Abstract
The general population, including children and adolescents, is exposed to 4-methylimidazole (4-MI) in the diet. 4-MI is a by-product of caramel color manufacturing. It has been previously classified as a possible human carcinogen, and displays potential reproductive toxicity.A follow up assessment of reproductive toxicity was conducted in rats utilizing the reproductive assessment by continuous breeding paradigm, which multiple generations were exposed to 4-MI in diet at 750, 2500, and 5000 ppm. 4-MI exposure was associated with delays in preputial separation and vaginal opening, impairment in reproductive performance, and concomitant histopathological findings in the prostate, testis, and epididymis at 2500 and 5000 ppm. The Lowest Observed Adverse Effect Level for reproductive (based on prostate atrophy) and developmental toxicity (based on delays in preputial separation and vaginal opening) was 750 ppm, equivalent to approximately 50–60 mg/kg bw/day.
Keywords: 4-methylimidazole, reproduction, development, testicular toxicity
1. INTRODUCTION
One of the main sources of 4-methylimidazole (4-MI) exposure to the general population is through diet. It is found in products such as carbonated beverages, barbecue sauce, and pancake syrup that contain caramel colors produced using ammonium compounds (namely Class III and Class IV caramel) (Hengel and Shibamoto 2013; Schlee et al. 2013). It is also found in coffee, cocoa, and roasted foods in which 4-MI is formed as a byproduct of non-enzymatic browning during thermal treatment (Casal et al., 2002; Mottier et al., 2017; Folmer et al., 2018).
In the United States, the FDA analyzed over 700 food and beverage samples collected from 2013 to 2015 and quantified the amount of 4-MI that may be present in foods containing added caramel color using the 2009–2012 NHANES 2-day dietary intake (Folmer et al., 2018). These data were used to develop a comprehensive dietary exposure assessment for 4-MI for the U.S. population aged 2 years or more, and several subpopulations. The highest (90th percentile) exposure scenario for the US population aged 2 years and older was found to be 2.8 μg/kg body weight (bw) per day.
Several recent studies on 4-MI containing products have also been conducted in Korea and Europe and findings indicate widespread potential for ingestion and exposure of 4-MI. A study from Belgium measured levels of 4-MI in over 500 food samples; concentrations as high as 2.8 mg/kg food were noted for 4-MI with an average daily intake of 0.4–3.7 μg/kg bw/day. Coffee, cola and beer contributed most to the dietary intake in Belgium (Fierens et al., 2018; Jacobs et al., 2018). Another study reported levels of 4-MI ranging from < 5 to 983 mg/kg in 40 commercial caramel colorings; these levels meet the Joint Expert Committee on Food Additives (JECFA) guidelines (Licht et al., 1992; IARC 2010). In general, higher 4-MI levels have been found in Class IV caramel colorings as shown by a study on 90 commercial orally consumed products with levels ranging from 112 to 1276 mg/kg of 4- MI (IARC 2010). 4-MI was also detected in pancake syrup at an average level of 11.5–38 micrograms per ¼ cup serving size, as well as in soft drinks at an average of > 29 micrograms per 12-ounce sample (Consumer Reports 2014 a, b; Folmer et al., 2018; Lee and Lee, 2016; Choi and Jung 2017). Occupationally, most exposure to 4-MI likely occurs dermally in the manufacturing of several pharmaceuticals, photographic chemicals, dyes and pigments, cleaning and agricultural chemicals, and rubber (Chan et al., 2008).
There are currently no data in humans directly linking 4-MI exposure to cancer. California added 4-MI to its list of probable carcinogens and has stipulated 29 μg per day (~0.4 mg/kg/day) as the “No Significant Risk Level” intake. IARC has also listed 4-MI as a probable carcinogen (IARC 2010). This was primarily based on chronic exposure studies previously conducted by the National Toxicology Program (NTP), which reported clear evidence of carcinogenic activity in male and female mice based on increased incidence of lung neoplasms when exposed to 312, 625 and 1250 ppm of 4-MI in feed (NTP 2004; 2006, Chan et al., 2008). Male F344/N rats exposed to 5000 ppm 4-MI in the diet for 14-weeks displayed adverse effects in the male reproductive system including lower testis weight, seminiferous tubule degeneration, atrophy of the prostate gland, and decreased sperm motility and density (NTP 2004, 2006). These findings suggest that 4-MI and structurally related compounds (Adams et al., 1998) could adversely affect reproductive performance. Based on these findings, and the potential for continuous long-term exposure of the general public throughout all stages of life including children, teenagers, and people of reproductive age, the NTP evaluated the potential for reproductive toxicity following 4-MI exposure over multiple generations utilizing the reproductive assessment by continuous breeding (RACB) paradigm previously described by Chapin and Sloan (1997). Feed was used as the route of exposure to mimic likely human exposure.
2. MATERIALS & METHODS
This NTP study was conducted at Research Triangle Institute (RTI) International in compliance with the U.S. FDA (1988, 1994) Good Laboratory Practice (GLP) regulations.
2.1. Animals and Husbandry
Adult Hsd:Sprague Dawley SD rats, the animal model typically used for NTP developmental and reproductive studies, were purchased from Harlan Laboratories (now called Envigo, Dublin, VA) and housed in solid-bottom, polycarbonate cages containing Sani-Chips® cage litter (P.J. Murphy Forest Products, Montville, NJ), suspended on automatic watering racks with filter sheets. The quarantine period was approximately two weeks. Adult males and females were singly housed except during cohabitation or when dams were with their respective litters. F1c weanlings were housed 2/sex/cage until approximately post-natal day (PND) 91. Irradiated NIH-07 Certified rodent diet (Zeigler Bros., Inc., Gardners, PA) was available ad libitum throughout the study. Tap water was available ad libitum via an automatic water delivery system (Edstrom Industries Inc., Waterford, WI). Water was periodically assessed and met NTP’s requirements. Animal rooms were maintained on a 12-hour light/dark cycle per day (~0600/1800 on/off) at 72 ± 3°F and relative humidity of 50% ± 15%. Sentinel animals were used on the study to confirm the health status.The study was reviewed and approved by the RTI IACUC. Animals were handled, cared for, and used in compliance with the NRC Guide for the Care and Use of Laboratory Animals (2011).
2.2. Chemical and Exposure
The dose formulation development was performed by Battelle Columbus under the NTP chemistry support contract. 4-MI (CAS# 822–36-6) was procured from Sigma-Aldrich (St Lois, MO; lot# 119H5114) and Alfa Aesar (Ward Hill, MA; lot# C07T016). The two lots were combined to make one homogeneous lot (lot# 051410) and used in studies. The identity of the combined lot was confirmed by infrared spectroscopy, and 1H and 13C nuclear magnetic resonance spectroscopy. The purity (> 99%) was determined by high performance liquid chromatography (HPLC) with ultra violet (UV) detection (215 nm) and gas chromatography with flame ionization detection. Throughout the study, unless otherwise noted, animals were continuously exposed to 4-MI mixed in certified NIH-07 feed (Zeigler Brothers, Inc., Gardners, PA). Prior to study start, stability of 4-MI in feed (50 to 5000 ppm) was confirmed for up to 42 days at ambient temperature. Concentration of 4-MI in feed was confirmed analytically throughout the study using a validated HPLC-UV method. All formulations were homogeneous and within 10% of target concentration. Analyses for dose concentration confirmation were conducted by Battelle (Columbus, OH) under the NTP chemistry support contract. Body weights and feed consumption collected throughout the study (pre-cohabitation, cohabitation, gestation, and lactation) were used to calculate 4-MI intake.
2.3. RACB Study Design
The RACB is a two-generation study design developed by the NTP for use in identifying potential hazards/toxic effects on male and female reproduction, to characterize that toxicity, and to define the dose-response relationships for the test article (Chapin et al., 1997). In this study, multiple breedings of the F0 and F1 generation were conducted to evaluate potential 4-MI-induced reproductive toxicity after continuous exposure to 0, 750, 2500, or 5000 ppm via the diet. These exposure levels are based on data from a dose-range finding study with a summary of select end-points in Supplemental Table 1 and all data are provided here: CEBS Tables. Male and female Hsd:Sprague Dawley SD rats, designated the F0 generation, were randomized into 4 groups of 23 rats/sex/dose-group. Rats were continuously exposed beginning two weeks prior to cohabitation until necropsy unless otherwise noted. Endpoints measuring postnatal development were collected and included evaluation of developmental landmarks and pubertal indices (Figure.1).
Figure 1:
Schematic of the Reproductive Assessment by Continuous Breeding (RACB) Design
Rats were administered 4-MI in the diet at 0, 750, 2500, or 5000 ppm ad libitum. The F0 adults were exposed during a 2-week prebreed exposure period, during cohabitation, gestation and lactation for the F1a, F1b, and F1c generations until necropsy. The F1c generation was exposed throughout life either indirectly via the mother during gestation and lactation and then by direct exposure to the dosed feed beginning on PND 28 until necropsy. The F2c generation was exposed to 4-MI via the mother during gestation and lactation. Reproductive and developmental performance was measured as described in the text. Additionally, a crossover mating was performed on the F0s following generation of the F1c to determine affected sex; during crossover mating, the 4-MI treated animals were crossed with naïve animals of the opposite sex.
Male and female rats of the first generation (F0; age 10–12 weeks, n = 23/sex) were cohabitated until evidence of mating or up to 15 days with the same partner three consecutive times to potentially produce a litter (labelled as F1a, F1b, and F1c). The F1a and F1b litters were euthanized on PND 4. The F1c litters were retained, weaned and exposure continued until they were sexually mature (~PND 95) upon which time they were paired with a life-long partner (avoiding sibling matings) to potentially produce three litters (labelled as F2a, F2b, and F2c). Similar to the F1, the F2a and F2b litters were euthanized on PND 4. The F2c litter and corresponding dams were evaluated through PND 28. The day of confirmed mating via presence of a vaginal plug or sperm was considered Gestation Day (GD) 0. The day of littering was designated PND 0. F0 animals were observed for moribundity and mortality at least twice daily and detailed clinical observations were conducted once daily. F0 male body weights were recorded twice weekly. F0 male feed consumption was recorded twice weekly except during the periods of cohabitation. F0 female body weights were recorded at least twice weekly during the prebreed interval, once daily during gestation on GD 0 to 21, and during lactation on PND 1, 4, 7, 13, 16, 19, 21, 25, and 28. Female feed consumption was recorded twice weekly during the prebreed interval, and for the following gestational and lactational intervals, respectively: GD 0–3, 3–6, 6–9, 9–12, 12–15, 15–18, and 18–21 during gestation, and for PND 1–4, 4–7, 7–10, 10–13, 13–16, 16–19, 19–21, 21–25, and 25–28.
Following weaning on PND 28, F1c animals were observed for survival at least twice daily and detailed clinical observations were conducted once daily. F1c male body weights were recorded once weekly until scheduled termination. F1c male feed consumption was recorded once weekly except during the periods of cohabitation. F1c female body weights and feed consumption were recorded at least once weekly during the prebreed interval, and during gestation and lactation for the same intervals indicated as the F0 females.
Sperm parameters were evaluated in the F0 (~PND 230), F1c-interim (~PND 95), and F1c terminal (~PND 220) groups. Anogenital distance (AGD) and body weight was recorded on PND 1 for all pups. AGD was measured using a stereomicroscope with a calibrated eyepiece diopter containing a micrometer. Individual pups were sexed and weighed on PND 1 and 4 (a and b litters) or 1, 4, 7, 13, 16, 19, 21, 25, and 28 (c litter). The “c” litters of the F1 and F2 generations were standardized to a litter size of 12 pups (6/sex/litter where possible) on PND 4. Male pups were evaluated for retention of areolae/nipples on PND 13 and for testes descent beginning on PND 14. F1 and F2 offspring that were removed due to death or moribundity during lactation received a gross necropsy.
The F1c offspring were weaned on PND 28 and were randomly assigned to either 1) the F1c-interim cohort (or “non-parental cohort”), which were necropsied at ~PND 95 to assess ongoing toxicity(up to 3/sex/litter; n = 20–56 (M); n = 27–58 (F) across the exposure groups), or 2) the F1c terminal cohort (or “parental cohort”) that was assigned to be paired three times (up to 2/sex/litter; n = up to 44/sex/group) to produce the F2 generation. The F2a and F2b litters were euthanized on PND4, and the F2c animals were euthanized on PND 28 after measuring similar developmental markers as with the F1 (AGD, testis descent, areolae/nipple retention). The acquisition of vaginal opening was evaluated in all F1c females beginning on PND 23 and the acquisition of preputial separation was evaluated in F1c males beginning on PND 35. Vaginal smears were collected from the F0 and F1c- terminal females for 16 consecutive days for evaluation of estrous cyclicity. Collection of vaginal smears began when the F0 females were approximately 33 weeks of age, and on PND 52 for females in the F1c terminal (parental) cohort.
2.3.1. Crossover Mating
Due to a decrease in litter size in the F1 generation, a crossover mating was conducted to investigate which sex was affected by the exposure. Following the third pairing, the F0 males exposed to 0, 2500 and 5000 ppm 4-MI were paired with naïve females. The 0 and 2500 ppm F0 females were paired with naïve males. The naïve animals were approximately 13–14 weeks of age. Due to the moribundity noted in the 5000 ppm females, a crossover mating was not conducted in that group. Cross-over mating animals were cohabited for seven days with females, and checked each morning for evidence of mating. Exposure to 4-MI containing feed was suspended during the overnight cohabitation period; all animals were presented with control diets during cohabitation. The females in the crossover mating were allowed to litter and their offspring were euthanized on PND 4. Live and dead pups were counted and sexed daily on PND 0 through 4, and live pups were weighed on PND 1 and 4. Females that did not deliver were euthanized 24 days after the last day of cohabitation and received an examination to determine pregnancy status.
2.3.2. Necropsy
F0 and F1 generation male and female rats were weighed at study termination, euthanized, subjected to necropsy, and select tissues were fixed and trimmed in for histopathology. The F0 males and females were approximately 33 and 39 weeks old respectively, at necropsy. F1c males and females not used for mating (F1c-interim cohort) were euthanized at approximately 95 days of age. F1c parental (F1c terminal) males and females were euthanized following completion of pairing for production of the F2c generation, corresponding to 218 to 224 days of age for the males and 241 to 261 days of age for the females. At the necropsy of the F2c offspring on PND 28, a complete gross evaluation was performed, and only gross lesions of the kidneys or testes were evaluated for histopathology. Male reproductive organs collected for histopathology from adult F0 and F1s included right testis and epididymis, dorsolateral prostate, ventral prostate, seminal vesicles with coagulating glands, bulbourethral glands (Cowper’s glands), and levator ani bulbocavernosus muscle complex (LABC). Female reproductive organs collected for histopathology included uterus, cervix, vagina, and left and right ovaries. Other tissues collected for histopathology included liver, kidneys, thymus, and endocrine glands (pituitary, thyroid, and adrenal). In addition to selected tissues, gross lesions were retained in fixative and trimmed in for histopathology.
Tissues were fixed in neutral buffered 10% formalin except the right testis and epididymis and left and right ovaries, which were initially fixed in modified Davidson’s fixative for approximately 24 hours prior to transfer to formalin. All tissues were embedded in paraffin and stained with hematoxylin and eosin except for the right testis and epididymis, which were stained with periodic acid-Schiff’s stain and counterstained with hematoxylin.
Cauda epididymal sperm motility and concentration, and testicular spermatid head counts were evaluated in the left testis and epididymis of F0 and F1c terminal males. Ovarian follicle counts for primordial follicles, primary follicles, secondary follicles, antral follicles, and atretic follicles were generated from 5 serial stepped sections from the left and right ovaries for F0 and F1c terminal females (n = 12–15/group).
2.3.3. Histopathology
Each of the tissues from all of the control and high-exposure (5000 ppm for F0, and 2500 ppm for F1c due to overt toxicity at 5000 ppm) animals were sectioned, stained, and examined. If treatment-related histopathologic effects were observed, then tissues from animals in the lower exposure groups were examined. For the F0 and F1 generations, tissues from all rats removed from study prior to scheduled necropsy and all reported gross lesions were evaluated grossly and histologically.
The data from this study was subjected to additional NTP pathology peer-review procedures. The slides, study lab report, individual animal data records, and pathology data were evaluated by a second, independent NTP quality assurance (QA) contract laboratory. A QA pathologist evaluated slides from reproductive and endocrine organs from all males and females, accessory sex glands from males, and mammary glands from females. Representative histopathology slides containing examples of lesions related to chemical administration (except for testicular lesions), examples of disagreements in diagnoses between the study laboratory and QA pathologist, or lesions of general interest were presented by the QA pathologist to a pathology working group (PWG) for review. The PWG consisted of the laboratory study pathologist, the NTP pathologist for the study, and 8 other pathologists experienced in rodent toxicologic pathology. Final diagnoses for reviewed lesions represent a consensus between the study pathologist, the reviewing (QA) pathologist, and the PWG. The slides of testis and epididymis from the F0 and F1c generations were reviewed by a third pathologist with an awareness of stages of the spermatogenic cycle. This pathologist then conducted a second PWG, and issued a report on the testis and epididymis representing the consensus diagnoses of the PWG.
2.4. Statistical Analyses
Statistical methods differed for F0 and F1 animals, since methods for the F1 animals needed to account for within litter correlation where present. For F0 reproductive performance endpoints (CEBS Table# R21; crossover mating, Table 5; Supp. Table 1), statistical analysis was performed by Cochran-Armitage (trend) and Fisher Exact (pairwise) two-sided tests (Gart et al., 1979). For F1 reproductive performance endpoints, analysis was performed using the Rao-Scott Cochran-Armitage procedure for both trend and pairwise tests (Rao and Scott, 1992). F0 litter size and survival endpoints (Table 5) were analyzed using Jonckheere’s test for trend. For pairwise comparison of dosed groups to controls, Shirley’s (1977) (as modified by Williams 1986) or Dunn’s (1964) method was used depending on detection of a significant trend. F1 litter size and survival endpoints were analyzed using a permutation test based on the Jonckheere trend test (Jonckheere 1954), and pairwise comparisons were made using a modified Wilcoxon test (Datta and Satten 2005) with the Hommel procedure (Hommel 1988) to adjust for multiple comparisons.
Table 5:
Average litter size across the pairings in each generation (average ± SEM)
| F0 | F1c | ||||||
|---|---|---|---|---|---|---|---|
| 0 ppm | 750 ppm | 2500 ppm | 5000 ppm | 0 ppm | 750 ppm | 2500 ppm | |
| Total Litter Size (PND 0)a | |||||||
| A | 14.6 ± 0.4** | 13.3 ± 0.6 | 9.9 ± 0.7** | 5.7 ±1.9** | 13.2 ± 0.5** | 11.0 ± 0.6* | 8.9 ± 0.8** |
| B | 14.2 ± 0.5** | 13.5 ± 0.6 | 9.2 ± 0.7** | 4.7 ± 0.9** | 15.3 ± 0.5** | 12.3 ± 0.5** | 10.6 ± 0.9** |
| C | 13.6 ± 0.6* | 12.8 ± 0.6 | 10.6 ± 1.2* | - | 10.7 ± 0.8 | 9.8 ± 0.7 | 9.4 ± 0.9 |
| 4-MI Male × Naïve Female | 13.8 ± 0.5 | - | 11.6 ± 1.1 | 11.7 ± 1.8 | - | - | - |
| 4-MI Female × Naïve Male | 11.4 ± 1.5 | - | 11.3 ± 1.7 | - | - | - | - |
| Live Litter Size (PND 0) | |||||||
| A | 14.0 ± 0.4** | 12.7 ± 0.6 | 7.8 ± 1.0** | 1.8 ± 1.6** | 11.1 ± 0.7** | 9.3 ± 0.7 | 7.2 ± 0.8** |
| B | 13.7 ± 0.5** | 12.7 ± 0.5 | 8.5 ± 0.6 | 0.5 ± 0.5** | 13.6 ± 0.6** | 11.9 ± 0.6 | 9.8 ± 0.8** |
| C | 12.2 ± 0.8 | 12.4 ± 0.5 | 9.8 ± 1.2 | - | 9.2 ± 0.7 | 8.8 ± 0.6 | 8.5 ± 0.9 |
| 4-MI Male × Naïve Female | 13.4 ± 0.4 | - | 10.6 ± 1.1 | 11.5 ± 1.9 | - | - | - |
| 4-MI Female × Naïve Male | 10.7 ± 1.4 | - | 11.0 ± 1.7 | - | - | - | - |
| Survival Ratio (PND 1 – 4) | |||||||
| A | 0.96 ± 0.02 | 0.98 ± 0.01 | 0.91 ± 0.04 | 0.10 ± 0.10 | 0.92 ± 0.05** | 0.93 ± 0.04 | 0.72 ± 0.08** |
| B | 0.97 ± 0.02 | 0.98 ±0.01 | 0.93 ± 0.03 | 0.67b | 0.97 ± 0.01* | 0.97 ± 0.01 | 0.91 ± 0.03* |
| C | 0.95 ± 0.03 | 0.98 ± 0.01 | 0.93 ± 0.04 | - | 0.89 ± 0.05 | 0.92 ± 0.05 | 0.88 ± 0.04 |
| 4-MI Male × Naïve Female | 0.97 ± 0.01 | - | 0.96 ± 0.02 | 0.99 ± 0.01 | - | - | - |
| 4-MI Female × Naïve Male | 0.99 ± 0.01* | - | 0.82 ± 0.10* | - | - | - | - |
| Survival Ratio (PND 5–28) | 0.98 ± 0.01 | 0.96 ± 0.02 | 0.95 ± 0.02 | - | 0.83 ± 0.07 | 0.93 ± 0.02 | 0.89 ± 0.03 |
| Average Live Litter Size/Pair | 13.2 ± 0.4 | 12.6 ± 0.3 | 9.2 ± 0.6 | 2.0 ± 1.0 | 11.4 ± 0.4** | 10.1 ±0.5** | 8.3 ± 0.6 |
F0 litter size and survival endpoints were analyzed using Jonckheere’s test for trend (0 ppm column) and Shirley’s or Dunn’s methods for pairwise comparison of controls to dose groups. F1c data were analyzed using the bootstrapped Jonckheere test for trend; pairwise comparisons used the Datta-Satten modified Wilcoxon test with the Hommel adjustment for multiple comparisons. Testing for trend and pairwise differences was not performed for sample sizes of 1 or 2.
p < 0.05
p < 0.01
n = 1 litter
To analyze attainment of developmental endpoints (testes descent (TD), vaginal opening (VO), preputial separation (PPS) in Table 6), trend and pairwise tests were based on mixed models for age at attainment with dose as a covariate and a random effect for litter followed by a Dunnett-Hsu adjustment for multiple comparisons. For VO and PPS, weaning weight was included as a covariate in mixed models. To calculate age at attainment adjusted for pup body weight, a linear model was fit to attainment age as a function of body weight. Then the estimated coefficient for body weight was used to adjust each pup’s observed attainment age based on the difference between its body weight and the overall mean body weight. For epididymal sperm endpoints (Table 7), F0 animals were analyzed using the Jonckheere trend test followed by Shirley’s or Dunn’s method for pairwise comparisons. F1 animals with littermates were analyzed with a bootstrapped Jonckheere trend test where litters were permuted across dose groups, and animals with the same maternal dam were sampled with replacement. Modified Wilcoxon tests were used for pairwise comparisons, with the Hommel adjustment for multiple comparisons.
Table 6:
Developmental markers (mean ± SEM) in male and female rats after 4-MI exposure
| 0 ppm | 750 ppm | 2500 ppm | |
|---|---|---|---|
| F1c Examined, Males (no. of litters) | 99 (18) | 115 (22) | 61 (15) |
| Areolae/nipples per litter | 0 | 0 | 0.14 ± 0.10 |
| Pups with areolae/nipples (%) | 0 (0)* | 0 (0) | 3 (4.92) |
| Litters with areolae/nipples (%) | 0 (0) | 0 (0) | 2 (13.33) |
| Day of testis descent | 16.7 ± 0.2* | 16.8 ± 0.2 | 17.1 ± 0.2 |
| F2c Examined,Males (litters) | 108 (25) | 133 (32) | 69 (20) |
| Areolae/nipples per litter | 0 | 0 | 0.17 ± 0.17 |
| Pups with areolae/nipples (%) | 0 | 0 | 3 (4.35) |
| Litters with areolae/nipples (%) | 0 | 0 | 1 (5.00) |
| Day of testis descent a | 18.1 ± 0.3** | 18.5 ± 0.4 | 19.4 ± 0.4* |
| F1c Examined, Males (litters) | 89 (18) | 100 (22) | 60 (15) |
| Age at PPS (PND) | 43.5 ± 0.4** | 46.2 ± 0.4** | 47.2 ± 0.6** |
| Body Weight at PPS (g) | 195.8 ± 2.7 | 205.6 ± 2.9* | 190.4 ± 3.4 |
| Body Weight at PND 28 (g) | 85.4 ± 1.6** | 80.4 ± 1.7 | 73.1 ± 2.9** |
| Adjusted age at PPS b | 44.3 ± 0.3* | 46.4 ± 0.4** | 46.4 ± 0.5* |
| F1c Examined, Females (litters) | 96 (19) | 111 (22) | 67 (15) |
| Age at VO (PND) | 33.8 ± 0.2** | 37.2 ± 0.3** | 39.4 ± 0.3** |
| Body Weight at VO (g) | 106.2 ± 2.0** | 117.2 ± 2.0** | 117.1 ± 1.8** |
| Body Weight at PND 28 (g) | 76.7 ± 1.6** | 71.3 ± 1.5* | 64.3 ± 2.3** |
| Adjusted age at VO b | 34.1 ± 0.2** | 37.2 ± 0.3** | 39.0 ± 0.3** |
Means of litter means for age at attainment are presented. Trend (0 ppm column) and pairwise tests for age at attainment were based on mixed models with dose as a covariate and a random effect for litter, with a Dunnett-Hsu adjustment for multiple comparisons. For PPS and VO, mixed models included weaning weight as a covariate. Mixed models for body weight at attainment and body weight at weaning included dose as covariate and a random effect for litter, with a Dunnett-Hsu adjustment for multiple comparisons.
p < 0.05
p < 0.01
Number of animals (litters) examined for testicular descent was 107 (25), 132 (32), and 68 (20) respectively.
Means of adjusted age at PPS and VO were calculated as the mean of the litter means of the weaning weight-adjusted attainment age for individual pups.
Table 7:
Epididymal sperm parameters (average ± SEM) of the F0, F1c Interim, and F1c Terminal male rats after 4-MI exposure
| Endpoint | Generation | 0 ppm | 750 ppm | 2500 ppm | 5000 ppm |
|---|---|---|---|---|---|
| Sperm/Cauda (106) | F0 | 180.6 ± 8.2** | 206.6 ± 7.9 | 167.0 ± 11.3 | 135.1 ± 7.3** |
| F1-interim | 187.8 ± 8.9 | 168.8 ± 6.5 | 153.8 ± 11.4 | - | |
| F1-terminal | 196.7 ± 9.0 | 190.2 ± 7.1 | 176.0 ± 10.6 | - | |
| Sperm/g Cauda (106) | F0 | 682.4 ± 29.7* | 752.7 ± 21.2 | 676.5 ± 38.3 | 589.5 ± 28.5 |
| F1-interim | 856.4 ± 26.1 | 780.1 ± 25.8* | 754.4 ± 40.8* | - | |
| F1-terminal | 759.6 ± 27.3 | 723.0 ± 26.2 | 721.2 ± 35.0 | - | |
| % Motile | F0 | 83.3 ± 2.1** | 80.1 ± 1.6 | 76.2 ± 1.8** | 71.9 ± 2.5** |
| F1-Interim | 68.9 ± 1.8 | 68.7 ± 2.0 | 61.9 ± 1.2** | - | |
| F1-terminal | 80.1 ± 1.5** | 77.4 ± 1.2 | 71.7 ± 2.9 | - | |
| % Progressively Motile | F0 | 70.0 ± 1.9* | 68.9 ± 1.3 | 67.2 ± 1.6 | 65.3 ± 2.5 |
| F1-interim | 57.6 ± 1.6 | 57.4 ± 1.7 | 51.4 ± 1.4* | - | |
| F1-terminal | 66.9 ± 1.3 | 66.5 ± 1.2 | 64.4 ± 2.9 | - |
Statistical analysis for F0 data performed by Jonckheere trend (0 ppm column) and Shirley or Dunn pairwise tests. For F1 animals with littermates, a bootstrapped Jonckheere trend test was used, with pairwise comparisons using the Datta-Satten modified Wilcoxon test with a Hommel adjustment. F0 n = 23, 23, 20, 21; F1 – interim n = 49, 56, 20; F1 – terminal n = 40, 44, 39.
p< 0.05
p<0.01
Organ and body weights (Tables 8 and 9; Supp. Tables 1, 4 and 5) in F0 animals were analyzed using a Jonckheere trend test and Williams or Dunnett (pairwise) tests. For organ and body weight endpoints in F1 animals, mixed models were fit with a random litter effect and a Dunnett-Hsu adjustment for both trend and pairwise analyses. Lesion incidence for F0 animals (Table 10) was analyzed using the Poly-3 trend and pairwise statistics. Lesion incidence for F1 animals was analyzed using the Cochran-Armitage test with a poly-3 adjustment for survival (Bailer and Portier, 1988) and a Rao-Scott modification for litter effect. Vaginal cytology data for F1 animals (Supp. Table 3) were analyzed using the modified Wilcoxon test with Hommel adjustment (cycle length, number of cycles). Tests for extended periods of estrus, diestrus, metestrus, and proestrus were constructed based on a Markov chain model proposed by Girard and Sager (1987).
Table 8:
Male rat body and reproductive organ weights (average ± SEM) for each generation after 4-MI exposure
| Endpoint | Generation | 0 ppm | 750 ppm | 2500 ppm | 5000 ppm |
|---|---|---|---|---|---|
| Body Weight (g) | F0 | 504.5 ± 4.4** | 477.5 ± 5.9** | 459.2 ± 7.5** | 455.2 ± 5.5** |
| F1-c interim | 395.0 ± 5.7** | 383.4 ± 4.9 | 338.9 ± 4.4** | - | |
| F1-c terminal | 497.3 ± 7.7** | 486.8 ± 6.7 | 443.8 ± 10.3** | - | |
| Right Testis Absolute (g) | F0 | 2.089 ± 0.023 | 2.089 ± 0.030 | 2.086 ± 0.028 | 2.066 ± 0.029 |
| F1-c interim | 1.928 ± 0.029 | 1.928 ± 0.023 | 1.897 ± 0.047 | - | |
| F1-c terminal | 2.095 ± 0.036 | 2.121 ± 0.026 | 2.182 ± 0.049 | - | |
| Left Testis Absolute (g) | F0 | 2.075 ± 0.026 | 2.074 ± 0.026 | 2.096 ± 0.026 | 2.062 ± 0.029 |
| F1-c interim | 1.931 ± 0.029 | 1.911 ± 0.023 | 1.884 ± 0.038 | - | |
| F1-c terminal | 2.090 ± 0.038 | 2.103 ± 0.030 | 2.150 ± 0.049 | - | |
| Right Epididymis Absolute (mg) | F0 | 681 ± 9** | 686 ± 10 | 628 ± 9** | 600 ± 9** |
| F1-c interim | 571 ± 9** | 564 ± 8 | 513 ± 12** | - | |
| F1-c terminal | 697 ± 9** | 697 ± 9 | 651 ± 11** | - | |
| Right Epididymis relative (mg/g) | F0 | 1.35 ± 0.02 | 1.44 ± 0.03* | 1.37 ± 0.03 | 1.32 ± 0.02 |
| F1-c interim | 1.45 ± 0.02 | 1.47 ± 0.02 | 1.52 ± 0.05 | - | |
| F1-c terminal | 1.41 ± 0.02* | 1.44 ± 0.02 | 1.47 ± 0.02 | - | |
| Left Epididymis Absolute (mg) | F0 | 704 ± 9** | 718 ± 13 | 657 ± 11** | 635 ± 10** |
| F1-c interim | 579 ± 10** | 561 ± 7 | 521 ± 12** | - | |
| F1-c terminal | 697 ± 13** | 701 ± 9 | 648 ± 12** | - | |
| Left Epididymis Relative (mg/g) | F0 | 1.40 ± 0.02 | 1.51 ± 0.04* | 1.44 ± 0.03 | 1.40 ± 0.02 |
| F1-c interim | 1.47 ± 0.02* | 1.47 ± 0.01 | 1.54 ± 0.04* | - | |
| F1-c terminal | 1.41 ± 0.02 | 1.45 ± 0.02 | 1.47 ± 0.02 | - | |
| Dorsolateral Prostate Absolute (mg) | F0 | 604 ± 26** | 492 ± 23** | 469 ± 15** | 421 ± 22** |
| F1-c interim | 402 ± 11** | 382 ± 12 | 330 ± 13** | - | |
| F1-c terminal | 539 ± 16** | 475 ± 19* | 449 ± 19** | - | |
| Dorsolateral Prostate Relative (mg/g) | F0 | 1.20 ± 0.05** | 1.03 ± 0.05** | 1.02 ± 0.03** | 0.93 ± 0.05** |
| F1-c interim | 1.02 ± 0.03 | 1.00 ± 0.03 | 0.97 ± 0.03 | - | |
| F1-c terminal | 1.08 ± 0.03 | 0.98 ± 0.04 | 1.01 ± 0.03 | - | |
| Ventral Prostate Absolute (mg) | F0 | 935 ± 28** | 785 ± 21** | 748 ± 29** | 515 ± 27** |
| F1-c interim | 561 ± 20** | 446 ± 14** | 355 ± 21** | - | |
| F1-c terminal | 825 ± 22** | 796 ± 27 | 591 ± 18** | - | |
| Ventral Prostate Relative (mg/g) | F0 | 1.85 ± 0.05** | 1.65 ± 0.05** | 1.63 ± 0.06** | 1.13 ± 0.06** |
| F1-c interim | 1.42 ± 0.05** | 1.16 ± 0.03** | 1.05 ± 0.06** | - | |
| F1-c terminal | 1.67 ± 0.05** | 1.64 ± 0.06 | 1.34 ± 0.05** | - | |
| Seminal Vesicle Absolute (g) | F0 | 1.946 ± 0.053** | 1.647 ± 0.045** | 1.520 ± 0.045** | 1.253 ± 0.042** |
| F1-c interim | 1.304 ± 0.032** | 1.186 ± 0.028* | 1.016 ± 0.035** | - | |
| F1-c terminal | 1.76± 0.04** | 1.69± 0.04 | 1.46± 0.05** | - | |
| Seminal Vesicle Relative (mg/g) | F0 | 3.87 ± 0.10** | 3.44 ± 0.11** | 3.29 ± 0.08** | 2.75 ± 0.10** |
| F1-c interim | 3.31 ± 0.06* | 3.10 ± 0.07 | 3.00 ± 0.09* | - | |
| F1-c terminal | 3.56 ± 0.08 | 3.48 ± 0.09 | 3.29 ± 0.10 | ||
| LABC Absolute (g) | F0 | 1.438 ± 0.028** | 1.369 ± 0.023 | 1.265 ± 0.034** | 1.236 ± 0.022** |
| F1-c interim | 1.161± 0.027* | 1.095 ± 0.024 | 1.052 ± 0.035 | - | |
| F1-c terminal | 1.341 ± 0.029* | 1.301 ± 0.034 | 1.222 ± 0.041 | - | |
| LABC Relative (mg/g) | F0 | 2.85 ± 0.06 | 2.87 ± 0.05 | 2.76 ± 0.06 | 2.72 ± 0.06 |
| F1-c interim | 2.94 ± 0.05 | 2.86 ± 0.06 | 3.11 ± 0.11 | - | |
| F1-c terminal | 2.70 ± 0.05 | 2.68 ± 0.06 | 2.75 ± 0.07 | - |
Statistical analysis for F0 data performed by Jonckheere trend (0 ppm column) and Williams or Dunnett pairwise tests. Statistical analysis for F1 animals with littermates was performed by mixed models with a random litter effect and a Dunnett-Hsu adjustment for both trend and pairwise analyses.
p< 0.05
p<0.01
“-“ = no animals examined due to early removal
F0 n = 20–23/group; F1-interim n = 20–56/group; F1-terminal n = 39–44/group
Table 9:
Mean body and ovarian weights (± SEM) of F0 and F1 females
| Generation | 0 ppm | 750 ppm | 2500 ppm | 5000 ppm | |
|---|---|---|---|---|---|
| Body weight (g) | F0 | 339.3 ± 6.3 ** | 291.7 ±2.3** | 300.7 ± 5.1** | 275.0 ± 4.1** |
| F1-interim | 242.6 ± 3.9 ** | 233.8 ± 2.7 | 217.8 ± 4.4 ** | - | |
| F1-terminal | 349.9 ± 6.5** | 325.5 ± 4.1 ** | 311.9 ± 3.8 ** | - | |
| Right Ovary Absolute (mg) | F0 | 65.6 ± 3.9** | 48.2 ± 3.7* | 61.3 ± 3.6 * | 39.7 ± 3.4 ** |
| F1-interim | 55.2 ± 1.4* | 52.8 ± 2.2 | 47.8 ± 1.6* | - | |
| F1-terminal | 84.3 ± 4.3 | 73.6 ± 3.2 | 70.8 ± 3.3 | - | |
| Right Ovary Relative (mg/g) | F0 | 0.19 ± 0.01 | 0.17 ± 0.01 | 0.20 ± 0.01 | 0.15 ± 0.01* |
| F1-interim | 0.23 ± 0.01 | 0.22 ± 0.01 | 0.22 ± 0.01 | - | |
| F1-terminal | 0.24 ± 0.01 | 0.23 ± 0.01 | 0.23 ± 0.01 | - | |
| Left Ovary Absolute (mg) | F0 | 66.1 ± 4.4** | 53.2 ± 4.8 | 66.0 ± 4.5 | 38.9 ± 3.7** |
| F1-interim | 58.5 ± 1.5* | 54.8 ± 2.4 | 51.1 ± 2.7 | - | |
| F1-terminal | 83.4 ± 4.1* | 75.4 ± 4.7 | 71.1 ± 2.4 | - | |
| Left Ovary Relative(mg/g) | F0 | 0.20 ± 0.01 | 0.18 ± 0.02 | 0.22 ± 0.01 | 0.14 ± 0.01* |
| F1-interim | 0.24 ± 0.01 | 0.23 ± 0.01 | 0.23 ± 0.01 | - | |
| F1-terminal | 0.24 ± 0.01 | 0.23 ± 0.01 | 0.23 ± 0.01 | - |
Statistical analysis for F0 data performed by Jonckheere trend (0 ppm column) and Williams or Dunnett pairwise tests. Statistical analysis for F1 animals with littermates was performed by mixed models with a random litter effect and a Dunnett-Hsu adjustment for both trend and pairwise analyses.
p< 0.05
p<0.01
“-“ = no animals examined due to early removal
F0 n = 12–22/group; F1-interim n = 27–58/group; F1-terminal n = 23–34/group
Table 10:
Incidences of selected histopathologic lesions of the F0, F1c Interim, and F1c terminal rats after 4-MI exposure
| Generation | 0 ppm | 750 ppm | 2500 ppm | 5000 ppm | |
|---|---|---|---|---|---|
| MALES | |||||
| Number of animals examined (litters) | F0 | (23) | (23) | (23) | (23) |
| F1-interim | 49 (18) | 56 (22) | 20 (8) | - | |
| F1-terminal | 40 (18) | 44 (22) | 40 (15) | - | |
| Prostate, Ventral Lobe – Atrophy | F0 | 0% ** | 9 [1.0]a 39%** |
20 [1.1] 87%** |
23 [2.4] 100%** |
| F1-interim | 4 [1.0] 8%** |
25 [1.0] 45%** |
17 [1.2] 85%** |
- | |
| F1-terminal | 4 [1.0] 10%** |
10 [1.0] 23% |
35 [1.5] 88%** |
- | |
| Testis – Degeneration | F0 | 1 [1.0] 5%** |
0% | 4 [1.8] 17% |
8 [1.6] 35%* |
| F1-interim | 4 [1.3] 8% |
6 [1.0] 11% |
1 [2.0] 5% |
- | |
| F1- terminal | 2 [1.0] 5% |
5 [2.0] 11% |
5 [1.2] 13% |
- | |
| Testis – Spermatid Retention | F0 | 2 [1.0] 9%* |
3 [1.0] 13% |
1 [1.0] 4% |
8 [1.3] 35%* |
| F1- interim | 0% | 3 [1.0] 5% |
4 [1.0] 20% |
- | |
| F1- terminal |
0% |
5 [1.2] 11% |
4 [1.0] 10% |
- | |
| Epididymis – Exfoliated Germ Cells | F0 | 1 [1.0] 5%** |
0 | 3 [1.7] 13% |
7 [1.3] 30%* |
| F1- interim | 3 [1.3] 6% |
5 [1.2] 9% |
4 [1.3] 20% |
- | |
| F1- terminal | 0% | 5 [1.2] 11% |
4 [1.3] 10% |
- | |
| Liver, Centrilobular Hepatocyte – Vacuolation | F0 | 0% * | NE | 0% | 19 [1.9] 83%** |
| F1- interim | 0% | NE | 0% | - | |
| F1- terminal | 0% | NE | 1 [1.0] 3% |
- | |
| FEMALES | |||||
| Number of animals examined (litters) | F0 | 23 | NE | 23 | 23 |
| F1-interim | 47 | 58 | 27 | ||
| F1-terminal | 40 | 43 | 40 | ||
| Kidney – Mineral | F0 | 1 [1.0] 4% |
NE | 2 [1.0] 9% |
1 [2.0] 4% |
| F1- interim | 9 [1.0] 19%** |
44 [1.3] 76%** |
20 [1.2] 74%** |
- | |
| F1- terminal | 8 [1.0] 20%** |
21 [1.1] 49%* |
27 [1.1] 68%** |
- |
Incidence with [avg. severity score] and percent incidence; Severity scores: 1=Minimal, 2=Mild, 3=Moderate, 4=Severe. Average severity scores were not used in statistical significance calculations.
Statistical analysis for the F0 animals was performed using the Poly-3 trend (0 ppm column) and pairwise statistics. Statistical analysis for F1 animals was performed using a Cochran-Armitage test with a poly-3 adjustment for survival and a Rao-Scott modification for litter effect. All tests were one-sided.
p< 0.05
p<0.01
NE = not examined (read-down) “-“ = no animals examined due to early removal
For continuous endpoints, extreme values were identified by the outlier test of Dixon and Massey (1957) for small samples (n≤20), and Tukey’s outer fences method (Tukey, 1977) based on three interquartile range intervals, for large samples (n>20). To identify outliers for continuous endpoints with litter effects, all observations across dose groups were fit to a linear mixed effects model with random litter effect, and the residuals were tested by dose group for outliers using Tukey’s outer fences method (Tukey, 1977). All flagged outliers were examined by NTP personnel, and implausible values were eliminated from the final analyses. Discrete count endpoints were manually reviewed for unusual values.
3. RESULTS
Due to the large amount of information collected under this study design, only the key findings are summarized here. All data including the dose range finding study and supplemental information cited in this manuscript are available via the National Toxicology Program’s Chemical Effects in Biological Systems (CEBS) database: https://doi.org/10.22427/NTP-DATA-002-01511-0000-0000-0
3.1. Survival, Food Consumption and Clinical Observations
Calculated doses (mg/kg) based on food consumption at various life stages are presented in Table 1. An average dose was calculated for the three gestational and lactational periods that produced the a, b, and c litters. For F0 animals, food consumption (g/kg body weight/day) of F0 females was up to 9% lower in the 5000 ppm group during pre-breed exposure and gestation of the a and b litters compared to the control group and sporadically lower in the 2500 ppm dose group at various stages of the study. For further details on food consumption see CEBS Tables I06, I08, I23.
Table 1:
Mean 4-MI intake (mg/kg/day) during the various phases of the RACB study
| 750 ppm | 2500 ppm | 5000 ppm | |
|---|---|---|---|
| F0 | |||
| Prior to Pairing (M, Study Days 0–14) | 47.9 | 144.6 | 260.1 |
| Prior to Pairing (F, Study Days 0–14) | 46.8 | 145.6 | 289.9 |
| Gestation (GD 1–21) a | 48.3 | 151.2 | 307.0 |
| Lactation (PND 1–4) a | 79.1 | 231.9 | 274.0 |
| Lactation (PND 1–13) b | 101.4 | 319.0 | - |
| F1c | |||
| Prior to Pairing (M) | 63.7 | 206.9 | - |
| Prior to Pairing (F) | 66.3 | 225.4 | - |
| Gestation (GD 1–21) a | 48.7 | 161.2 | - |
| Lactation (PND 1–4) a | 79.6 | 278.7 | - |
| Lactation (PND 1–13) b | 93.8 | 306.5 | - |
Average intake (mg/kg/day) of dams across the three breeding (A, B, C) periods
Intake of the dams (mg/kg/day) during lactation period of the C litter
“-“ = No 5000 ppm data due to early removal from the study
In the F0 generation, there was an exposure-related increase in number of females that were euthanized moribund or found dead in the mid and high dose groups respectively (Table 2). Several of these animals were removed around the time of parturition with dystocia or retained placentas/fetuses observed in most of them (Table 3). The F0 gestation length of the F1a litter was marginally longer (0.6 days) than controls, and the magnitude of the effect increased in the subsequent breeding (F1b) to statistically significant 1.2 days suggesting difficulties in parturition (CEBS Table R02). Due to this finding, the 5000 ppm F0 animals were not mated a third time. For the F1c females, there was a higher number of euthanized moribund animals in the 2500 ppm group, of which several animals displayed perturbed parturition and dystocia (Table 3). A higher number of females were observed to have a convulsion in the 5000 ppm group (39% in the F0) and 16% in the 2500 ppm group (F1) compared to controls (Supplemental Table 2). Animals exhibiting convulsions were not associated with moribundity or parturition or with a specific time period. Convulsions have previously been noted in NTP studies with 4-MI (NTP 2006).
Table 2:
Number of animals per removal reason during the 4-MI RACB study
| Generation/Sex | Removal | 0 ppm | 750 ppm | 2500 ppm | 5000 ppm |
|---|---|---|---|---|---|
| F0 | |||||
| Male | Start | 23 | 23 | 23 | 23 |
| Removed: Partner Loss | 0 | 0 | 3 | 2 | |
| Terminal | 23 | 23 | 20 | 21 | |
| Female | Start | 23 | 23 | 23 | 23 |
| Moribund/Found Dead | 1 | 1 | 4 | 11 | |
| Terminal | 22 | 22 | 19 | 12 | |
| F1c | |||||
| Male | Weaned | 99 | 115 | 60 | -a |
| Removal: Otherb | 10 | 15 | 0 | ||
| Moribund/Found Dead | 0 | 0 | 1 | ||
| Interim (Nonparental) | 49 (18)c | 56 (22) | 20 (8) | ||
| Terminal (Parental) | 40 (18) | 44 (22) | 39 (15) | ||
| Female | Weaned | 99 | 117 | 67 | -a |
| Removal: Otherb | 12 | 15 | 0 | ||
| Moribund/Found Dead | 3 | 5 | 8 | ||
| Interim (Nonparental) | 47 (17) | 58 (22) | 27 (12) | ||
| Terminal (Parental) | 37 (19) | 39 (22) | 32 (15) | ||
| F2c | |||||
| Male | Terminal (PND 28) | 107 | 132 | 69 | |
| Female | Terminal (PND 28) | 98 | 132 | 76 |
The 5000 ppm F0 group was removed from the study
Removed and not evaluated
Number of Litters in parentheses
Table 3:
Incidence of perturbed parturition across the three pairings per generation a
| Generation | 0 ppm | 750 ppm | 2500 ppm | 5000 ppm | |
|---|---|---|---|---|---|
| Dystocia | F0 | 0 | 0 | 1 | 5 |
| F1 | 0 | 0 | 4 | - | |
|
| |||||
| Retained Fetus/Placentas | F0 | 0 | 0 | 1 | 1 |
| F1 | 1 | 1 | 1 | 0 | |
|
| |||||
| Total Perturbed Parturition b | F0 | 0 | 0 | 2 | 6 |
| F1 | 1 | 1 | 5 | - | |
Number of females displaying evidence of mating across all three pairings: F0 n = 66, 68, 60, 21; F1 n = 109, 122, 104
Incidence of animals displaying dystocia or retained fetus/placentas
3.2. Body Weights
Body weights of the F0 males were approximately 10% and 13% lower in the 2500 ppm and 5000 ppm groups, respectively, compared to the controls throughout most of the study (Figure 2a). Over study days 0–14, F0 females also showed dose-related weight decreases (Figure 2b).
Figure 2.
Mean body weights versus day for male and female rats at different developmental stages for control and dosed animals. An average of the litter mean weights are shown in (e)-(h). Due to apparent moribundity/mortality, F0 females in the 5000 ppm dose group were removed prior to mating for the C litter and are not shown in (c)-(h).
For the F0 production of the F1a and F1b litters, dam body weights on GD 21 were 4–5%, 14–15%, and 19–25% lower in the 750, 2500, and 5000 ppm groups respectively (CEBS #I04). Body weights of the F0 dam during gestation of the F1c litter showed similar dose-related decreases relative to controls (Figure 2c). In the 2500 ppm and 5000 ppm groups, reduced dam body weights during gestation is likely partially related to smaller litter sizes.
F0 maternal weights on lactational day 4 were also reduced by similar magnitudes compared to controls (5–7% for 750 ppm a, b, c litters; 13–14% for 2500 ppm a, b, c litters, 25–31% for 5000 ppm a, b litters). These trends continued over lactational days 1–28 for the F0 females during F1c lactation (Figure 2d). Similar patterns were exhibited in the 750 and 2500 ppm groups compared to controls during the F1c production of the F2 litters.
The F1c male and female pup body weights were lower than controls in a dose-dependent manner over PND 1–28 (Figure 2e and f), which was up to 20% lower in the 2500 ppm by the end of lactation. A similar magnitude of effect occurred in the F2c pups. Post-weaning weights of the F1c were ~10–15% lower than controls in males and females in the 2500 ppm group ((Figure 2g and h; note that the 5000 ppm group was removed).
3.3. Reproductive Performance
The F0 5000 ppm group displayed decreases in percent mated females/pair, percent littered/pair and percent littered that mated relative to control (Table 4). For further information on each breeding (a, b or c), see CEBS Table# R21. Exposure to 4-MI did not appear to affect mating and littering of the F0 and F1c 750 and 2500 ppm groups relative to controls. When reproductive performance for each pair across the matings was evaluated for F0 animals, there was a decrease in number of litterings/pair in the F0 5000 ppm group relative to controls, and no evidence of a dose-related effect on litters/pair at 2500 ppm or lower for the F0 and F1c pairings (Table 4, CEBS Table #R21).
Table 4:
Average reproductive performance of the three pairings pairs (A, B, C) of the F0 and F1c following 4-MI exposure, including crossover mating of exposed F0 males or F0 females with naïve partners
| Generation | 0 ppm | 750 ppm | 2500 ppm | 5000 ppm | |
|---|---|---|---|---|---|
| Mated/Pair | |||||
| Average of A, B, C Pairings | F0 | 97.0 | 98.6% | 96.8% | 48.1%a |
| Average of A, B, C Pairings | F1c | 92.4% | 97.0% | 94.0% | - |
| 4-MI Male × Naïve Female b | F0 | 87.0%** | - | 80.0% | 33.3%** |
| 4-MI Female × Naïve Male b | F0 | 68.2% | - | 73.7% | - |
| Littered/Pair | |||||
| Average of A, B, C Pairings | F0 | 88.2% | 97.1% | 83.4% | 32.4%a |
| Average of A, B, C Pairings | F1c | 81.8% | 84.3% | 78.8% | |
| 4-MI Male × Naïve Female b | F0 | 69.6%** | - | 75.0% | 28.6%* |
| 4-MI Female × Naïve Male b | F0 | 54.5% | - | 52.6% | - |
| Littered/Mated | |||||
| Average of A, B, C Pairings | F0 | 90.8% | 98.6% | 86.2% | 75% a |
| Average of A, B, C Pairings | F1c | 87.9% | 86.9% | 83.7% | |
| 4-MI Male × Naïve Female b | F0 | 80.0% | 93.8% | 85.7% | |
| 4-MI Female × Naïve Male b | F0 | 80.0% | - | 71.4% | - |
| Number of Litters/Pair | F0 | 2.6 ± 0.2 | 2.9 ± 0.1 | 2.4 ± 0.2 | 0.1 ± 0.1 |
| F1c | 2.5 ± 0.1 | 2.5 ± 0.1 | 2.3 ± 0.1 |
Paired only two times (A and B); removed from study prior to third pairing (C)
Analysis of crossover mating was performed by Cochran-Armitage (trend) and Fisher Exact (pairwise) 2-sided tests. Trend results appear in the 0 ppm column.
p < 0.05
p < 0.01
Due to the reduction in mating and littering, a cross-over mating of 4-MI exposed or control males to naïve females, as well as 4-MI exposed or control females to naïve males was conducted following the F1c littering to determine the effects in each exposed sex. In the cross-over mating, there was a significant reduction in mated/pair, but not littered/mated in the 5000 ppm exposed males mating with naïve females (Table 4). All females that did not deliver were found to be non-pregnant suggesting an effect on fertility in males. There was no evidence of a exposure-related effect on reproductive performance of 2500 ppm females mated with naïve males or 2500 ppm males mated with naïve females (Table 4). The 5000 ppm females were not included in the cross-over mating due to moribundity associated with parturition.
3.4. Litter Parameters
There was a statistically significant trend of smaller total litter size with increasing dose in the F0 and F1c matings compared to the controls (Table 5). The number of total pups and live pups per litter on PND 0 (resulting from F0 matings) was significantly lower in the 2500 and 5000 ppm groups relative to controls. A similar statistically significant decrease in litter size was observed in the F1c 2500 ppm group relative to controls. Consistent with this finding was a decrease in the average number of live pups produced per pair across the dose groups relative to controls (Table 5). In the cross-over mating of the F0 generation, there were no statistically significant decreases in PND 0 total and live litter size in the 4-MI exposed males mated with naïve females or the 4-MI exposed females mated with naïve males (Table 5).
Pup survival (PND 1–4) was lower in the F0 5000 ppm group and the F1c 2500 ppm groups relative to controls (Table 5). In the crossover mating of the 4-MI exposed males with naïve females, significant dose-related effects on survival (PND 1–4) were not observed, although there was a decrease in survival when 2500 ppm 4-MI exposed females were mated with naive males. Survival of pups from PND 5 – 28 did not show significant exposure-related effects in the F0 and F1c 750 and 2500 groups relative to controls (Table 5).
3.5. Markers of Development
There was no consistent pattern in changes with anogenital distance of male and female pups, a marker of altered endocrine activity (CEBS Table #R04). A small number of pups were observed to have areolae or nipples in the F1c and F2c generations in the 2500 ppm group (Table 6) with a historical incidence of 2/382 pups and litter incidence of 2/80 litters. There was a significant trend toward delayed day of testicular descent in the F1c and F2c generation, and the delay in the 2500 ppm F2c males was significant by pairwise comparison to the controls (Table 6).
Statistically significant delays in preputial separation (PPS) and vaginal opening (VO), markers of pubertal development, were seen in male and female F1c offspring in the 750 and 2500 ppm groups relative to controls (Table 6). Male body weight at PPS attainment was statistically significantly higher (5%) in the 750 ppm group, with the 2500 ppm group similar to controls. Female body weight at attainment of VO was significantly higher (~10%) in the 750 and 2500 ppm groups. Body weight at weaning (PND 28) was used as a covariate in the analysis of PPS and VO since body weight is a known factor in pubertal progression. Weights at weaning were significantly reduced (10% or less) in dosed males and females compared to controls (CEBS Table #R16). After adjustment for weaning weight, PPS in the 750 and 2500 ppm groups was significantly delayed compared to controls (~2 days). Females exposed to 750 and 2500 ppm also displayed significant delays (3 and 6 days) in VO after adjustment for weaning weight.
3.6. Sperm Analysis and Estrous Cycle
F0 rats exposed to 5000 ppm displayed significantly lower sperm count per cauda epididymis and reduced count per mg cauda epididymis decreased sperm concentration per g cauda epididymis (25% and 14%, respectively), and sperm were significantly less motile than controls (Table 7). At the F1-interim time point, concentration per g cauda was significantly less than that of controls for the 750 ppm and 2500 ppm groups (9% and 12% respectively), with significantly decreased motility observed in the 2500 ppm group. However, for the F0 and F1-terminal time points, concentration per g cauda in rats exposed to 750 ppm and 2500 ppm was not significantly different from controls. Spermatid counts in the testis were not affected by exposure across F0 and F1c generations and time points (CEBS # R06). Extended diestrus was observed in the F0 5000 ppm group (p = 0.016), and there was a statistically significant increase noted in total cycle length at the 2500 ppm F1c females compared to controls (CEBS Table Vag Cyt and Supplemental Table 3).
3.7. Organ Weights
Organ weights were evaluated for the F0 (~PND 230), F1c-interim (~PND 95), and F1c terminal (~PND 220) groups. At necropsy, F0 male body weights in exposed groups were lower (up to 10%) compared to respective controls (Table 8). The magnitude of dose-related decreases across generations and time points was fairly similar, with body weights 2–5% lower in the 750 ppm dose groups and 9–14% lower in the 2500 ppm groups relative to controls. Testis absolute weights were unaffected by 4-MI exposure. Absolute weights of the epididymides (right and left) were 7–9% lower compared to controls in the 2500 ppm groups across generations and time points, and were reduced 10–12% in in the F0 5000 ppm groups compared to controls (Table 8). However, relative epididymides weights to body weight were not lower.
Across the F0 and F1c generations and timepoints, absolute dorsolateral prostate weights were often significantly lower compared to controls (6–19% in the 750 ppm groups, 15–22% in the 2500 ppm groups and 30% in the F0 5000 ppm group), with the F0 males displaying the largest decreases relative to controls (Table 8). Relative dorsolateral weights were also significantly lower in the F0 generation, but not the F1c generation. Absolute and relative ventral prostate weights were often significantly lower in the 750 ppm groups (3–20%), in the 2500 ppm groups (20–35%), and in the F0 5000 ppm group (45%) with the F1c-interim males being the most sensitive in terms of magnitude of difference (Table 8).
Absolute seminal vesicle weights were generally 4–15% lower in the 750 ppm groups, 16–22% in the 2500 ppm groups, and 36% in the F0 5000 ppm group, while relative seminal vesicle weights were mostly significantly reduced in the F0 generation (Table 8). Dosed F0 males showed the largest decreases in seminal vesicle weights on absolute and relative scales compared to controls. Absolute LABC weights were significantly decreased in F0 2500 and 5000 ppm exposure groups compared to controls with decreasing trends observed in the F1c timepoints. However, when LABC was adjusted for body weight, significant differences between dosed groups and controls were not observed. In general, the magnitude of difference between dosed groups and controls across the reproductive organs was fairly consistent and did not increase with exposure length within the F1 generation (age PND 90 for F1c-interim cohort vs PND 220 for terminal cohort).
For females, there was a statistically significant decrease in terminal body weight in all dose-groups in the F0, F1-interim, and F1-terminal groups relative to controls. Weights were 4–14%, 10–11%, and 19% lower in the 750, 2500, and 5000 ppm groups. There was a statistically significant decrease in the right and left ovarian weight (absolute and relative) in the 5000 ppm group and lower absolute weights of the right ovary in all dose groups compared to controls in the F0, and 2500 ppm groups in the F1-interim groups (Table 9).
There were no consistent effects attributable to 4-MI in the other males and females’ non-reproductive tissues examined. Changes in liver and kidney weights in male and female rats appeared secondary to body weight reductions (Supplemental Tables 4 and 5)
3.8. Histopathology
Histopathologic lesions related to treatment in male rats administered 4-MI included prostatic atrophy, testicular degeneration, testicular spermatid retention, exfoliated germ cells in the epididymis, and hepatocellular vacuolation. Histopathologic lesions related to treatment in female rats administered 4-MI included mineralization in the kidney (Table 10).
Treatment-related lesions in the prostate gland included atrophy of the ventral lobes in the F0 and F1 generations without exacerbation (increase in incidence and/or severity) in the F1 generation (Table 10). The incidences of atrophy of the ventral lobes of the prostate gland were significantly greater in the 750 ppm or greater F0 males, the 750 ppm or 2500 ppm F1c-interim males, and the 2500 ppm F1c-terminal males. Prostate atrophy was generally minimal to mild severity except in the 5000 ppm group, which was generally of mild to moderate severity in the F0 generation.
The ventral prostate of the treated rats with atrophy appeared at low magnification, to be smaller with lighter staining (Fig. 3B) than the normal ventral prostate of the controls (Fig. 3A). Microscopically, the luminal secretions within the acini of the atrophic glands were less abundant, and more often absent in 4-MI exposed males, in comparison to the normal prostate glands of the controls. When present, the luminal content in dosed- males was pale and flocculent (Fig. 3D, F) compared the more densely hyalinized secretory material normally observed in control animals (Fig. 3C, E). The epithelium lining the acini was more often very flattened in the atrophic prostates of 4-MI exposed males(Fig. 3F) by comparison to the control animals in which there were more acini lined by low cuboidal to tall columnar epithelium (Fig. 3E). Fibrous thickening of the interstitium was occasionally observed in atrophic prostates of 4-MI exposed males.
Figure 3.
Ventral prostate gland histopathology for F0 control and 5000 ppm 4-MI exposed males. Compared to the prostate from the control male, the prostate from the 4-MI treated male is smaller and the luminal contents are paler staining (compare control in A to F0 5000 ppm in B, 1x magnification). The luminal secretions in 4-MI treated prostates were often decreased, pale and occasionally flocculent (F0 5000 ppm in D, 10x; F, 40x) compared to the more densely eosinophilic secretory material in control prostates (C, 10x; E, 40x). Also, the epithelium lining the prostates of 4-MI treated males was often flattened and attenuated (F, 40x) when compared to normal control prostates lined by cuboidal to columnar epithelium (E, 40x). H&E.
Exposure-related lesions in the testis included degeneration and spermatid retention, which were statistically significantly increased in the F0 animals dosed with 5000 ppm 4-MI when compared with control males (Table 10).
Microscopically, testicular degeneration was seen as one or more degenerative changes in the testis that typically included focal or segmental loss of germ cells from Sertoli cell cytoplasm, disorganization of germ cell layers, focal tubular vacuolation, and/or evidence of depletion/degeneration of elongating and/or round spermatids. Less frequently, degeneration involved abnormal residual bodies (in stages VII, VIII, and IX), apoptosis of individual germ cells in certain stages, or multinucleated spermatids (Fig. 4A). In low numbers of stage VII or VIII tubules in all dose groups, apoptosis of occasional round spermatids or pachytene spermatocytes was observed as a degenerative change. In some of these tubules and others, there were decreased numbers of round and/or elongating spermatids within the tubule (Fig. 4A). Degeneration in some cases centered on tubules in the region around the rete testis, accounting for and excluding the tubuli recti. At the lowest dose, 750 ppm, degeneration was diagnosed in ~11% of the F1 generation males (both F1c-interim and terminal) but in no F0 animals. In the F1 animals in the 750 ppm dose group, the germ cell dropout within an individual tubule was often more focal than the more extensive dropout seen at higher doses. Also, at 750 ppm in the F1 animals, degeneration more frequently involved smaller vacuoles that likely represented missing single cells and depletion of elongating and/or round spermatids along with decreased tubular diameter.
Figure 4.
Histopathology of the germinal epithelium in the testis of F0 control and 5000 ppm 4-MI males (A, 20x, PAS/H). In an early-stage tubule (E), there is a focal area of germ cell dropout (marked with an oval), as well as apoptosis of germ cells (arrows). The mid-stage tubule (M) nearest the center of the image has depletion of elongating spermatids. A late-stage tubule (L) has depletion of elongating spermatids and spermatocytes, with decreased tubular diameter and disorganized germinal epithelium. Vacuoles are also present (*). The cauda of the corresponding epididymis (B) for the testis in panel A contains exfoliated germ cells and debris from the testis (40x, H&E). A stage XI tubule from an F1 generation male dosed with 2500 ppm 4-MI shows a cluster of retained mature spermatids at the luminal surface of the germinal epithelium (C, 40x). The mature spermatids (step 19 spermatids) should be released during stage VIII, and therefore should not be present in a stage XI tubule (step 11 spermatids), as seen in a control F0 male (D, 40x).
In addition to testicular degeneration, spermatid retention was also noted in the testis in all male exposure groups, as well as in 2 F0 control males (Table 10). As with degeneration, the incidences of spermatid retention were significantly increased in the F0 animals dosed with 5000 ppm (35%). The incidence of spermatid retention at 750 ppm or greater was slightly increased in the F1c males (both F1c-interim and terminal cohorts) compared to the control groups, but these increases were not statistically significant. Spermatid retention was characterized microscopically by the presence of mature, step 19, spermatids at the luminal surface, within the germinal epithelium, or in the basal Sertoli cell cytoplasm in stage IX-XII tubules (compare Fig. 4C to control in Fig. 4D). Rarely, concurrent controls in this study had up to 5 tubules/testis with up to 3 (or more) retained spermatids. Therefore, spermatid retention was considered present in a testis if there were at least 8 tubules each containing over 3 retained spermatids. Spermatid retention could be evaluated when sufficient tubules had retained spermatids without the aforementioned degenerative changes. In most cases, spermatid retention was typically seen as a sole testicular lesion in the testis. In fewer cases, spermatid retention was seen in conjunction with other degenerative changes (i.e., vacuolation, focal cell dropout, disorganization, or depletion of round or elongating spermatids), either within the same tubule or in other tubules within the same testis. When retained spermatids were observed in a tubule with concomitant vacuolation, cell dropout, or other degenerative changes, degeneration was recorded and spermatid retention was not diagnosed separately. In 8 animals (5 F0, 1 F1c-interim, and 2 F1c-terminal), both degeneration and spermatid retention were diagnosed in the same testis. The testes from these animals contained >7 tubules with >3 retained spermatids per tubule (and no concomitant other degenerative changes in each tubule) within the same testis in which other tubules displayed sufficient degenerative changes to warrant the separate diagnosis of degeneration.
The incidences of exfoliated germ cells in the epididymis was significantly increased in the F0 animals dosed with 5000 ppm 4-MI compared to controls (Table 10). Microscopically, exfoliated germ cells in the epididymis were characterized by the presence of intact, round cells intermixed with cell debris and mature sperm within the epididymal duct lumina that exceeded the level of occasional sloughed cells in control animals (Fig. 4B). These minimal to mild occurrences of exfoliated germ cells in the epididymis were generally associated with minimal to moderate degeneration of the germinal epithelium of the testis.
The incidences of hepatocellular vacuolation of the centrilobular hepatocytes of the liver was significantly greater in the F0 male rats dosed with 5000 ppm 4-MI (83%), and this effect was not present in the F1 generation (Table 10). Hepatocellular vacuolation was consistent with microvesicular and macrovesicular fatty change (CEBS Supplemental Fig. 1), and has been previously seen and described in F344/N rats in prior NTP studies of 4-MI (14-week or 2-year; NTP 2004, NTP 2006).
Incidences of mineralization in the kidney of female rats were significantly increased in F1c females (both interim and terminal) administered 750 ppm or 2500 ppm (Table 10). Mineralization was primarily a lesion of the F1 females and not the F0 females, suggesting that the effect was specific to that generation. Microscopically, mineral was characterized by focal to multifocal variably sized deposits of mineral that were randomly distributed along the corticomedullary junction in female rats (compare CEBS Supp. Fig. 2B to control kidney in Supp. Fig. 2A). Mineralization at the corticomedullary junction in females is often seen as a background lesion and has been shown to be linked to diet in female rats (Frazier et al. 2012). The significance of renal mineralization in F1 females in this study is unknown.
There were no other indications of renal damage from 4-MI exposure. There was a low incidence of renal tubule necrosis, which was present in only three F0 females at 5000 ppm that were euthanized due to moribund condition. However, these three F0 females had dystocia and were euthanized due to moribundity. The renal tubule necrosis in these three animals was may have been associated with ischemic hypoxia to the kidneys owing to moribundity.
There were also statistically significant increases in primordial, antral and atretic follicles in the F0 5000 ppm group, and a significant increase in atretic follicles in the 750 ppm F0 group (CEBS Supplemental Table 3). However, since there was no discernible pattern noted, the toxicological significance of these data was not apparent.
4. DISCUSSION
The NTP evaluated the potential for 4-MI to induce reproductive and developmental toxicity in a RACB paradigm based on evidence from previous 14-week studies that implicated the male reproductive system as a potential target (NTP 2006; Chan et al., 2008). The current RACB study design enables a more in-depth evaluation of potential effects on reproductive performance across time and generations.
Overall, findings from this study indicate that 4-MI exposure is associated with reproductive toxicity at all dose levels tested as demonstrated by a reduction in mating (5000 ppm) and smaller litter sizes (750, 2500, 5000 ppm). These findings were associated with decreases in male reproductive organ weights, histological changes in male reproductive tissues, and changes to sperm parameters. In females, there were higher incidences in perturbed parturition in the 2500 and 5000 ppm F0 dams, coupled with delays in vaginal opening in 4-MI exposed F1c offspring in 750 and 2500 ppm groups. A NOAEL was not established in this study.
The reduced mating and littering in the 5000 ppm exposure group was apparent in the a and b littering for the F0 females. Due to the moribundity/mortality associated with pregnancy and perturbed parturition in the females, this exposure group was not continued on to produce F1c litters. In the 2500 ppm exposure group, there was no clear effect on fertility (e.g. mating, littering). However, reduced litter sizes were noted in the 2500 ppm group at PND 0 in the F0 and F1c matings, and in the 750 ppm group of the F1c mating.
There were some changes noted in sperm parameters with reductions in sperm counts, but possibly due to variability, these effects were not consistent across generations or timepoints except for reductions in sperm motility. Spermatid counts within the testis were not affected by exposure, but there were notable histopathological lesions within the testis and epididymis that were exposure related. Testicular degeneration was observed and incorporated a variety of changes. Degeneration (apoptosis) of pachytene spermatocytes and round spermatids in stage VII and VIII tubules was noted. This cell- and stage-specific finding is consistently demonstrated to be the most sensitive indicator of decreased testosterone in the testis (Russell and Clermont 1977). This particular finding was seen as a component of the germinal epithelium degeneration at doses as low as 750 ppm in the F1 generation, and at higher doses in the F0 generation. Prolonged or marked depletion of testosterone can lead to a general degeneration or depletion of elongating spermatids, which are present in most stages of spermatogenesis. Because the presence of elongating spermatids regulates seminiferous tubule fluid secretion by the Sertoli cell, depletion of elongating spermatids can then lead to contraction of tubular lumen size and diameter (Creasy 2008). Depletion of elongating spermatids, sometimes associated with decreased tubular diameter, was observed in this study as a component of degeneration at doses as low as 750 ppm in the F1 generation, or at higher doses in the F0 generation. Similarly, in the previous 13-week study of 4-MI (NTP, 2004), the incidence of testicular degeneration was significantly increased in the 5000 and 10000 ppm group males.
A second microscopic lesion observed in the testes in this study was spermatid retention. Spermatid retention is a functional disturbance in the process of spermiation, resulting in delayed spermiation. It can be directly test-article induced due to abnormalities in the Sertoli cell or result from hormonal disturbance (i.e., testosterone deficiency) (Creasy et al. 2012). The mature spermatids should normally be released during stage VIII. The normal control testis may have a few retained spermatids at the lumen of stage IX or phagocytized at base of stage XII tubules, as seen in concurrent control animals in this study, but are typically very few in number of spermatids (1–2, rarely 3) and typically only in 1 or 2 tubules/testis (Creasy 2008). Spermiation (sperm release) during stage VIII is an androgen-dependent process. Spermatid retention is a subtle but important change often associated with abnormal sperm number, motility, or morphology, and potentially, decreased fertility (Creasy et al. 2012). In addition, percent motile sperm was significantly decreased in the 2500 and 5000 ppm groups. Spermatid retention in this study occurred both independently, and in the presence of other degenerative changes in the 4-MI exposed animals. There was no exacerbation of testicular degeneration or spermatid retention in subsequent generations following 4-MI exposure.
The histologic lesions of degeneration (apoptosis) of pachytene spermatocytes in stages VII and VIII and spermatid retention are suggestive of low testosterone owing to 4-MI exposure. In younger rats exposed to short duration of food restriction, similar lesions have been demonstrated, but were accompanied by greater body weight decrements than in this study of 4-MI. Rehm et al. (2008) demonstrated findings of degeneration (apoptosis) of pachytene spermatocytes, mainly in stage VII, and spermatid retention in young rats (6–12 weeks at study start) when subjected to food restriction for a short period of time (2 or 6 weeks). Food restriction for 2 weeks resulted in 15–18% lower terminal body weights, and for 6 weeks resulted in 22–27% lower terminal body weights, when compared to ad libitum fed animals (Rehm et al. 2008). In contrast, terminal male body weights decrements of animals dosed with 4-MI were generally < 10% and were not considered to be the cause of the apoptosis in stages VII and VIII nor the spermatid retention. Terminal body weights in 4-MI exposed males were only 2–5% lower at 750 ppm, 9–14% lower at 2500 ppm, and 10% lower at 5000 ppm. Moreover, in older Sprague Dawley rats subjected to food restriction over an extended period of time, testis histopathology and testis and epididymis weights have been demonstrated to not generally be affected by up to a 30% reduction in terminal body weight (Chapin et al. 1993).
A corresponding epididymal lesion to degeneration and depletion of germ cells in the testis was seen as the presence of exfoliated germ cells and debris in the duct of the epididymis following 4-MI exposure. Exfoliated germ cells were present in 30% of the animals at 5000 ppm, which corresponded to the 35% of the males diagnosed with germinal epithelium degeneration in the testis. The presence of sloughed germ cells in the duct lumen of the epididymis is a common sequela of spermatogenic disruption in the testis.
Several male reproductive organ weights were reduced, with the prostate being the most sensitive reproductive tissue affected. Ventral prostate atrophy was noted in all exposure groups in both F0 and F1 generations. This finding was not exacerbated in the F1 generation even though exposure initiated in utero. The finding of prostate atrophy, which was most apparent in the ventral lobes, corresponded to statistically significant decreases in absolute and relative prostate weights. While prostate and seminal vesicle weights have previously been shown to decrease with lower body weight gains of 10% or more compared to controls (Chapin et al. 1993), the prostate weight reductions in 4-MI exposed animals exceeded losses expected due to body weight reductions, which were minimal in this study. In the current 4-MI study, F0 and F1 animals dosed with 750 ppm 4-MI had only 2–5% lower terminal body weights compared to controls, yet had 4–20% lower ventral prostate weights compared to controls. F0 or F1 animals dosed with 2500 ppm had 9–14% lower terminal body weights than controls, yet had 20–36% lower ventral prostate weights. Atrophy of the ventral prostate was previously reported in an NTP study as a compound related effect following 13-weeks of 4-MI exposure at the higher concentrations of 5000 and 10000 ppm compared to those noted in the current study (NTP, 2004).
In addition to reproductive toxicity, there were also indications of developmental toxicity to the reproductive system following 4-MI exposure. There were delays in pubertal development with preputial separation and vaginal opening in the 750 and 2500 ppm males and females delayed respectively. Low body weight can lead to delayed pubertal development, but 4-MI effects were still apparent after adjusting for body weight at weaning. There was a low incidence of areolae/nipples (F1c and F2c) and delayed testis decent (F2c) in the male pups of the 2500 ppm group could be due to inhibition of androgen signaling. However, there were no indications of malformations to the male reproductive tract such as hypospadias, or malformed epididymides or testis.
Cross-over mating of exposed males with naïve females and exposed females with naïve males was conducted in the F0 generation to determine an affected sex (i.e. findings due to effects in males and/or females). The cross over with exposed females to naïve males was limited to one exposure (2500 ppm) due to the toxicity observed in the 5000 ppm group and there were no effects observed in the female cross-over consistent with findings in the previous breedings. However, in the male cross-over (exposed males × naïve females), there was a significant decrease in the number of matings and litterings (primarily due to the reduced matings) in the 5000 ppm group that was consistent with findings in the F0 A and B matings. This indicates an effect on the male ability to successfully mate.
Among non-reproductive tissues, the liver appeared to be a target of toxicity in males, with hepatocellular vacuolation, primarily of centrilobular hepatocytes at 5000 ppm. The hepatocellular vacuolation may be related to altered lipid metabolism and hepatic injury, and has been noted before in previous NTP 14-week and 2- year studies (NTP 2004, 2006). These effects may be attributed to 4-MI’s effects on cytochrome P450 as previously noted in the literature (Hargreaves et al., 1994, Zarn et al., 2003), which could impact testosterone clearance. Although we did not measure specific cytochrome P450 enzymes in this study, other indications of enzyme induction such as increased liver weights or hepatocyte hypertrophy were not observed. Furthermore, liver lesions were only observed in the top dose of 5000 ppm, while reproductive organ weight decreases were observed at 750 ppm.
There are multiple lines of evidence suggesting that exposure to 4-MI may be inhibiting steroidogenesis in both males, and females. In 4-MI exposed females, there were functional changes of the reproductive system including dystocia along with delay in VO. Structurally related azole fungicides including ketoconazole, epoxiconazole and propiconazole have been shown to cause adverse effects similar to those noted in this study and include dystocia, increased litter loss, fetal death at birth and/or reduced litter size (Chambers et al., 2014). Estrogens are important for maintaining pregnancy, and it has previously been shown that CYP-450 inhibition by azoles in dams results in inhibiting steroidogenesis thereby causing placental/fetal disruption (Fisher et al., 1998; Chambers et al., 2014).
The steroidogenesis-related effects in males was supported by histopathological findings including ventral prostate atrophy and lower reproductive organ weights, as well as histopathological findings of degeneration and spermatid retention in the testis and the presence of exfoliated germ cells in epididymis in addition to a delay in PPS. Certain changes observed microscopically in the testis in male rats treated with 4-MI in this study, specifically the apoptosis of spermatocytes in stage VII/VIII tubules as well as spermatid retention, are consistent with effects of low testosterone. The effects on reproductive tissues and developmental markers of male sex differentiation and maturity indicate inhibited androgen action either via decreased steroid production or inhibition of the androgen receptor (McIntyre et al., 2001, 2002, Bowman et al., 2003, Hotchkiss et al., 2004, Taxvig et.al., 2007, 2008). The inhibition of steroid hormone synthesis by 4-MI has been previously shown in the literature by structurally- related fungicides (Taxvig et al., 2007, 2008) and structurally related compounds such as azole fungicides have been reported to block testosterone synthesis in humans and rodents by inhibiting CYPs (Feldman et al., 1986, Hume et al., 1983, Taxvig et al., 2008, Chambers et al., 2014). Ketoconazole administration in humans has been shown to decrease serum testosterone without a change in gonadotropin levels, suggesting a direct effect on the testis (Shurmeyer and Nieschlag 1984). Similar in vivo and in vitro studies in rodents have demonstrated that ketoconazole and other imidazoles decrease testosterone production (Shurmeyer and Nieschlag 1984; Adams et al. 1998).
In the current study, we noted a LOAEL of 50–60 mg/kg/day. This value is lower than previous findings from NTP subchronic studies in which reproductive tissue weights were altered in male F344/N rats at 300+ mg/kg/day, and histopathology lesions in other tissues at 160+ mg/kg/day (NTP 2004). Some of these differences may be due to rat stocks and study design considerations. The current LOAEL is comparable to a previous NTP chronic study which noted an increase in alveolar/bronchiolar adenomas and alveolar/bronchiolar adenomas or carcinomas (combined) 40 and 80 mg/kg/day respectively in female mice, and non-neoplastic lesions in the liver, lung, heart, and pancreas of female rats at 60 mg/kg/day. An EFSA panel conducted in 2011, concluded that an intermediate dose of 625 mg 4-MI/kg diet, equivalent to 80 mg 4-MI/kg bw/day could be considered to be a NOAEL based on the lack of genotoxicity of 4-MI, also noting that alveolar/ bronchiolar neoplasms occured spontaneously at high incidence in B6C3F1 mice in the NTP chronic studies (EFSA 2011). Subsequently, Morita et al., (2016) estimated a tolerable daily intake of 0.8 mg/kg/day (48 mg in terms of 60 kg body weight per person), based on the NOAEL of 80 mg/kg/day evaluated by EFSA after applying a safety factor of 100. The FDA/CFSAN currently estimates the upper 90th percentile of dietary intake for a high-exposure scenario in the US population 2 years and older to be approximately 2.8 μg/kg bw/day, well-below levels at which toxicity was noted in these studies (Folmer et al., 2018).
In conclusion, this is the first study to show that 4-MI is a reproductive and developmental toxicant in male and female rats, which occurred at all concentrations tested. The Lowest Observed Adverse Effect Level for reproductive (based on prostate atrophy in males) and developmental toxicity (based on delays in preputial separation and vaginal opening) was 750 ppm, equivalent to approximately 50–60 mg/kg bw/day. Generally, the effects did not appear to worsen from F0 to the F1 generation or with prolonged exposure. Given the occurrence of 4-MI in foods containing caramel coloring including soft drinks, these reproductive and developmental data may aid in the risk assessment of this caramel coloring byproduct.
Supplementary Material
ACKNOWLEDGEMENTS
This work was supported by the Intramural Research Program of the National Institute of Environmental Health Sciences/National Institutes of Health. The authors thank Dr. Jack Bishop for his involvement with the study design; Dr. Gordon Flake, NTP for his involvement in the histopathology; Dr. Angela King-Herbert, NTP for animal support; Dr. Linda Kooistra, Charles River Laboratories, for serving as the Quality Assessment Pathologist and Pathology Working Group Coordinator; Dr Rochelle Tyl, RTI International for her involvement with the study conduct; Dr. Jennifer Fostel, NTP, for providing website support for the data; Dr. Grace Kissling for her help on the initial statistics; and Drs. Nigel Walker, Michelle Hooth, Vicki Sutherland, and Kembra Howdeshell for their critical reviews of the manuscript.
Funding:
This work was supported (in part) by the National Institutes of Health [NIEHS contract HHSN273201600011C].
List of Abbreviations
- 4-MI
4-methylimidazole
- CFSAN
Center for Food Safety and Applied Nutrition
- CYP
Cytochrome P450
- DART
Developmental and Reproductive Toxicology
- FDA
Food and Drug Administration
- IACUC
Institutional Animal Care and Use Committee
- IARC
International Agency for Research on Cancer
- HPLC
High performance Liquid Chromatography
- UV
ultraviolet
- JECFA
Joint Expert Committee on Food Additives (JECFA)
- NTP
National Toxicology Program
- QA
Quality Assurance
- RTI
Research Triangle Institute
Footnotes
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
REFERENCES
- Adams ML, Meyer ER, and Cicero TJ (1998). Imidazoles Suppress Rat Testosterone Secretion and Testicular Interstitial Fluid Formation In Vivo. Biology of Reproduction 59:248–254. [DOI] [PubMed] [Google Scholar]
- Agresti A. (1990) Categorical Data Analysis. John Wiley and Sons, New York. [Google Scholar]
- Bailer AJ, and Portier CJ (1988). Effects of treatment-induced mortality and tumor-induced mortality on tests for carcinogenicity in small samples. Biometrics. 44:417–431. [PubMed] [Google Scholar]
- Bowman CJ, Barlow NJ, Turner KJ, Wallace DG, Foster PM (2003). Effects of in utero exposure to finasteride on androgen-dependent reproductive development in the male rat. Toxicol Sci. 74(2):393–406 [DOI] [PubMed] [Google Scholar]
- Casal S, Fernandes JO, Oliveira MBPP, Ferreira MA. (2002). Gas chromatographic – mass spectrometric quantification of 4-(5-)-methylimidazole in roasted coffee after ion-pair extraction. J Chrom A. 976:285–291 [DOI] [PubMed] [Google Scholar]
- Chambers JE, Greim H, Kendall RJ, Segner H, Sharpe RM, Van Der Kraak G (2014). Human and ecological risk assessment of a crop protection chemical: a case study with the azole fungicide epoxiconazole. Crit Rev Toxicol.44(2):176–210. [DOI] [PubMed] [Google Scholar]
- Chan PC, Hill GD, Kissling GE, Nyska A. (2008). Toxicity and carcinogenicity studies of 4-methylimidazole in F344/N rats and B6C3F1 mice. Arch Toxicol. 82(1):45–53 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chapin RE, Gulati DK, Barnes LH, Teague JL (1993). The Effects of Feed Restriction on Reproductive Function in Sprague-Dawley Rats. Fundam.Appl Toxicol 20, 23–29. [DOI] [PubMed] [Google Scholar]
- Chapin RE, Sloane RA (1997). Reproductive assessment by continuous breeding: evolving study design and summaries of ninety studies. Environ Health Perspect. 105 Suppl 1:199–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Choi SJ, Jung MY. (2017). Simple and fast sample preparation followed by gas chromatography-tandem mass spectrometry (GC-MS/MS) for the analysis of 2- and 4-methylimidazole in cola and dark beer. J Food Sci.82:1044–1052. [DOI] [PubMed] [Google Scholar]
- Consumer Reports. (2014a). Consumer reports: caramel color: the health risk that may be in your soda. Last updated February 10, 2014. accessed 2019 April 16. https://www.consumerreports.org/cro/news/2014/01/caramel-colorthe-health-risk-that-may-be-in-your-soda/index.htm
- Consumer Reports. (2014b). Consumer reports: caramel color in pancake syrup may pose a health risk. Last updated April 28, 2014. accessed 2019 April 16. https://www.consumerreports.org/cro/news/2014/03/caramel-coloring-inpancake-syrup/index.htm
- Cox DR (1972). Regression models and life-tables. J. R. Stat. Soc. Series. B Stat. Methodol. 34:187–220. [Google Scholar]
- Creasy DM (2008). Endocrine disruption: a guidance document for histologic evaluation of endocrine and reproductive tests. Part 2: Male reproductive system. http://www.oecd.org/chemicalsafety/testing/40580640.pdf
- Creasy D, Bube A, de Rijk E, Kandori H, Kuwahara M, Masson R, Nolte T, Reams R, Regan K, Rehm S, Rogerson P, Whitney K. (2012). Proliferative and Nonproliferative Lesions of the Rat and Mouse Male Reproductive System. Toxicology Pathology 40: 40S. DOI: 10.1177/0192623312454337 [DOI] [PubMed] [Google Scholar]
- Cunha SC, Senra L, Cruz R, Casal S, Fernandes JO (2016). 4-Methylimidazole in soluble coffee and coffee substitutes. Food Control. 63:15–20. [Google Scholar]
- Datta S, Satten GA. (2005). Rank-sum tests for clustered data. J. Am. Stat. Assoc. 100:908–915. [Google Scholar]
- Dixon WJ, and Massey FJ Jr. (1957). Introduction to Statistical Analysis, 2nd ed., pp. 276–278, 412. McGraw-Hill Book Company, Inc., New York. [Google Scholar]
- Dunn OJ (1964). Multiple comparisons using rank sums. Technometrics. 6:241–252. [Google Scholar]
- Dunnett CW (1955). A multiple comparison procedure for comparing several treatments with a control. J. Am. Stat. Assoc. 50:1096–1121. [Google Scholar]
- [EFSA] European Food Safety Authority. (2011). Scientific opinion on the re-evaluation of caramel colors (E150a,b,c,d) as food additives, EFSA panel on food additives and nutrient sources added to food (ANS). EFSA J. 9(3):2004. [Google Scholar]
- European Commission. (2012). Commission regulation (EU) No. 231/2012 of 9 march 2012 laying down specifications for food additives listed in annexes II and III to regulation (EC) No 1333/2008 of the European Parliament and of the Council. Off J Eur Union. L 83:32–34. [Google Scholar]
- Feldman D. (1986). Ketoconazole and Other Imidazole Derivatives as Inhibitors of Steroidogenesis. Endocrine Reviews. 7(4): 409–420. [DOI] [PubMed] [Google Scholar]
- Fierens T, Van Holderbeke M, Cornelis C, Jacobs G, Sioen I, De Maeyer M, Vinkx C, Vanermen G. (2018). Caramel color and process contaminants in foods and beverages: part II - Occurrence data and exposure assessment of 2-acetyl-4-(1,2,3,4-tetrahydroxybutyl)imidazole (THI) and 4-methylimidazole (4-MEI) in Belgium. Food Chem. 255:372–379. [DOI] [PubMed] [Google Scholar]
- Folmer DE, Doell DL, Lee HS, Noonan GO, Carberry SE (2018). A U.S. population dietary exposure assessment for 4-methylimidazole (4-MEI) from foods containing caramel colour and from formation of 4-MEI through the thermal treatment of food. Food Addit Contam Part A Chem Anal Control Expo Risk Assess. 2018 Oct;35(10):1890–1910. [DOI] [PubMed] [Google Scholar]
- Frazier KS, Seely JC, Hard GC, Betton G, Burnett R, Nakatsuji S, Nishikawa A, Durchfeld-Meyer B, Bube A. 2012. Proliferative and nonproliferative lesions of the rat and mouse urinary system. Toxicol Pathol. 2012 Jun;40(4 Suppl):14S–86S. [DOI] [PubMed] [Google Scholar]
- Gart JJ, Chu KC, and Tarone RE (1979). Statistical issues in interpretation of chronic bioassay tests for carcinogenicity. J. Natl. Cancer Inst. 62:957–974. [PubMed] [Google Scholar]
- Girard DM, and Sager DB (1987). The use of Markov chains to detect subtle variation in reproductive cycling. Biometrics. 43:225–234. [PubMed] [Google Scholar]
- Greaves P. (2012). Male Genital Tract. In Histopathology of Preclinical Toxicity Studies, 4th ed., pp. 615–666. Elsevier, Amsterdam, the Netherlands. [Google Scholar]
- Hargreaves MB, Jones BC, Smith DA, Gescher A (1994). Inhibition of p nitrophenol hydroxylase in rat liver microsomesby small aromatic and heterocyclic molec ules. Drug Metab Dispos. 22(5):806–10. [PubMed] [Google Scholar]
- Hengel M, Shibamoto T. (2013). Carcinogenic 4 (5)-methylimidazole found in beverages, sauces, and caramel colors: chemical properties, analysis, and biological activities. J Agric Food Chem. 61:780–789. [DOI] [PubMed] [Google Scholar]
- Hommel G (1988) A stagewise rejective multiple test procedure based on a modified Bonferroni test, Biometrika. 75:383–386. [Google Scholar]
- Hotchkiss AK, Parks-Saldutti LG, Ostby JS, Lambright C, Furr J, Vandenbergh JG, Gray LE Jr.(2004). A mixture of the “antiandrogens” linuron and butyl benzyl phthalate alters sexual differentiation of the male rat in a cumulative fashion. Biol Reprod. 71(6):1852–61 [DOI] [PubMed] [Google Scholar]
- Houben GF and Penninks AH (1994). Immunotoxicity of the colour additive Caramel Colour III: a review of complicated issues in the safety evaluation of a food additive. Toxicology 91, 289–302. [DOI] [PubMed] [Google Scholar]
- Hume AL, & Kerkering TM (1983). Ketoconazole (Nizoral, Janssen Pharmaceutica, Inc.). Drug Intelligence & Clinical Pharmacy, 17(3), 169–174. [DOI] [PubMed] [Google Scholar]
- Hsu JC (1992). The factor analytic approach to simultaneous inference in the general linear model. J. Comp. Graph. Stat. 1:151–168. [Google Scholar]
- International Association on Research for Cancer (IARC) (2010). IARC Monographs on the Evaluation of Carcinogenic Risks of 4-methylimidazole to Humans. Last accessed April 16, 2019. https://monographs.iarc.fr/wp-content/uploads/2018/06/mono101-015.pdf
- Jacobs G, Voorspoels S, Vloemans P, Fierens T, Van Holderbeke M, Cornelis C, Sioen I, De Maeyer M, Vinkx C, Vanermen G. (2018). Caramel colour and process byproducts in foods and beverages: part I – Development of a UPLC-MS/MS isotope dilution method for determination of 2-acetyl-4-(1,2,3,4-tetrahydroxybutyl)imidazole (THI), 4-methylimidazole (4-MEI) and 2-methylimidazole (2-MEI). Food Chem. 255:348–356. [DOI] [PubMed] [Google Scholar]
- Jonckheere AR (1954). A distribution-free k-sample test against ordered alternatives. Biometrika. 41:133–145. [Google Scholar]
- Lamb JC IV (1985). Reproductive toxicity testing: evaluating and developing new testing systems. J Am Coll Toxicol 4:163–171 [Google Scholar]
- Lee S, Lee KG (2016). Analysis and risk assessment of 4 (5)- methylimidazolein brown colored foods and beverages. Food Add Contam. 9:59–65. [DOI] [PubMed] [Google Scholar]
- Licht BH, Shaw K, Smith C, Mendoza M, Orr J and Myers DV, (1992a). Development of specifications for caramel colours. Food and Chemical Toxicology 30(5), 383–387. [DOI] [PubMed] [Google Scholar]
- Licht BH, Shaw K, Smith C, Mendoza M, Orr J. and Myers DV, (1992b). Characterization of caramel colours I, II and III. Food and Chemical Toxicology 30(5), 375–382. [DOI] [PubMed] [Google Scholar]
- McIntyre BS, Barlow NJ, Foster PM (2001). Androgen-mediated development in male rat offspring exposed to flutamide in utero: permanence and correlation of early postnatal changes in anogenital distance and nipple retention with malformations in androgen-dependent tissues. Toxicol Sci. 62(2):236–49. [DOI] [PubMed] [Google Scholar]
- McIntyre BS, Barlow NJ, Foster PM (2002). Male rats exposed to linuron in utero exhibit permanent changes in anogenital distance, nipple retention, and epididymal malformations that result in subsequent testicular atrophy. Toxicol Sci. 65(1):62–70. [DOI] [PubMed] [Google Scholar]
- Morita T, Uneyama C.(2016). Genotoxicity assessment of 4 methylimidazole: regulatory perspectives.Genes Environ. 1;38:20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Morrissey RE, Lamb JC 4th, Morris RW, Chapin RE, Gulati DK, Heindel JJ (1989). Results and evaluations of 48 continuous breeding reproduction studies conducted in mice. Fundam Appl Toxicol. 13(4):747–77. [DOI] [PubMed] [Google Scholar]
- Mottier P, Mujahid C, Tarres A, Bressaire T, Stadler R. (2017). Process-induced formation of imidazoles in selected foods. Food Chem. 228:381–387. [DOI] [PubMed] [Google Scholar]
- National Toxicology Program (NTP) (2004). Toxicity Studies of 2- and 4-Methylimidazole (CAS Nos. 693–98-1 and 822–36-6) Administered in Feed to F344/N Rats and B6C3F1 Mice. Toxicity Report Series No. 67. NIH Publication No. 04–4409. U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health; Research Triangle Park, NC [Google Scholar]
- National Toxicology Program (NTP) (2006). NTP Technical Report on the Toxicology and Carcinogenesis Studies of 4-mMethylimidazole (CAS No. 822–36-6) in F344/N Rats and B6C3F1 Mice (Feed Studies). Technical Report Series No. 535. NIH Publication No. 05–4471. U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health; Research Triangle Park, NC [Google Scholar]
- Rao JNK, Scott AJ (1992). A simple method for the analysis of clustered binary data. Biometrics. 48:577–585. [PubMed] [Google Scholar]
- Rehm S, White TE, Zahalka EA, Stanislaus DJ, Boyce RW, Wier PJ. (2008). Effects of food restriction on testis and accessory sex glands in maturing rats. Toxicol Pathol 36(5): 687–694. [DOI] [PubMed] [Google Scholar]
- Russell LD, and Clermont Y. (1977). Degeneration of germ cells in normal, hypophysectomized and hormone treated hypophysectomized rats. Anat Rec 187, 347–366. [DOI] [PubMed] [Google Scholar]
- Schurmeyer T, and Nieschlag E. (1984). Effect of ketoconazole and other imidazole fungicides on testosterone biosynthesis. Acta Endocrinologica 105(2):275–280. [DOI] [PubMed] [Google Scholar]
- Schlee C, Markova M, Schrank J, Laplagne F, Schneider R, Lachenmeier DW. 2013. Determination of 2-methylimidazole, 4-methylimidazole and 2-acetyl-4-(1,2,3,4-tetrahydroxybutyl) imidazole in caramel colours and cola using LC/MS/MS. J Chom B. 927:223–226. [DOI] [PubMed] [Google Scholar]
- Shirley E. (1977). A non-parametric equivalent of Williams’ test for contrasting increasing dose levels of a treatment. Biometrics. 33:386–389. [PubMed] [Google Scholar]
- Tarone R. (1975). “Tests for Trend in Life Table Analysis.” Biometrika. 62:679–682. [Google Scholar]
- Taxvig C, Hass U, Axelstad M, Dalgaard M, Boberg J, Andeasen HR, Vinggaard AM (2007). Endocrine-disrupting_activities_in vivo_of the_fungicides_tebuconazole_and_epoxiconazole. 100(2):464–73. [DOI] [PubMed] [Google Scholar]
- Taxvig C, Vinggaard AM, Hass U, Axelstad M, Metzdorff S, Nellemann C (2008). Endocrine-disrupting properties in vivo of widely used azole fungicides Int J Androl. 31(2):170–7 [DOI] [PubMed] [Google Scholar]
- Tukey JW (1977). Easy summaries – numerical and graphical. In Exploratory Data Analysis, pp 43–44. Addison-Wesley, Reading, MA. [Google Scholar]
- Tzatzarakis M, Vakonaki E, Moti S, Alegakis A, Tsitsimpikou C, Tsakiris I, Goumenou M, Nosyrev AE, Rizoa AK, Tsatsakis AM. (2017). Quantification of 4-Methylimidazole in soft drinks, sauces and vinegars of Greek market using two liquid chromatography techniques. Food Chem Tox.107:565–571. [DOI] [PubMed] [Google Scholar]
- U.S. Food and Drug Administration (FDA) (1988). Good Laboratory Practice Regulations for Nonclinical Laboratory Studies. Code of Federal Regulations (CFR) (April 1), pp. 229–243. [Google Scholar]
- U.S. Food and Drug Administration (FDA) (1994). Good Laboratory Practice Regulations for Nonclinical Laboratory Studies. Code of Federal Regulations (CFR) (March 21), 56FR12300. [Google Scholar]
- Williams DA (1971). A test for differences between treatment means when several dose levels are compared with a zero dose control. Biometrics. 27:103–117. [PubMed] [Google Scholar]
- Williams DA (1972). The comparison of several dose levels with a zero dose control. Biometrics. 28: 519–531. [PubMed] [Google Scholar]
- Williams DA (1986). A note on Shirley’s nonparametric test for comparing several dose levels with a zero-dose control. Biometrics. 42:183–186. [PubMed] [Google Scholar]
- Yamaguchi H, Masuda T. (2011). Determination of 4 (5)- methylimidazolein soy sauce and other foods by LC-MS/ MS after solid-phase extraction. J Agric Food Chem. 59:9770–9775. [DOI] [PubMed] [Google Scholar]
- Zarn JA, Bruschweiler BJ and Schlatter JR (2003) Azole fungicides affect mammalian steroidogenesis by inhibiting sterol 14alpha-demthylase and aromatase. Environ Health Perspect 111:255–261. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.