Ziracin-Induced Congenital Urogenital Malformations in Female Rats

Abstract

Spontaneous hypospadias is seldom observed in rats in contrast to its occurrence in 1 out of 250 human births. Ziracin, an antibacterial of the everninomycin class under development for serious enterococcal, staphylococcal, and streptococcal infections, caused anomalies of the external genitalia in F1 female rats and decreased reproductive performance. To characterize the urogenital malformations and determine the period of sensitivity to the effects of Ziracin during development, pregnant rats (F0) were administered 60 mg/kg IV of Ziracin from GD6 to LD21, GD6 to 13, GD14 to the last day of gestation or LD0 to 21. Controls received saline or placebo from GD6 to LD21. Ziracin-induced changes occurred in F1 rats exposed from GD6 to LD21 and GD14 to the last day of gestation, indicating that the period of sensitivity to Ziracin was from GD 14 to the last day of gestation. The urogenital abnormalities consisted of cranial displacement of the urethral opening within the vagina from its normal location at the tip of the genital tubercle. When the urethrovaginal junction occurred at the distal third of the vagina, it created an urogenital cloaca. As a result, ascending infections were seen in the urinary and genital tract. No differences in survivability, body weight, and date of vaginal opening were observed in F1 females. The estrous cycles were slightly prolonged. The mating and fertility indices were decreased as a result of the urogenital anomalies. The mammary glands of pregnant F1 females were underdeveloped, thus F2 pups from affected F1 females had a decreased survival rate. Although the cause of these effects is not known, the findings are consistent with a potential hormonal mechanism.

Introduction

Ziracin, a fermentation product of Micromonospora carbonacea, is an oligosaccharide Gram-positive antibacterial compound of the everninomicin class, under development for serious infections caused by glycopeptide-resistant enterococci (VRE), staphylococci (S. epidermidis, a.k.a. GISE, and S. aureus (GISA)), and S. pneumoniae (Ganguly, 2000; Jones et al., 2001; Chu et al., 2002; Zhong et al., 2002). While the specific site of binding has not yet been determined, studies have shown that Ziracin inhibits protein synthesis in bacteria by interacting with the large ribosomal subunit (Belova et al., 2001; Chu et al., 2002). The pharmacokinetics of Ziracin has been previously described (Lin et al., 2000). Ziracin is neither mutagenic in a bacterial/mammalian microsome mutagenicity assay nor clastogenic in a human peripheral blood cell chromosome aberration assay with and without metabolic activation (unpublished company reports). Intravenous doses of 30 mg/kg and 60 mg/kg were well tolerated in studies of 3-month duration in monkeys and rats, respectively. No Ziracin-related effects on fertility or on pregnancy and fetal development (unpublished company reports) were observed in standard reproductive toxicity studies (ICH guidelines, 1994) conducted in rabbits and rats at doses up to 40 and 45 mg/kg, respectively, in which term fetuses were examined after exposure of the dam during GD6-17 (rat) or GD7-19 (rabbit). However, in peri- and postnatal reproductive toxicity studies in rats (Segment III), findings were noted in the offspring of F0 female rats dosed with 15, 30, and 60 mg/kg of Ziracin.

In the Segment III study during which pregnant rats were administered Ziracin intravenously at doses up to 60 mg/kg from GD7 to postnatal day 20, the mating and fertility indices of the F1 females were reduced in a dose-related fashion. In addition, these F1 females exhibited anomalies of the external genitalia characterized by malformations of the vaginal opening and a shortening of the urethrovaginal distance. Urinary tract infection with or without obstruction by calculi caused the death of 4 F1 females across all treated groups and was attributed to the vaginal and urethral anomalies. There were no effects on male fertility or organ weights.

The present study was conducted to further characterize the Ziracin-induced changes in the urogenital tract of F1 females born to dams exposed during gestation. The first objective was to characterize the alterations in F1 reproductive capacity and the morphogenesis of the genitalia malformations. The second was to evaluate the potential contribution of the Placebo formulation to these developmental effects, as only a saline control group was used in the original Segment III study. The final aim was to determine the period of sensitivity to the developmental effects of Ziracin during specific intervals of gestation and/or lactation.

Materials and methods

Animals and Husbandry

Sprague–Dawley rats (Crl:CD[SD]IGS BR VAF/Plus) from Charles River Laboratories, Kingston, NY, were mated at the supplier. The day that evidence of mating was detected was designated as gestation day 0 (GD0) for that female. The day of delivery was designated lactation day 0 (LD0). If delivery was not completed on the day of onset, the following day or the day of completion was designated as lactation day 1 (LD1).1 F0 females were housed individually in suspended, stainless-steel, wire-mesh cages until transferred to solid-sided, stainless-steel maternity cages containing nesting material (CareFresh, Canada) when the first set of females reached GD18. At weaning or shortly thereafter, the offspring (F1) were housed individually in wire-mesh cages except during cohabitation when they were housed 2 per cage. After mating, the F1 rats were housed individually in the same manner as the F0 females. PMI Nutrition International, Inc. Certified Rodent LabDiet 5002 and tap water were provided ad libitum. There were no significant variations in the environmental conditions (72 ± 4°F, 40–70% relative humidity and a 12-hour light/dark cycle) of the animal room. The use of animals in this study followed the guidelines provided in the NRC Guide for the Care and Use of Laboratory Animals (1996) and the Animal Welfare Act (Code of Federal Regulations, Title 9), and the protocol was approved by the Institutional Animal Care and Use Committee.

Test and Control Articles

Ziracin Sterile Powder for Injection, U.S.P [Formula 3146] contained Ziracin:Meglumine (NMG):Hydroxypropyl β Cyclodextrin (HPβCD) at a ratio of 1:3:5 with Polysorbate 80 and mannitol and a concentration of 20 mg/mL. The Placebo [Formula 3149] contained NMG: HPβCD at a ratio of 3:5 with Polysorbate 80 and mannitol. Sterile 0.9% Sodium Chloride for Injection, U.S.P (Abbott, Abbott Park, Illinois) was used as the saline control. Dosing solutions were prepared daily by reconstitution with sterile water for injection and administered intravenously the same day at the dose of 60 mg/kg, which was deemed the maximal dose because it was associated with decreased pup survival and fertility in the original Segment III study. Analytical results indicated that the average dosing concentrations used in this study were within specifications (±10%).

Study Design

The rats were randomized into 6 groups (n = 10 rats/ group). All Ziracin-treated groups received 60 mg/kg/day during specific intervals of gestation and/or lactation in order to determine the window of sensitivity to the developmental effects observed in the original segment III study. Group 3 treatment covered the period from gestation day 6 (GD6) to lactation day 21 (LD21), group 4 treatment was limited to GD6 to GD13, group 5 to GD14 to the last day of gestation, and group 6 covered the lactation period from LD0 to 21. Rats were dosed with 0.9% sodium chloride on the days they were not administered Ziracin for the period of GD6 to LD21. Control groups received 0.9% sodium chloride (group 1) and Placebo [Formula 3149] (group 2), respectively, from GD6 to LD 21.

F₀ Measurements

Females were examined daily for changes in appearance and behavior from GD0 until LD21. Body weight was measured on gestation days 0, 6, 9, 12, 14, 18, and 21, and on lactation days 0 (when possible), 1, 4, 7, 14, and 21. Food consumption was measured for the intervals of gestation days 0 to 6, 6 to 12, 12 to 14, and 14 to 21, and lactation days 1 to 4, 4 to 7, 7 to 10, and 10 to 14. F₀ females were sacrificed by CO₂ inhalation on LD21 or within 1 week thereafter. A female that did not produce a litter was sacrificed on GD25. All F₀ dams were examined for visceral changes and implantation sites were counted.

F₁ Antemortem Measurements

The total number of live and dead pups were recorded daily until LD21. F₁ pups were culled to 10 per litter on LD4 to allow uniform pup growth and female pups were preferentially retained. The pup survival rate was calculated for lactation day(s) 0, 1–4, and 5–21. From LD0 until 21 pups were observed daily. They were sexed, weighed on LD0, 1, 4, 7, 14, and 21, and identified individually with a paw tattoo prior to weighing and anogenital distance measurements. The presence of nipples in female pups was assessed on LD1. The anogenital distance of F₁ females was measured on LD1, 7, 14, and 21 and on postnatal days 28, 35, 42, and 49 (Clark, 1999). The distance from the urethral to the vaginal orifice was measured on postnatal days 42, 49, and at study termination.

After measurement of the anogenital distance on LD21 3 females per litter were randomly selected and retained for postweaning measurements and breeding. Postweaning body weights were measured at approximately weekly intervals prior to and during cohabitation or until sacrifice for females with no positive sign of mating. F₁ females were checked daily for vaginal opening from postnatal day 25 until all females tested positive or until postnatal day 42, whichever came first. They were weighed on the day they tested positive for vaginal opening.

Estrous cycles of F₁ females were evaluated by vaginal cytology for 3 weeks prior to mating (starting at approximately 9 weeks of age) and during the mating period until positive evidence of mating was detected. The results obtained during the premating period were used to analyze estrous cycles. Mean estrous frequency during the 3-week period and the number of litters producing at least 1 female having irregular cycles (defined as shorter than 4 days or longer than 6 days and/or including 3 or more consecutive days of estrous) were determined.

At approximately 12 weeks of age, the F₁ females were placed with a nonstudy male on a 1-to-1 ratio until mating. If a pair of rats failed to mate after 2 weeks, the female was placed in cohabitation with a proven male for up to 1 week.

F₁ Postmortem Measurements

One female/litter was selected from five litters in each group for interim sacrifice on LD1, 4, 21, and at 7 weeks of age (postnatal days 49/50). At study termination, all females were sacrificed within 1 week after completion of delivery and after F2 pups were sacrificed on LD4, that is between LD4 and 7, with the last LD7 being on study day 128. The urethro-vaginal distance was measured with a calibrated caliper.

All females were examined for visceral changes, implantation sites were counted and the kidneys, adrenal glands, mammary glands, pituitary gland, ovaries, oviducts, uterus (plus cervix), vagina, vulva, and gross findings were collected in 10% neutral buffered formalin and processed to hematoxylin-stained slides for light microscopic examination. Kidneys, ovaries, and adrenal glands were weighed prior to fixation. At tissue trimming, the vagina was sectioned sagittally from the cervix to the vulva to include the urethra, urethral opening, and urinary bladder left in its normal anatomical position. Serial sections of the vagina were performed and examined.

F₂ Measurements

Litter size (number born alive and dead), and the sex, body weight, and observations of individual offspring were recorded on LD0 (when possible) and 1 and the length of gestation was calculated. Thereafter, litters were examined daily for number of pups until LD4. Stillborn and dead pups were sexed and examined for the presence of milk in the stomach and external abnormalities. Survival was calculated until LD4 when all surviving F₂ pups were sacrificed by CO₂ inhalation.

Data Analysis

A 1-way analysis of variance was used for continuous data. When the probability of the overall “F” statistic for any set of data was ≤0.05, the dose groups for that set of data were compared with the control group via Dunnett’s “t” procedure, in which α levels of 0.01 and/or 0.05 were used for comparison. For noncontinuous data a Kruskal-Wallis test, a nonparametric analysis of variance, was used to test for statistical significance. When the probability of the overall “H” statistic for any set of data was ≤0.05, the dose groups for that set of data were compared with the control group via the Kruskal-Wallis (2 × 2) procedure, in which α levels of 0.01 and/or 0.05 were used for comparison. Categorical data were compared via the Chi-square test. When the probability of the overall statistic for any set of data was 0.05, the dose group for that set of data was compared to the control group via Fisher’s exact test, in which α levels of 0.01 and/or 0.05 were used for comparison.

Ziracin was administered to the F₀ (maternal) generation and therefore the F₀ litter should serve as the unit of analysis for all F₁ and F₂ data (Clark, 1999). The F₀ litter was used as the unit of analysis for the following parameters for the F1 generation: lactation body weights, anogenital distance, nipple retention, estrous cycles, and vaginal opening. The litter could not be used as the unit of analysis for postnatal body weights, urethro-vaginal distance, F₁ litter and delivery data, and F₂ data due to the nonstandard study design and computer limitations. However, because the number of litters per group was similar, the differences between calculations with and without the litter as the unit of analysis are negligible and have no effect on the interpretation of the study results.

Results

F₀ Generation

Maternal (F₀) administration of 60 mg/kg Ziracin or placebo during gestation and lactation was well tolerated. There were no effects on the dams or offspring (F₁) survival during lactation.

F₁ Generation

There were no effects in the F₁ generation attributed to the placebo. In the F₁ generation, the Ziracin-related effects were limited to the females in groups 3 and 5, in which the dams received Ziracin from GD6 to LD21 (group 3) and from GD14 to the last day of gestation (group 5). There were no differences in the viability of pups at birth or during lactation. There were no other differences in body weight among the groups. All the F1 adult body weights during gestation and lactation were comparable to the saline control group. The mean anogenital distance was increased in groups 3 and 5 to 3.93 mm and 3.62 mm versus 3.08 mm in the saline control at postnatal day 7 and to 18.85 mm and 18.81 mm versus 16.56 mm in the saline control at postnatal day 49 (Table 1). The urethro-vaginal distance was significantly reduced in these groups to 0.00 mm and 0.17 mm compared to 5.48 mm in the saline control at day 42 and to 0.00 mm in both groups compared to 4.94 mm in the saline control at termination (Table 2).

This drastic diminution of the urethro-vaginal distance in groups 3 and 5 reflects the Ziracin-induced urogenital malformation involving the external genitalia (Table 3 and Figures 1–5). The urethra and vagina are normally completely separated and the urethra opens at the tip of the genital tubercle, which is ventral to the vagina (Figures 1a and 1b). The urethra of most F₁ females born to dams exposed to Ziracin during the period of gestation, which included GD14 through the last day of gestation (group 5), opened as a slit displaced dorsally towards the vaginal opening anywhere from along the shaft, to the base of the genital tubercle, to some distance within the vagina (Figures 2a and 2b). The junction apparently occurred more frequently at 1 of 2 levels: urethrovulvar at the level of the clitoris in mild cases or urethrovaginal at approximately one-third within the vagina in moderate cases (Figure 3).

The more severe urethrovaginal fusion thus created an urogenital cloaca in the distal third of the vagina. Alongside this anomaly, the preputium of the clitoris and associated integument were reduced in size. In minimal cases, the tip of the genital tubercle was blunted and the fossa clitoridis, ducts of clitoral glands, and urethra connected obliquely at the base of the clitoris. In mild to moderate cases, the preputium of the clitoris and associated integument was much reduced in size and the opening of the vulva was enlarged as a result. This malformation became evident by external examination in all F₁ females during the period of postnatal days 28 to 49 and was detectable by microscopic examination in some females on LD4, the earliest time point examined (Figures 4a and 4b). It was minimal in young females (LD4 and LD21) and became gradually more severe as the rats matured so that at the end of the study, the preputium of the clitoris and associated integument were greatly reduced in size in most F1 females.

The resulting distal cloaca predisposed the rats to ascending urinary tract infections and/or urinary incontinence in a limited number of females. Four females from group 3 and 1 from group 5 were either found dead or sacrificed in moribund condition on days 89, 119, 106, 87, and 93, respectively. At necropsy these females had white granular material in the inguinal area, urine-stained fur, distended bladder, dilated renal pelvis and ureter, multifocal to coalescing tan areas in the renal cortex and/or adhesions in the abdominal cavity. At the end of the study, a host of inflammatory changes, often chronic, had developed in F1 females. These included mild to severe vaginitis, abscesses, which seemed centered on the clitoral glands or ducts, and endometritis. Inflammation in the urinary tract consisted of urethritis, cystitis with urothelial ulceration, hyperplasia and squamous metaplasia, and pyelonephritis (Figure 5). Chronic irritation of urinary incrustations led to a severe necrotic dermatitis over the inguinal area of 1 female. Inflammatory changes in the kidney caused a slight increase (approximately 12%) in organ weight. There were no differences in the weight of the adrenal glands or ovaries.

The age and body weight on the day in which vaginal opening was observed was comparable among all groups (Table 4). There were no differences in nipple retention in any group. The mean duration of estrous cycles was increased in group 3 females to 6.10 days versus 5.14 and 5.17 days in the saline and placebo groups (Table 5). This corresponded to a decrease in the total number of estrous cycles within the 3-week premating period for group 3. There were no other Ziracin-related changes in the duration or number of estrous cycles.

The mating and fertility indices were decreased in groups 3 and 5 to 33.3% and 29.2% and 55.6% and 85.7%, respectively, versus 100% in the saline control. Two out of the 5 pregnant females in group 3 were found to have implantations (resorptions) but no fetuses. The small number of pregnant F1 females in these groups precluded a complete assessment of parturition parameters (duration of gestation, pup sex ratio, and pup weight at delivery). Hypoplasia of the mammary gland, characterized by ducts without any acinar development, was seen in a few F1 females from groups 3 and 5. During lactation, the mammary glands of the maternal F₁ females in groups 3 and 5 were underdeveloped. Histologically, only minimal secretory mammary gland development occurred in lactating F1 females compared to the normal postpartum glandular appearance seen in females from other groups.

F₂ Generation

In the F2 generation, Ziracin-related effects were only observed in the offspring from groups 3 and 5 as expected. The F₂ pup survival was greatly decreased in groups 3 and 5 with 30.8% and 0% of pups alive on postnatal day 4 versus 96.8% in the saline control. The low survival of the offspring correlated to the reduced mammary gland development in the maternal F₁ females. Likewise, body weights on postnatal day 1 for F₂ pups were decreased to 5.3 and 5.4 g in those 2 groups compared to 6.6 g in the saline control. There were no other remarkable clinical observations for the F₂ pups. There were no effects attributed to the Placebo.

Discussion

This study demonstrated that Ziracin-induced congenital malformations in the offspring (F1) of female rats dosed during gestation. Macroscopically, Ziracin-exposed F1 females exhibited malformations of the external genitalia, characterized by reduction of the distance between the vulva and urethral orifices, a cleft genital tubercle and malpositioning of the urethral opening into the vagina, thus creating an urogenital cloaca. In addition, a dose-related increase in secondary inflammatory changes was seen in the kidneys, urinary bladder, urethra, and vagina and was attributed to impaired urine outflow, ascending urinary tract infection and/or urolithiasis. Interestingly, only the female offspring showed adverse findings. Males born to Ziracin-treated dams were not affected (unpublished company reports). The dams were not adversely affected. Lower survival was also observed in the F2 pups, which was attributed to the inability of the underdeveloped mammary gland of the F1 to supply adequate amount of milk to the neonates.

The current study was designed to identify the period of sensitivity during which the offspring were sensitive to the teratogenic effects of the drug. Females born in groups 3 and 5, dosed from GD6 to LD21 and from GD14 to the last day of gestation respectively, were the only ones affected. No effects were seen when Ziracin was administered earlier to dams, from GD6 to GD13, or later during lactation, from LD1 to LD21. This identifies the period of sensitivity of the female fetus to the developmental effects of Ziracin as from GD14 to the last day of gestation since this was the common period of administration to both affected groups 3 and 5. The lack of effects in the other groups exonerates both early embryonic exposure and exposure during lactation. Indeed, affected pups born to females exposed during GD14 to the last day of gestation were not exposed through the milk. Ziracin, having a half-life of 6 hours in the rat (Lin et al., 2000), was virtually eliminated from the dam’s plasma when she started lactating. And additionally, no significant oral absorption of Ziracin from milk to the pups was detected in another study that assessed the transfer of radiolabeled Ziracin into milk of lactating rats and absorption by suckling pups (unpublished company reports). The dose of 60 mg/kg was selected based on the results of a previous study in which most rats exhibited anomalies of the external genitalia at that dose (unpublished company report). The lack of abnormal findings in the placebo control group in this study indicated that the placebo formulation did not play a causative role in the observed reproductive toxicity.

The anomaly seen with Ziracin corresponds to female hypospadias as described in the literature (Williams and Bloomberg, 1976). Morphogenetically, female hypospadias results from incomplete development of the urethra and the ventral wall of the lower part of the vagina (Williams and Bloomberg, 1976). In rats, gestation day 14–15 corresponds to the time when the Mullerian ducts, fused to form the anterior part of the vagina, opened into the urogenital sinus (Forsberg and Kalland, 1981). Differentiation of the external genitalia commences around GD19 and continues postnatally. The lack of abnormal findings in the ovaries, uterus, upper part of the vagina, and rectum attests to the normal development of the urogenital ridges, partitioning of the cloaca by the urorectal septum, fusion and canalization of the paired Mullerian ducts, and regression of the mesonephric ducts. It appears that Ziracin, when present during late gestation, alters normal development and differentiation of the lower vagina, urethra, and genital tubercle.

The fact that such a limited anatomical area is affected implicates the urogenital sinus, which forms the urethra and the lower third of the vagina, and the genital tubercle, including the genital swellings and urethral folds, as targets of altered morphogenesis (Inomata et al., 1985).

Hypospadias is a relatively common congenital malformation in humans, averaging 1 per 250 male and female births (Baskin et al., 2001). In contrast, spontaneous hypospadias is seldom observed in rats. Because of its predictability and exquisite sensitivity, the rat in utero exposure model is recognized as the model of choice to identify endocrine disruptors and hormonally active compounds that impact the development of the reproductive tract (Gray and Kelce, 1996). Hypospadias has been induced in this model by compounds with antiandrogenic activity, natural and synthetic estrogens, antiestrogens, TCDD and retinoic acid (Vannier and Raynaud, 1980; Henry and Miller, 1986; Gray and Ostby, 1995; Gray and Kelce, 1996; Flaws et al., 1997; Baskin et al., 2001; Bowman et al., 2003). More specifically, prenatal administration of androgens causes, in female offspring, an array of malformations, which impact the lower part of the vagina, Wolffian duct-derived structures and external genitalia by creating a cleft phallus (Wolf et al., 2002).

Androgenization of the vagina ranges from agenesis of the lower vagina with vaginal atresia to junction of the vagina with the urethra in a way that mimics a masculinized pattern of development. Similarly, Ziracin, given during gestation, altered the development of the genital tubercle and the urogenital sinus in females, forming an urethrovaginal junction. This pattern appears compatible with the hypothesis that Ziracin exerts an androgenic effect on the developing organism. Increased anogenital distance in females and the absence of notable effects in males (except at very high doses) have been seen with prenatally administered androgens (Wolf et al., 2002). There was no delayed vaginal opening in our F1 females and normal estrous cycles were observed with the exception of only a slight increase in duration. Both delayed vaginal opening and abnormal estrous cycles have been observed following masculinization of gonadotropin release. The absence of these findings in this study is consistent with the fact that the sensitive period for masculinization of gonadotropin release is the neonatal period and not the prenatal one (Wolf et al., 2002). Reduced mating and pregnancy indices were noted in Ziracin-treated rats but were considered secondary to the anatomical defects and genito-urinary infections.

Hypoplasia of the mammary gland and poor postpartum secretory activity in F1 females could also be explained by hormonal disturbances during gestation, impairing normal development. Indeed, studies have demonstrated that permanent effects on mammary gland maturation follow exposure to hormonally active toxicants around GD15–19 (Fenton et al., 2002; Rayner et al., 2004). In human females, virilization of the fetus is the most common malformation of the urogenital sinus and is generally due to congenital adrenal cortical hyperplasia (Williams and Bloomberg, 1976; Zaontz and Packer, 1997; Dasgupta et al., 2003). In this syndrome, excessive levels of androgens are produced from hyperplastic adrenal glands under ACTH stimulation, as an enzymatic defect in the synthesis of cortisol causes the loss of negative feedback on the pituitary axis. An adrenal gland origin for the androgenic activity can be excluded in this case because the adrenal glands of Ziracin-treated rats were indistinguishable from those of controls.

Although a small percentage of cases results from single gene mutations, hypospadias is generally believed to have a multifactorial etiology, in which various genetic and environmental factors participate to different degrees (Baskin et al., 2001). Mutations in the HOXA13 gene have been demonstrated to be at the basis of the pathogenesis in the human Hand-Foot-Uterus syndrome (Baskin et al., 2001; Mauch and Albertine, 2002; Perriton et al., 2002). In mouse mutants, combinations of transgenic Hoxa-13/Hoxd-13 mice exhibit various abnormalities of the genitalia with malpositioning of the vaginal, urethral and/or anal openings, cloaca, and absent genital tubercle (Warot et al., 1997). Most recently, Hox and Sonic hedgehog (Shh) genes and associated signaling pathways have been shown to play a crucial role in the development and differentiation of the distal urethra and genital tubercle (Haraguchi et al., 2001; Ogino et al., 2001; Perriton et al., 2002; Suzuki et al., 2002, 2003). Ziracin-induced dysmorphogenesis was restricted to the distal urethra and vagina and genital tubercle making these genes potential candidates. There were no anomalies detected in the limbs or the external ear, which could have suggested similarities with known genetic syndromes in humans and animals. However, inappropriate gene expression as a result of endocrine disruption provides a plausible mechanism for Ziracin effects. Hox genes, a number of key signaling proteins, and growth factors implicated in urethrogenital morphogenesis have been shown to be influenced by hormones and/or compounds with hormonal activity (Gupta et al., 1996; Baskin et al., 2001). Some of these factors are involved in epithelial-mesenchymal interactions and regulation of cell proliferation and apoptosis, shown to play a role in the teratogenicity of other compounds. For instance, retinoic acid causes genitourinary malformations partly through interference with Shh and Hox genes expression and decreased cell proliferation (Mesrobian et al., 1994; Mauch and Albertine, 2002). Also, the basis for the clefting of the clitoris and hypospadias induced by TCDD is believed to be alteration of cell proliferation and apoptosis (Gray and Ostby, 1995; Flaws et al., 1997). Differences in gene expression between control and Ziracin-exposed tissues should help elucidate these questions.

Ziracin is a novel oligosaccharide antimicrobial of the everninomicin class. Avilamycin, the only other antimicrobial of that class, is approved for use in veterinary medicine as preventative for swine dysentery and growth promotant in pigs (Merck Index, 2001) and has not, to the authors’ knowledge, been associated with birth defects. Clinical experience showed that Ziracin is generally safe and well tolerated in healthy volunteers and patients (Foster and Rybak, 1999). There were no reports of drug-related congenital malformations in the literature on the everninomycin family. Hypospadias can also occur as a nonspecific malformation associated with low birth weight (Baskin et al., 2001). In this study, however, no differences in body weight could support this hypothesis. The absence of deleterious effects in males is compatible with an androgenic effect of Ziracin. It is well known that early patterning of the genital tubercle and urethra occurs independently of steroid hormones and that later development is under the control of androgens in the male (Baskin et al., 2001; Perriton et al., 2002; Suzuki et al., 2002). Androgens given prenatally cause minimal effects in the male offspring at very high doses where a slight decrease in anogenital distance and reduction in glans penis weight are observed (Wolf et al., 2002). In conclusion, this study demonstrated that Ziracin administered prenatally from GD14 to the last day of gestation induces congenital malformations of the female urogenital tract (increased anogenital distance, abnormal external genitalia, and urethrovaginal junction) and decreased development of the mammary gland. These data, combined with the lack of effects in male offspring, suggest a potential androgenic action of Ziracin although further experiments would be needed to ascertain this mechanism.

Although the significance of these findings for human safety is not known, in view of the reproductive findings in the offspring of Ziracin-treated rats, women of child-bearing potential were counseled to use effective methods of contraception during treatment.