replication and subsequent neoplastic transformation.

3) Recently a third mechanism through activation of peroxisome proliferator-activated receptors (PPARα) has been accepted by most of the experts in this field. The possible mechanisms of hepatocarcinogenicity and the species differences with respect to these hepatic effects of PP will be discussed in this section.

Peroxisome proliferation:

Peroxisomes are cytoplasmic organelles present in all cell types and contain a number of hydrogen peroxide generating oxidases, catalase (which catalyses the degradation of hydrogen peroxide), and a fatty acid ß-oxidation enzyme system. Peroxisome proliferation is characterised by increased peroxisome volume density (resulting from an increase primarily in the number of peroxisomes, although size may also be increased), changes in morphology, and induction of peroxisomal enzyme activities.

Other reported effects of PPs in hepatocytes of rats and mice include mitochondrial proliferation (with changes in enzyme activities), increase in the number of lysosomal bodies (with changes in enzyme activities and lipofuscin deposition), and induction of some microsomal enzyme activities. While marked effects have been observed in hepatocytes, only minor effects have been observed in certain other tissues. The term PPs covers substances, which in rodents induce peroxisome proliferation and liver enlargement (hepatomegaly), the latter being due to both hepatocyte hyperplasia (increased replicative DNA synthesis and cell division) and hypertrophy. Hepatomegaly and peroxisome proliferation is an early event during exposure to DEHP and has been observed in rats from about 14 days of exposure and throughout the exposure period at dose levels from 0.05% DEHP in the diet.

It has been suggested that liver tumour formation following prolonged administration of PPs arises from a sustained oxidative stress to rodent hepatocytes due to an imbalance in the production and degradation of hydrogen peroxide. The imbalance in the hydrogen peroxide production and degradation might be a result of that catalase is induced to a much lesser extent than peroxisomal ß-oxidation enzymes. The increased level of hydrogen peroxide in hepatocytes might, either directly or via other reactive oxygen species (e.g., hydroxyl radicals) cause DNA damage and subsequent neoplastic transformation, or apoptosis, e.g. by oxidative damage to intracellular membranes which then can lead to increased cell turnover thus increasing the probability of spontaneous tumour formation or to cell death.

However, there is evidence suggesting that the level of oxidative damage in vivo may be too low to account entirely for the carcinogenicity of PPs. Following prolonged administration (up to 79 weeks) of DEHP to rats (Tamura et al., 1990a,b), a 20-fold increase in peroxisomal b-oxidation activity was found after 2-4 weeks of treatment with a gradual decrease from week 20 to week 79 but remaining at an 8-10-fold higher level than control levels. Catalase activity increased (2-3-fold) after short-term treatment and remained at this level throughout the treatment period. The hepatic hydrogen peroxide level also increased but only 1.2-1.7-fold. As the hepatic hydrogen peroxide levels increased only slightly and did not correspond to the increase in peroxisomal b-oxidation activity, these results indicate that a large part of the hydrogen peroxide produced by peroxisomal ß-oxidation could be rapidly scavenged by catalase. This could be explained by the fact that the maximal hepatic catalase activity in vitro is thousands of times greater than the corresponding ß-oxidation activity in untreated animals. After treatment with peroxisome proliferator, the maximal activity of catalase is still thousands of times greater, even though catalase is only up-regulated 2- 3-fold and acyl-CoA oxidase increases as much as 20-fold. This indicates that liver tumours produced after long-term administration of DEHP might not be due only to the oxidative stress introduced by the enhanced peroxisomal ß-oxidation. This conclusion is strengthened by the fact, that even when oxidative stress was induced by the administration of buthionine sulfoximine at a dose that drastically lowered the endogenous glutathione pool in liver, the potent peroxisome proliferator, nafenopin, failed to induce unscheduled DNA synthesis, or increase the level of DNA single strand breaks in hepatocytes from animals treated in vivo (Nilsson et al., 1991). These authors also suggested, that the modest increase in the levels of the oxidation product 8-hydroxydeoxy-guanosine (8-OHdG), and that was found in liver DNA from animals treated with the potent peroxisome proliferators over long periods of time (Kasai et al., 1989), most probably represents an artifact from isolation of the DNA where no precautions had been taken to prevent e.g. OH-radicals to be formed through Fenton-like reactions (Nilsson et al., 1991).

In conclusion, induction of peroxisome proliferation alone is not enough to explain the hepatocarcinogenic activity of DEHP in rodents.

Exposure of Leydig cells to PPs prevented cholesterol transport into the mitochondria after hormonal stimulation and inhibited steroid synthesis, without altering total cell protein synthesis or mitochondrial and DNA integrity (Gazouli 2002).

PPs also reduced the levels of the cholesterol-binding protein peripheral-type benzodiazepine receptor (PBR) because of a direct transcriptional inhibition of PBR gene expression in MA-10 Leydig cells.

MA-10 cells contain mRNAs for PPARalpha and PPARbeta/delta, but not for PPARgamma.

In vivo treatment of mice with PPs resulted in the reduction of both testis PBR mRNA and circulating testosterone levels, in agreement with the proposed role of PBR in steroidogenesis.

By contrast, liver PBR mRNA levels were increased, in agreement with the proposed role of PBR in cell growth/tumor formation in non steroidogenic tissues. However, PPs did not inhibit testosterone production and testis PBR expression in PPARalpha-null mice. These results suggest that the antiandrogenic effect of PPs is mediated by a PPARalpha- dependent inhibition of Leydig cell PBR gene expression.

Hepatocyte proliferation and apoptosis:

Hepatocyte proliferation (characterised by increased replicative DNA synthesis and cell division, and hypertrophy) is an important response in rodent liver to PPs and has been implicated in the mechanisms of rodent hepatocarcinogenesis. Hepatocyte proliferation occurs in the tumours that develop in rats and mice after administration of PPs and is seen in lesions that are the direct progenitors of tumours Thus, at least for the more potent PPs, a close correlation has been found between induction of sustained replicative DNA synthesis and the potency of various PP with respect to induction of liver tumors (Marsman et al., 1988, 1992). However, the association of cell proliferation with tumour formation in rodent liver is complex and the magnitude of response for hepatomegaly and hepatocyte proliferation is not entirely predictive of eventual tumour yield. For PPs, it is important to differentiate short-term from prolonged stimulation of cell replication. Hepatomegaly is evident during the first few days of administration of PPs and is largely due to transient hepatocyte proliferation that subsides after several days as liver weight reaches a new plateau. Hepatomegaly has also been seen in rodent liver after prolonged administration of some, but not all, PPs. There are some indications that suppression of hepatocellular apoptosis also occurs during the induction of hepatomegaly. The increase in liver weight is dependent on continued exposure and is reversible upon cessation of exposure and it has been suggested that this reversal could be related to large increases in hepatocellular apoptosis.

An increased rate of cell proliferation can be a critical effect both in tumour initiation, by increasing the frequency of spontaneous mutations and the rate of conversion of DNA adducts into mutations before they are repaired, and in tumour promotion by facilitating the promotion of initiated cells. This effect could be strengthened by suppression of hepatocellular apoptosis, which lead to increasing the number of mutant hepatocytes susceptible to replication and subsequent neoplastic transformation. It has been reported recently (James et al., 1998), that DEHP and its metabolite MEHP can induce DNA synthesis and inhibit hepatocyte apoptosis in rats and mice hepatocytes in both in vivo and in vitro studies. Also a tumour-promoting activity of DEHP was observed in the liver of mice whereas a promoting activity in the liver of rats is equivocal. As first suggested by Préat et al. (1986a, 1986b), in addition to cell proliferation PPs appear to have an important role in promoting the selective growth of basophilic preneoplastic foci. Thus, whereas phenobarbital causes an increase in the number of preneoplastic foci in a liver initiated by e.g. diethylnitrosamine, a potent PP like nafenopin or WY-14,643 does not increase the number of such foci appreciably, but causes a great increase in the size of these foci. Further, the basophilic foci induced by PPs seem to have a much higher likelihood to progress to hepatocellular carcinomas by boosting the selection of transformed cells (Préat et al., 1986a, 1986b; Cattley and Popp, 1989). Development of foci and adenomas depends on the continuous exposure to PPs, where cessation of exposure results in disappearance of benign lesions, and continued exposure is essential for progression to malignant tumors (Marsman and Popp, 1994; Miller and Cattley, 1996).

In conclusion, induction of hepatocyte proliferation combined with suppression of hepatocellular apoptosis could play a major role in the hepatocarcinogenicity of DEHP.

Activation of PPARα:

Recent investigations have demonstrated the central role of a class of nuclear receptors, the peroxisome proliferator-activated receptors (PPARs), in mediating the effects of PPs (reviewed by Cattley et al., 1998). In the presence of PPs or fatty acids, the PPAR receptors induce the transcriptional regulation of PP-responsive genes.

Out of four different subtypes, the subtype PPARα that is strongly expressed in tissues catabolizing fatty acids (liver, digestive mucosa, kidney proximal tubules, muscle and retina) seems to be the most important with respect to processes associated with peroxisome proliferation. The expression of PPARα may be, to some extent, affected by glucocorticoids (Lemberger et al., 1994, 1996), providing an explanation of the previously observed effects on peroxisome proliferation induced by fasting and stress caused e.g. by hypothermia (de Duve, 1983; Reddy and Lalwani, 1983). Gene transcription is affected through a heterodimeric receptor complex involving PPARα and the retinoid X receptor (RXR) that is activated by PPs and 9-cis-retinoic acid (present endogenously). The activated receptor complex regulates transcription via binding to the promoter regions of peroxisome responsive genes, e.g. those involved in the ß-oxidation of fatty acids. However, the fact that potent PPs will act as hypolipidemic drugs in man without causing detectable peroxisome proliferation seems to indicate, that other genes are also affected by PPARα , e.g. the promoter regions of the human apolipoprotein genes (apo A-I, apo A-II, apo C-III). The fact that long-term administration of DEHP in the feed down to a concentration of 200 ppm increases the dolichol content of lysosomal membranes (Edlund et al., 1986) also bears evidence of the multifaceted effects induced by PPs.

When DEHP (1.2% in the diet) was administered to Sv/129 mice entirely lacking the PPARα , none of the responses typical for peroxisome induction (such as increase in the number of peroxisomes, induction of replicative DNA-synthesis, and hepatomegaly) found in the wild type mouse could be detected (Ward et al., 1998). Further, whereas the wild type mouse fed DEHP exhibited typical lesions in liver, kidney and testis, no signs of liver toxicity was detected in the "knockout mouse". On the other hand, evidence of lesions in kidneys and testes were also found in the latter, indicating PPARα independent pathways for induction of toxicity in these organs. In another study with PPARα null mice fed the potent peroxisome proliferator, Wy-14,643 at 0.1% in the diet for 11 months, no indication of replicative DNA synthesis in the liver, and no increase in the incidence of liver tumors were observed. In contrast, after 11 months administration, 100% of the wildtype mice exhibited multiple hepatocellular neoplasms (Peters et al., 1997). Although it is established that PPARα could mediate liver cell proliferation and hepatocarcinogenesis of the studied PPs, the pathways involved have not been elucidated.

A report from 2003 (Klaunig J.E. et al, 2003) provides an in-depth analysis of the state of the science on several topics critical to evaluating the relationship between the mode of action (MOA) for PPARα agonist, such as DEHP, and the human relevance of related animal tumours. Topics include a review of existing tumor bioassay data, data from animal and human sources relating to the MOA for PPARα agonists in several different tissues, and case studies on the potential human relevance of the animal MOA data. The summary of existing bioassay data discloses substantial species differences in response to peroxisome proliferators in vivo, with rodents more responsive than primates. It is clear from the published studies that humans possess a functional PPARα, and that peroxisome proliferators activate the human receptor. It is clear that some of the genes modulated by these chemicals in humans differ from those regulated by rodent PPARα. The available epidemiological and clinical studies are inconclusive, but, nonetheless, do not provide evidence that peroxisome proliferators, such as DEHP, cause liver cancer in humans.

In conclusion, the available, studies on transgenic mice, indicate that PPARs may play a central role in mediating the hepatotoxic effects of DEHP, such as increase in the number of peroxisomes, induction of replicative DNA-synthesis, and hepatomegaly. Also, it has been demonstrated that PPARs is required in mediating the hepatocarcinogenic effects of the PPs Wy-14,643 in mice.

Species differences

Marked species differences with respect to hepatic response to PPs are apparent, were rats and mice seem to exhibit the highest sensitivity. Guinea pigs and monkeys are relatively insensitive, while Syrian hamsters have demonstrated a sensitivity intermediate between these two groups of mammals. In a comparative study (14 days) of rats and hamsters, the liver weight of hamsters was significantly increased only at 1000 mg/kg b.w. per day with no significant increases in peroxisomal enzyme activities, whereas a significant and dose-dependent (from 100 mg/kg b.w. per day) increase in rat liver weight as well as in peroxisomal enzyme activities was observed (Lake et al., 1984). In marmosets, the liver weight was not affected and microscopic examination revealed only a slight increase in peroxisomes following administration of 2000 mg/kg b.w. per day for 14 days. In rats, hepatomegaly, marked peroxisome proliferation, and increased peroxisomal enzyme activities was observed following a similar dosage regimen (Rhodes et al., 1986). In another comparative study (21 days) with rats and cynomolgus monkeys, no treatment related changes in liver weight and peroxisomal enzyme activities were observed in monkeys at dose levels up to 500 mg/kg b.w. per day, whereas in rats marked effects on the same parameters were seen at a similar dosage regimen (Short et al., 1987).

In vitro studies using primary hepatocyte cultures from rodents and primates have supported the in vivo findings. Thus, whereas a number of different PPs have caused peroxisome proliferation in rat and mouse hepatocytes, several investigations have demonstrated a lack of activity in primate and human cells (reviewed in: Ashby et al., 1994; IARC, 1995; Elcombe et al., 1997). MEHP, a metabolite of DEHP, did not stimulate peroxisome proliferation in human cells, although a marked response was obtained in rat hepatocytes (Butterworth et al., 1989).

The potential human response to PPs has been examined in liver biopsies obtained from patients treated with hypolipidemic drugs. Morphometric measurements in liver biopsies did not reveal evidence for peroxisome proliferation. The potential carcinogenic risk of hypolipidemic therapy with fibrates (clofibrate and gemfibrozil, both being potent PPs) has been evaluated in two limited clinical trials with no evidence for carcinogenesis obtained (IARC, 1996). No relevant data are available on humans exposed to DEHP. But in a study where liver biopsies had been obtained from dialysis patients, who are exposed to significant quantities of DEHP leached from PVC dialysis tubings, Ganning et al. (1987) reported an increased number of peroxisomes in exposed individuals. However, this claim was based on two electron micrographs from two different patients, where an apparent increase in the number of peroxisomes was found in one specimen.

In rodent liver, hepatomegaly and peroxisome proliferation require expression of functional PPARα (Lee et al, 1995). The slight or no responsiveness of human liver to some effects of PPs, such as hepatomegaly and peroxisome proliferation, could be explained by a low level of PPARα found in human livers (1-10% of the level found in rat and mouse liver), as well as observations of genetic variations that render the human PPARα receptor less active as compared to PPARα expressed in rodent liver (Tugwood et al., 1996; Palmer et al., 1998; Woodyatt et al., 1999).

In conclusion, the available data indicate a quantitative species differences in the response to the hepatic effects of DEHP and in the activation of PPARα.

Effects of phthalate exposure on perinatal testosterone production

From Scott 2009 (Steroidogenesis in the Fetal Testis and Its Susceptibility to Disruption by Exogenous Compounds. Endocrine Reviews, December 2009, 30(7):883–925).

The effects of phthalate exposure on perinatal testosterone production (assessed as AGD) in humans has been investigated in several studies, the results of which are conflicting. The first cross-sectional study (in the United States) examined 85 boys at 2–36 months of age and found a negative correlation between AGD (corrected for body weight) and the level of certain phthalate metabolites, including monobutyl phthalate (MBP) and monoethyl phthalate (MEP), found in maternal urine during pregnancy (Swan 2005). A recent expansion of this study to include a total of 106 boys has confirmed the negative correlation between AGD and maternal (urinary) levels of phthalates, including MEP, MBP, monoethyl hexyl phthalate (MEHP) and the further MEHP metabolites, monoethyl hydroxyhexyl phthalate and monoethyl oxohexyl phthalate (Swan 2008). Both of these studies also demonstrated that AGD correlated to penile volume/length (Swan 2005, Swan 2008) and the incidence of cryptorchidism (Swan 2005), similar to rat studies (Welsh 2008).

Another study of 73 pregnant Mexican women in a hospital-based cohort investigated the association between exposure to MEHP, monobenzyl phthalate, MEP, and MBP during pregnancy and AGD in male newborns (Bustamente-Montes 2008). This study found a statistically significant association between MEP exposure and reduced AGD and also between monobenzyl phthalate exposure and reduced penis length and width. These studies are consistent with the possibility that perinatal phthalate exposure can inhibit testosterone production in the male fetus (during the masculinization programming window), resulting in reduced AGD and penile volume/length, as well as inhibiting normal testis descent; such effects are in broad, but not total, agreement with studies in rats detailed below. However, one point of disagreement is with regard to diethyl phthalate and its metabolite MEP.

In contrast to the above-mentioned studies, a smaller (prospective) study in Taiwan, involving 33 boys, found no relationship between levels of MBP or MEHP, measured in AF or maternal urine, during pregnancy, and AGD of the male offspring (Huang 2009).

This finding fits with two in vitro studies that have investigated the effects of MBP or MEHP on human fetal testis testosterone production. In the first study, second-trimester human fetal testis explants were cultured with MBP in short-term culture, but there was no effect on basal or hCG-stimulated testosterone production (Hallmark 2007). In the second study, exposure of first-trimester human fetal testis explants to MEHP in the presence or absence of LH/hCG found no effect on testosterone production or on steroidogenic enzyme expression (Lambrot 2009).

The latter study did, however, demonstrate that MEHP has negative effects on germ cells, as it did in fetal rat testis explants, which makes the lack of effect on testosterone production more convincing.

An increase in SHBG would result in less free testosterone because more would be able to bind to SHBG, and this would result in an increase in LH because of reduced negative feedback, thus explaining the increased LH-free testosterone ratio observed without the need to invoke direct inhibition of steroidogenesis by the phthalates. Phthalates are known to act on the liver, which is also the source of SHBG, but it is unknown whether phthalate exposure affects SHBG production.

A study of men working in a PVC factory, and thus occupationally exposed to phthalates, showed a modest and significant reduction in serum free testosterone in workers with high levels of urinary MBP and MEHP, compared with unexposed workers (Pan 2006). However, a cross-sectional study of 295 men attending an andrology clinic in Massachusetts found no association between phthalate levels in urine and serum levels of testosterone (Duty 2005), although in this instance phthalate exposure would have been notably lower than for the PVC workers.

Moreover, if the rat is any guide, then adult human Leydig cells may be relatively insensitive to the effect of phthalates. Thus, treatment of prepubertal rats with 200 mg/kg/d diethyl hexyl phthalate (DEHP) from postnatal d 21–35 caused a 50% reduction in serum testosterone levels (Akingbemi 2001), whereas the same or higher doses administered to adult rats had little or no effect (Akingbemi 2001, Agarwal 1986). Although one study has shown a massive inhibitory effect (>90%) of MEHP on LH-stimulated testosterone production by adult rat primary Leydig cells over 2 h of culture, this was found only after exposure to 1 mM MEHP, and no effect was found with a 10-fold lower dose (Jones 1993).

In contrast to the data in humans, there is unequivocal evidence that certain phthalates can profoundly inhibit testosterone production by the fetal rat testis. Thus, administration of phthalates to pregnant rats during the last week or so of gestation results in reduced AGD and reproductive malformations, including hypospadias, in the Long Evans (Gray 1999), Sprague-Dawley (Carruthers 2005) and Wistar (Ema 1998) strains of rat. These changes are consistent with reduced fetal androgen exposure, and this has been demonstrated directly after in utero exposure to a number of phthalates, including DEHP (Wilson 2004, Gray 2000, Parks 2000), DBP (Wilson 2004, Fisher 2003, Mylchreest 2002, Schultz 2001), butyl benzyl phthalate (Wilson 2004), diisobutyl phthalate (Boberg 2008), diisononyl phthalate (Gray 2000), and diisoheptyl phthalate (McKee 2006).

Several of these compounds have been shown to cause dose-dependent suppression of fetal testicular testosterone production, and for the most potent (DEHP and DBP), this occurs at doses above 100–250 mg/kg/d.

Other phthalates, such as diethyl phthalate, dimethyl phthalate, dioctyl phthalate, and diisodecyl phthalate, do not affect fetal rat testicular testosterone production or AGD (Gray 2000, Howdeshell 2008).

Molecular analyses in fetal rat testes after in utero exposure to phthalates has shed light on the potential mechanisms via which phthalates suppress testicular testosterone production. Several of the key genes involved in steroidogenesis are down-regulated after in utero exposure to DBP or MEHP. These genes are StAR, HMG-CoA synthase, and SRB1 (all involved in cholesterol uptake/transport), and the steroidogenic enzymes Cyp11a, 3beta-Hsd, and Cyp17 (Schultz 2001, Lahousse 2006, Lehmann 2004, Plummer 2007). Suppression of these various enzymes provides a convincing explanation for the phthalate-induced reduction in fetal testicular testosterone production.

Based on present evidence, it appears that all of the phthalates that affect testosterone production by the fetal testis do so by similar mechanisms, although the dose-response characteristics may differ; in this regard, DBP and DEHP are the most potent and are approximately equipotent, based on the above-cited studies.

In contrast to the consistent effects of DBP and DEHP on fetal testicular testosterone production in the rat, data for exposure of fetal mice to DBP or DEHP have produced equivocal results. A detailed study showed that administration of single or multiple doses of DBP (up to 1500 mg/kg/d) or MEHP(up to 1000 mg/kg/d) to pregnant mice did not reduce testicular testosterone levels or affect the expression of the steroidogenic enzyme genes, as seen in the rat (Gaido 2007); this was shown in two strains of mice (C57Bl6, G3H/HeJ).

In contrast, a recent study in C57Bl6 mice treated with 100, 200, or 500 mg/kg/d DEHP from e12–e17 reported that this dose-dependently induced hypospadias on e19, with males from the top dose group exhibiting a 75.7% incidence of hypospadias and a 13% reduction in AGD (Liu 2008). In addition to conflicting with the study by Gaido et al. (Gaido 2007), this study also conflicts with data for the rat exposed to similar levels of DBP or DEHPbecause effects on AGD (measured postnatally) are of similar magnitude to those reported for the mice, but the rates of hypospadias reported are considerably less in the rats, ranging from 12–37% (Gray 2000, Fisher 2003, Mylchreest 2000), an observation that is explained by the relatively poor suppression of testosterone levels by DBP in rats during the masculinization programming window (Scott 2008).

No other toxicological studies involving fetal exposure of mice to phthalates that cause fetal testis effects in rats have reported hypospadias (Heindel 1989, Shiota 1982), although it is uncertain whether or not this was specifically sought. Another study has reported that DEHP can reduce insulin-like factor 3 mRNA expression by fetal mouse Leydig cells in vivo and in vitro, but effects on steroidogenesis were not studied (Song 2008). The conflict over fetal testis effects of phthalates in mice is not clarified by studies on isolated Leydig cells from postnatal mice because positive effects on steroidogenesis have been reported (Gunnarsson 2008), whereas negative effects of MEHP have been reported on MA-10 tumor Leydig cells (Dees 2001). However, a very recent study (Lehraiki 2009) perhaps reconciles these disparate findings for phthalate effects in mice. It shows that whether MEHP has inhibitory or stimulatory effects on steroidogenesis in fetal mouse testes cultured over 1–3 d depends on fetal age, culture duration, and the presence/ absence of LH. Notably, no inhibition of basal testosterone production was observed at any age, but in particular at e13.5 (corresponding to part of the masculinization programming window), and this is in marked contrast to the studies in rats.

Johnson et al. (2012) concluded that recent studies on fetal testis xenografts suggest that the human fetal testis responds like the mouse and is refractory to phthalate induced inhibitionof testosterone production. This conclusion is based on experimental data with xenocrafted fetal testis of humans, rats and mice reported by Mitchell et al. (2012) and Heger et al. (2012). .