Abstract
The potential adverse effects of resorcinol, delivered via drinking water at 0, 120, 360, 1000, and 3000 mg/L (palatability limit), were assessed in a regulatory guideline compliant two-generation reproduction study in Crl:CD(SD) rats. Expanded end points of thyroid gland (TG) function were added because of clinical case reports indicating human TG toxicity. Average daily resorcinol intake (mg/kg) at the 3000 mg/L concentration was 233 in F0 and F1 males, whereas in females it was 304 (premating/gestation) and 660 (lactation). No resorcinol ingestion-related clinical signs of toxicity were observed. Furthermore, neither gross morphologic anomalies nor effects on reproductive function or thyroid hormone levels were detectable. Body weight reductions occurred in 3000 mg/L F0 and F1 animals and were more pronounced in males. However, there was no evidence of either cumulative toxicity in the second generation or of enhanced sensitivity to resorcinol in pregnant/lactating females. Water intake was lower in 3000 mg/L rats of both generations and intermittently, to a lesser extent, at 1000 mg/L; however, concurrent feed intake and utilization were unaffected. Decreased TG follicular colloid content (conventional histopathology; confirmed by quantitative stereomicroscopy) in the 3000 mg/L F0 males was attributed to resorcinol but not considered adverse. The 3000 mg/L intake level appeared to have caused an adaptive thyroid response to a new homeostatic level with no adverse physiological consequences in either males (the more susceptible gender) or females. There were no differences in TG histology in F0 rats of either sex at 1000 mg/L. Thus, resorcinol intake at maximum palatability via a route and mode relevant to potential human exposures via contaminated drinking water at presently unknown environmental concentrations caused no detectable adverse effects on any reproduction or TG end points. The 3000 mg/L resorcinol exposure level was the no-observed-adverse-effect level (NOAEL) for parental systemic and offspring toxicity, while 1000 mg/L was the no-observed-effect level (NOEL).
Keywords
Resorcinol (CAS no. 108-46-31; 1,3-benzenediol; m-hydroxybenzene; m-dihydroxyphenol) is used in large quantities in the manufacture of tires and rubber goods, in the production of resins and resin adhesives, and in tanning and dyeing applications. Much smaller quantities are incorporated into pharmaceutical creams to treat acne and into cosmetic products, notably hair dyes. Resorcinol was first introduced into medicinal use in the 19th century to treat, amongst other conditions, ulcerating skin lesions. Thyroid gland (TG) toxicity, manifested as reversible goiters, has been reported in some patients upon prolonged treatment with ointments containing quite variable concentrations (ranging from 1% up to 50%) of resorcinol (Berthezene et al. 1973, Bull and Fraser 1950; Guinet, Tourniaire, and Peyrin 1967; Klem 1930; Pascher 1978; Strakosch 1943). Clinical case reports also revealed symptoms of hyperexcitability and CNS stimulation (e.g., Klem 1930; Pascher 1978; Bontemps et al. 1995; Duran et al. 2004). A more recent review of the toxicology of resorcinol that included occupational epidemiology studies and a risk assessment perspective found no evidence that exposure to resorcinol at levels greater than present in the general environment causes thyroid dysfunction (Lynch, Delzell, and Bechtel 2002). The authors of that review paper do not state any environmentally occurring concentrations of resorcinol in the human habitat.
In 2003, the European Union’s Committee of Toxicity, Ecotoxicity and the Environment (CSTEE) concurred with the conclusions of two reports that had placed resorcinol on a list of chemicals with endocrine disruption potential (CSTEE 2003). Resorcinol was included in that category because of its well-documented, reversible goitrogenic action on the human TG upon prolonged application of resorcinol ointments in patients with ulcerating skin disorders. Those side effects have been reported in numerous clinical case reports published over the span of several decades.
Several experimental animal studies have shown that thyrotoxic effects could not be induced by conventional dosing methods such as subcutaneous (SC) injection or oral bolus (original references in Doniach and Logothetopoulos 1953). Even high-dose gavage administrations in lifetime carcinogenesis bioas-says in mice and rats revealed no adverse effects on the TG (NTP 1992; see below). The only notable symptoms elicited by high bolus doses were hyperexcitability and transient signs of central nervous system (CNS) stimulation, which subsided soon after dosing (Doniach and Logothetopoulos 1953; Merker et al. 1982; NTP 1992). In contrast, when resorcinol was dissolved in oil and injected SC to rats twice daily for variable lengths of time up to 69 days, TG histology showed goitrogenic effects and hyperplasia. Other experiments with resorcinol diacetate, a slow release formulation, administered to rats also revealed adverse TG effects (Doniach and Logothetopoulos 1953). These earlier studies suggest that rapid metabolism occurred in rats and that the pharmaco(toxico)kinetics of resorcinol needed to be experimentally altered in order to allow manifestation of TG toxicity (Welsch, 2008).
The most comprehensive set of subchronic and chronic resorcinol toxicity data derives from a series of studies sponsored by the US National Toxicology Program (NTP). The study design included comprehensive mutagenicity testing and bioassays in two rodent species that were conducted in the early 1980s. The data reveal no evidence of carcinogenicity or TG-specific toxicity, even at doses up to 520 mg/kg/day in rats or 450 mg/kg/day in mice for 13 weeks or 150–225 mg/kg/day (5 days/week) for 2 years in rats and mice (NTP 1992). Some unexplained gender-specific adrenal gland fresh weight effects occurred in the NTP 13-week studies, without clear dose relationships and in opposite directions in rats and mice (NTP 1992). Absolute and relative adrenal gland weights were significantly increased in all surviving male rat dosed groups. In contrast, significant decreases in absolute and relative adrenal gland weights occurred in several of the male mice dosing groups.
There is unequivocal evidence that resorcinol is readily absorbed from the gastrointestinal tract of rats. When radiolabeled resorcinol was administered as a single oral bolus dose to Fischer 344 rats at 112 and 225 mg/kg, at least 90% of the label associated with the 112 mg/kg dose was excreted in the urine within 24 h. Only trace amounts were detected in major tissues (Kim and Matthews 1987). These data show that the chemical undergoes rapid and efficient metabolism, and disposition occurs primarily via urine. Glucuronide conjugates accounted for ~70% of the urinary metabolites, and there was also sulfate conjugation. Resorcinol elicited no developmental toxicity in conventional testing guideline compliant studies in either Sprague-Dawley rats (DiNardo et al. 1985) or New Zealand white rabbits (Spengler, Osterburg, and Korte 1986).
A review of the resorcinol toxicology database disclosed that a two-generation reproduction study was a major data gap. The preferred species for such evaluations compliant with testing guidelines (Organization for Economic Coperation and Development [OECD] and US Environmental Agency [US EPA]) is the rat. That species was therefore chosen for the study whose outcome is reported in this communication. An advantage in the context of potential TG toxicity of resorcinol is that rats are more sensitive than humans to thyroid-disrupting agents (Choksi et al. 2003; Jahnke et al. 2004), thus providing an increased margin of safety when extrapolating to potential human exposure. The selected route of administration was oral via drinking water, because this is an acceptable route and mode of administration in regulatory guideline compliant reproduction studies. In addition, drinking water intake of resorcinol is relevant to one of the anticipated potential routes of exposure of the human population at large via drinking water that might be contaminated. All testing guideline reproduction end points were recorded following ingestion of maximum palatable resorcinol concentrations via the drinking water to meet the hazard detection objectives of a testing guideline compliant regulatory study. There are at present neither published values of estimated daily human environmental exposure levels nor environmental concentrations in drinking water available. In light of the human clinical case reports of TG toxicity, special attention was paid to the TG in the study design. The gland’s micromorphology and functional status were assessed in parents and offspring. There were no detectable adverse effects, either on reproductive performance or on TG histopathology and hormonal parameters upon ingestion of resorcinol at the maximum palatable concentration.
The present study was conducted at WIL Research Laboratories, LLC, under contract to the Resorcinol Task Force, Ltd.
MATERIALS AND METHODS
Test Chemical
Resorcinol (CAS no. 108-46-31) was supplied as solid United States Pharmacopeia (USP) purity grade, off-white flakes by Indspec Chemical (Pittsburgh, PA). Identity of the neat chemical was verified by the supplier at the beginning and end of the study with proton nuclear magnetic resonance (NMR), carbon-13 NMR, and ultraviolet (UV) spectrometry. The purity was determined to be greater than 99.8% by capillary gas chromatography.
Resorcinol has excellent water solubility (Merck & Co. 2001) and is stable in water if protected from ambient light. Formulations were prepared by dissolving resorcinol in reverse-osmosis-purified municipal water (Ashland, OH) at the target concentrations of 120, 360, 1000, and 3000 mg/L on a weekly or twice-weekly basis. The solutions were analyzed periodically (first and third weeks of dispensation and approximately monthly thereafter) by a validated high-performance liquid chromatography (HPLC)/UV method. Least-squares regression was used to generate an equation relating the relative response at 254 nm of the resorcinol solution peak with that of the calibration standards (resorcinol at 5.0 to 350 mg/ml in deionized water). Results indicated that the drinking water formulations used in the present study were homogeneous, stable and within 10% of the target concentrations. Stability of resorcinol in drinking water for 29 days at room temperature and protected from light had previously been confirmed (Nemec 2003).
Animals and Husbandry
Virgin male and female Crl:CD(SD) rats (30 days old upon arrival; males 100 to 175 g; females 70 to 100 g) were obtained from Charles River Laboratories, Raleigh, NC. All rats were identified individually by ear tags after arrival. Selected F1 and F2 weanlings were also uniquely identified by ear tags on postnatal day (PND) 21 at the time of their selection. That number was formed by combining the number of their mother with a hyphenated offspring number, based on interdigital ink numberings made on PND 4. In addition, each study male and female (F0 and F1 parents and retained F2 offspring) were assigned a unique study number. All data collected during the course of the entire study were tracked by these numbers.
During the 14-day acclimation period, animals were observed twice daily for general health, mortality, and moribundity. Rats were initially gang-housed by sex for 3 days and thereafter individually housed, except during mating, in stainless steel suspended wire mesh cages. Feed (basal diet from PMI Nutrition International, Certified Rodent LabDiet 5002) was available ad libitum throughout the acclimation period and during the entire duration of the study.
Bottled reverse osmosis-purified (on-site) drinking water was available ad libitum to all animals throughout the acclimation period and to the control group during the entire study. Water bottles had ball-bearing sipper tubes to minimize spillage and thus enhance the accuracy of water consumption measurements. Bottles were wrapped with aluminum foil to protect resorcinol-containing solutions from exposure to ambient light in the animal room. Water bottles were changed and sanitized once per week. The animals were maintained on a 12-h light/12-h dark photoperiod at 71°F ± 5°F and 30% to 70% humidity. Lights were on from 0600 to 1800 h. Animals in this study were held, cared for, and used in compliance with the Animal Welfare Act (AWA 1966, as amended) and the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources et al. 1996). The animal facilities at WIL Research Laboratories, LLC, are accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC International).
Study Design
Animals in good general health were allocated to groups based on body weight stratification and randomized in a block design by a computer-generated program (WIL Toxicology Data Management System). For the F0 generation five groups of 30 males and 30 females were mated to yield at least 20 pregnant females per group at or near term. Upon study initiation all rats had drinking water available at all times that contained resorcinol at concentrations of 0, 120, 360, 1000, and 3000 mg/L. These resorcinol exposure concentrations for the full two-generation reproduction study were arrived at by preceding palatability assessments of short duration (2 weeks). The drinking water concentrations were raised up to 10000 mg/L, a level that was not accepted by the rats. By contrast the concentration of 3000 mg resorcinol/L was tolerable and was designated as the maximum palatable concentration, which was selected as the top concentration.
The F0 and F1 animals were approximately 6 or 4 weeks old, respectively, when 10 weeks of prebreeding exposures to resorcinol-containing drinking water began. Upon completion of that phase and with resorcinol exposures continuing, rats were randomly bred 1:1 within treatment groups to produce the F1 generation. Individual females were placed into the home cage of a male (nonsibling) for up to 14 days. During the mating period, the appropriate resorcinol drinking water concentration was available, but individual water consumption was not recorded. Females were examined daily for mating success. Following positive evidence of mating (copulation plug and/or vaginal lavage with microscopic verification), females were considered pregnant. This day was designated as gestation day (GD) 0. Even when there was no evidence of mating at the completion of the mating period, such females were transferred to plastic maternity cages with nesting material. Individual water consumption was recorded on GD 1, 4, 7, 14, and 21.
Male and female F1 offspring (30/sex/group) were randomly selected at weaning (PND 21; first day of postnatal life immediately after birth is designated as PND 0) to compose the F1 parental generation. Following weaning, animals selected from the F1 generation for further use as parents for the F2 generation were group housed for 1 week (PNDs 21 to 28). During that week they ingested water containing their group-specific resorcinol concentration and water consumption was monitored by group. On PND 28, animals were transferred to their respective, previously assigned 10-week prebreeding resorcinol exposure groups and housed singly. The breeding procedure (30 animals/sex/group; avoiding sibling matings) was then replicated to produce the second generation. The control group consumed reverse osmosis–purified water under otherwise identical conditions as the resorcinol-exposed groups. After mating, resorcinol exposure continued for the males until 1 day prior to termination and for the females throughout gestation and lactation until 1 day prior to euthanasia. The entire resorcinol exposure duration for surviving F0 males and females was 129 to 135 days, whereas F1 males and females selected to produce the second generation of offspring consumed resorcinol-supplemented drinking water for 144 to 157 days.
Following weaning of the litters in the F1 and F2 generations, the dams continued consumption of their group-specific drinking water and were individually housed in suspended wire-mesh cages until the scheduled study termination. The offspring of the F0 and F1 generations (F1 and F2 animals, respectively) were potentially exposed to resorcinol in utero and via milk through nursing during PNDs 0 to 21. Furthermore, in later stages of postnatal life (>PND 15) some direct exposure via drinking water of the F1 offspring selected for breeding may have occurred in their maternal home cage.
Observations and Measurements
All parental animals (F0 and F1) were observed twice daily for general appearance and behavior, as well as for any signs of toxicity related to resorcinol intake. In light of the previously noted hyperexcitability and signs of CNS stimulation both in human subjects and in rats, particular attention was paid to any symptoms indicative of nervous system stimulation. Detailed physical examinations were recorded weekly for all parental animals. Females were also observed twice daily during the period of expected parturition and at parturition for dystocia (prolonged or delayed labor) or other difficulties delivering their anticipated litters.
Individual body weights of F0 and F1 males were recorded weekly throughout the study and prior to the scheduled necropsy. Individual body weights of F0 and F1 females were also measured weekly until evidence of copulation was detected and thereafter on GDs 0, 4, 7, 11, 14, 17, and 20 as well as lactation days (LDs) 1, 4, 7, 14, and 21. Parental feed consumption was recorded on the same days as the body weight measurements, except during the mating period when measurement of individual feed consumption was suspended. Parental water consumption was measured twice a week, except during the breeding period. Mean resorcinol ingestion (mg/kg/day) for each group was determined by dividing the concentration of resorcinol in the water (mg/L) by the ml/kg/day (= g/kg/day) water consumption value for each recorded interval.
Estrous cyclicity was monitored by daily vaginal lavages and light microscopy. The slides from each F0 and F1 female were evaluated daily, beginning 3 weeks prior to pairing and continuing until mating was observed. Females were allowed to deliver naturally and nurture their young to weaning on PND 21. On the day of parturition, the numbers of stillborn and live offspring were recorded. The sex of the newborns was determined. All offspring were examined for external malformations on PND 0. Stillborn and intact offspring, dying between PNDs 0 and 4, were necropsied using a fresh tissue dissection technique (Stuckhardt and Poppe 1984). A detailed gross necropsy was performed on any offspring dying after PND 4 and prior to weaning.
To reduce variability in postnatal growth and development, 8 offspring/litter (4/sex when possible) were randomly selected on PND 4 using a computer-generated selection procedure, except for litters with fewer than eight newborns. Litters were examined daily for survival and any changes in appearance or behavior (including nesting and nursing behavior). Each newborn was individually weighed and received a detailed physical examination on PNDs 1, 4, 7, 14, and 21. Individual sex determinations of all offspring and those selected for survival were performed on PNDs 0, 4, and 21.
Each F1 male offspring selected as a parent for the F2 generation was examined daily for balanopreputial separation beginning on PND 35 (Korenbrot, Huhtaniemi, and Weiner 1977). All selected F1 female offspring were examined for vaginal opening beginning on PND 25 (Adams et al. 1985). These observations continued daily until all animals had attained those two end points. Body weights of the adolescents were recorded on the day of acquisition of these postnatal developmental landmarks.
Following weaning of the offspring, serum concentrations of thyroid-stimulating hormone (TSH), thyroxine (T4), and triiodothyronine (T3) were analyzed in 15 randomly selected F0 and F1 adult males and females of each group at the scheduled study termination date. These hormones were also determined in one surplus offspring/sex/litter from 15 randomly selected F1 and F2 litters per group at the scheduled PND 21 necropsies. In addition, TG hormone analyses were performed on 15 randomly selected F1 and F2 litters per exposure group culled on PND 4. Blood from all offspring was pooled within litters without regard to sex. A solid-phase, chemiluminescent enzyme immunoassay procedure (DPC Immulite) was used to quantitatively measure the total circulating levels of T3 and T4, whereas TSH levels were quantitatively determined using magnetic separation via a commercially procured radioimmunoassay kit (DPC 125I assay system, Amersham Pharmacia Biotech).
Samples of sperm from the right epididymis were collected from each adult F0 and F1 male and evaluated for the percentage of progressively motile sperm. Motile sperm were evaluated using the Hamilton-Thorne HTM-IVOS (Integrated Visual Optical System) Version 12.1 computer-assisted sperm analysis (CASA) system. Sperm morphology was evaluated by light microscopy with modifications of the wet-mount evaluation technique (Linder et al. 1992). The left testis and epididymis of all F0 and F1 males in all groups was evaluated for homogenization-resistant spermatid head counts and daily sperm production rate (testis only; Blazak, Ernst, and Stewart 1985) using the HTM-IVOS system.
Necropsy and Histopathologic Examination
Surviving F0 and F1 adults were necropsied once their offspring had been weaned. The stage of estrus on the day of study termination was determined for all F0 and F1 females in the control and 3000 mg/L groups. Selected F0 and F1 parental tissues and organs were fixed by immersion in neutral-buffered 10% formalin for possible histopathological examination.
Microscopic evaluations were performed on the following tissues of all F0 and F1 parental animals from the control and high-exposure groups and those that had died or were euthanized in extremis: adrenal glands, cervix, right epididymis (caput, corpus, and cauda), coagulating gland, ovaries, oviducts, pituitary gland, prostate gland, seminal vesicles, stomach, right testis, thyroid gland, uterus, vagina, and gross internal lesions (all groups). Periodic acid–Schiff (PAS) and hematoxylin staining were used for the right testis and epididymis, and hematoxylin-eosin staining was used for all other tissues. Quantitative histopathologic evaluation of five sections of the inner third of each ovary (including enumeration of primordial follicles) was conducted on 10 F1 females from the control and 3000 mg/L groups (Bolon et al. 1997; Bucci et al. 1997). A qualitative assessment for the presence or absence of growing follicles, astral follicles and corpora lutea was also performed. In addition, microscopic examinations of the TGs were conducted for all F0 animals in the 1000 and 3000 mg/L groups. The reproductive organs (right testis and epididymis, prostate, seminal vesicles, coagulating gland, uterus, oviducts, ovaries, cervix, and vagina) of all F0 and F1 parental animals in the 120, 360, and 1000 mg/L groups that failed to mate or produce offspring or otherwise exhibited potential reproductive dysfunction (e.g., abnormal estrous cyclicity or andrological changes) were also examined by microscopy.
Organ weights obtained from all F0 and F1 parental animals included adrenals, brain, total and cauda epididymides (weighed separately), kidneys, liver, ovaries, pituitary, prostate, seminal vesicles with coagulating glands and accessory fluids, spleen, testes (weighed separately), fixed TGs (10% neutral-buffered formalin), and uterus with oviducts and cervix.
On PND 21, a complete necropsy similar to that performed on parental animals (with emphasis on potential developmental sequelae and reproductive organ system morphology) was conducted on F1 offspring not selected for resorcinol exposure and breeding as well as on F2 offspring. Brain, spleen, thymus gland, and fixed TG (see above) weights were also recorded for these animals (one/sex/litter only, when available).
In addition to conventional histopathology of the TG, a quantitative stereomicroscopic analysis of thyroid colloid content was conducted on 15 randomly selected F0 males and females in the control and 3000 mg/L groups and 15 randomly selected F0 males in the 1000 mg/L group. For stereomicroscopy, TG images were captured using a microscope-mounted digital camera directly onto a personal computer equipped with image-capture software. An intersection grid (total of 216 intersection points, approximately 85 μm apart) was digitally constructed and superimposed upon each thyroid gland image. The number of grid intersections (points) that fell on areas of colloid in each image was manually counted and divided by the total number of points that fell anywhere on the TG. The resulting value was equal to the volume fraction of colloid in the entire specimen (Elias, Hyde, and Scheaffer 1983). Two TG lobes were scored from each animal, and an average value was obtained. Individual averages were then subjected to statistical analyses.
Bioanalytical Determination of Resorcinol in F1 Parental Plasma Samples
The lack of any readily detectable TG-directed adverse (goitrogenic) effects, despite high daily resorcinol intake by the F0 parental animals, prompted considerations regarding the metabolic fate of resorcinol upon protracted ingestion via drinking water. Additional studies were therefore designed as the study was in progress to determine whether free resorcinol was detectable in the bloodstream of F1 parental animals. Blood samples were obtained during the week prior to study termination. Under isoflurane anesthesia blood was collected from the retro-orbital sinus. The timing of the blood collections destined for the analysis of free resorcinol was very critical and was correlated as closely as logistically feasible to the drinking behavior of the animals. Rats with ad libitum access to drinking water display two episodes of particularly intense water consumption during the dark phase of a 12-h dark/12-h light photoperiod (Spiteri 1982; Johnson and Johnson 1990). The first drinking episode lasts for about 2 h and starts soon after the onset of the dark cycle, whereas a second 2-h burst of drinking activity occurs just prior to the onset of the light cycle. In the present study, the light phase began at 0600 h. Therefore, blood collections were conducted during the first hour of the light cycle. This timing was used in an attempt to be as close as possible to maximal resorcinol blood levels resulting from the second period of intense drinking activity.
For these pilot studies blood was collected from 5 control group males and females and analyzed in parallel with the following randomly selected samples from a total of 15 blood samples/treatment group collected from either gender after 144 to 157 consecutive days of resorcinol ingestion: 120 mg/L (7 females); 360 mg/L (7 males and 7 females); 1000 mg/L (7 males and 7 females); and 3000 mg/L (10 males and 10 females).
Plasma samples were isolated and analyzed for resorcinol concentration using HPLC/MS (mass spectrometry) in the negative electrospray ionization (ESI) mode. The lower limit of quantitation (LLOQ) concentration was 100 ng resorcinol/ml.
Statistical Analyses
All statistical analyses were conducted using two-tailed tests (except as noted below) for a minimum significance level of 5%, comparing each resorcinol-exposed group to the control group. Data obtained from nongravid animals were excluded from statistical analyses. Parental mating, fertility, copulation, and conception indices were evaluated by the chi-square test with Yates’ correction factor (Hollander and Wolfe 1999). Parental body weight, food and water consumption data, estrous cycle and gestation lengths, precoital intervals, implantation sites, unaccounted for implantation sites, number of newborns, live litter sizes, newborn body weights and weight changes, balanopreputial separation and vaginal patency data (day of acquisition and body weight), absolute and relative organ weights, sperm production rate, epididymal and testicular sperm numbers, ovarian primordial follicle counts, and percentages of colloid content in the thyroid gland were subjected to a one-way analysis of variance (ANOVA; Snedecor and Cochran 1980) among all groups. If the ANOVA was significant, Dunnett’s test (Dunnett 1964) was used for the pairwise comparisons to the control group. Serum hormone concentrations were subjected to a parametric one-way ANOVA to determine intergroup differences as described above. If the ANOVA revealed statistically significant (p<.05) intergroup variance, a one-tailed Dunnett’s test was used to compare the resorcinol-exposed groups to the control group. Sperm motility and morphology as well as proportional postnatal offspring survival and sex at birth were statistically analyzed by the Kruskal-Wallis nonparametric ANOVA test (Kruskal and Wallis 1952) followed by the Mann-Whitney U test (Kruskal and Wallis 1952), when appropriate. Histopathologic findings at the scheduled F0 and F1 necropsies were evaluated using a two-tailed Fisher’s exact test (Steel and Torrie 1980).
Regulatory Compliance
The present study was conducted in compliance with the OECD 416 (OECD 2001) and the US EPA TSCA OPPTS 870.3800 (US EPA 1998) testing guidelines. In recognition of the known adverse effects of resorcinol on the TG, the study design went beyond the regulatory testing guidelines by including expanded thyroid functional assessments. All evaluations adhered to the Principles of Good Laboratory Practice (US EPA 1989; OECD 1997).
RESULTS
Resorcinol Consumption and Blood Level Analysis
The average daily intake of resorcinol in the 3000 mg/L animals across both generations ranged from 233 mg/kg for males to 304 (premating and gestation) and 660 (lactation) mg/kg for females (Table 1). Profoundly increased water consumption with concomitantly much higher intake of a water-delivered test chemical during lactation is commonly observed in nursing animals (Christian et al. 2002a, 2002b).
The pilot bioanalytical measurements revealed blood resorcinol concentrations of 116 to 621 ng/ml in one F1 male and two F1 females from the 3000 mg/L group (Table 2). The very low plasma resorcinol levels found in only 3 of 20 blood samples obtained from the highest (3000 mg/L) exposure group indicated that, as anticipated based on published data, the chemical was readily absorbed and rapidly eliminated from the systemic circulation, most likely by efficient metabolism.
Parental Systemic End Points
There were no resorcinol-related parental deaths or clinical signs of toxicity at any of the four exposure levels evaluated (data not shown). There were no signs and symptoms of any CNS-related toxicity. Such symptoms, lasting for 1 to 2 h, were noted when comparable daily doses of resorcinol (225 or 450 mg/kg in male rats) were given by oral bolus dosing (gavage) in studies conducted by the National Toxicology Program (NTP 1992). Modest body weight reductions occurred in both sexes and generations among animals consuming drinking water with 3000 mg resorcinol/L (Figures 1 and 2; F0 generation only). These effects consisted primarily of decreased cumulative body weight gains in 3000 mg/L F0 animals for the overall pre-mating period (F0 females) or the entire resorcinol exposure time (F0 and F1 generation males). There were no clear trends in week-to-week body weight gains among these animals. However, body weights were up to 6.3% lower in F0 females during the last 2 weeks of the premating period and up to 7.1% lower in F1 males during the entire study. Body weights in the 3000 mg/L F0 females were decreased during the first week of gestation (up to 5.5%), throughout lactation (up to 8.4%), and after the lactation period ended (6.3%). Body weights were also reduced in the F1 females consuming 3000 mg resorcinol/L drinking water during lactation (up to 6.1%) and after the lactation period ended (up to 7.0%). Body weights and weight gains were unaffected in males and females exposed to 120, 360, and 1000 mg resorcinol/L in both generations.
Reduced water consumption was recorded among the 3000 mg/L F0 and F1 parental animals during the premating period (females) or the entire two generations (males) and for the F1 offspring that were gang-housed by litter during PNDs 21 to 28 (Figures 1 and 2 for F0 data). Often water intake was also lower in the 1000 mg/L group males and females, although these reductions were less pronounced and the onset occurred later than in the 3000 mg/L group. Water consumption in the 1000 mg/L group was consistently less than in the concurrent control group beginning on study days 21 to 24; however, slight decreases were also observed, although inconsistently, earlier in the premating period. The reduced water intake in the 1000 mg/L group continued during the first week of gestation, whereas in the 3000 mg/L group females drank less throughout gestation and lactation. However, feed intake and utilization in the 1000 and 3000 mg/L groups in both generations were unaffected. The decreases in water consumption at these levels were therefore attributed to resorcinol exposure but were not considered an adverse effect. Both water and feed consumption as well as feed utilization in both generations were unaffected in males and females of the 120 and 360 mg resorcinol/L treatment groups.
Reproductive Function
F0 and F1 Adults
When evaluated across two generations, reproductive performance (regularity and duration of estrus, mating, fertility, conception and copulation indices, numbers of days between pairing and coitus, and gestation length), spermatogenic end points (testicular and epididymal sperm concentrations and sperm production rate, motility, progressive motility and morphology), and parturition were unaffected by exposure to resorcinol (Tables 3 and 4). There were no macroscopic or microscopic findings in the reproductive organs and no effects on reproductive organ weights (data not shown).
F1 and F2 Offspring
The numbers of F1 and F2 newborns, live litter sizes, sex ratios at birth, and postnatal survival were unaffected by parental exposure to resorcinol (Table 3). No test article–related effects were observed on neither F1 nor F2 offspring body weights or body weight gains (Table 3). The general physical condition and appearance of the offspring during the preweaning period were indistinguishable from that of the control animals. There were no effects on the acquisition of sexual developmental landmarks (balanopreputial separation and vaginal patency) in the F1 offspring (Table 5).
Parental and Offspring Thyroid/Pituitary Hormone Levels
There were no biologically significant resorcinol exposure-induced changes in T3, T4, or TSH serum concentrations in any of the treatment groups of either sex when F0 and F1 adults were evaluated at necropsy (Table 6) or when these TG functional status parameters were determined in selected F1 and F2 offspring on either PND 4 (pooled by litter) or PND 21 (Table 7). Mean TSH concentrations were slightly elevated in an apparent resorcinol exposure concentration-related manner in the 360, 1000, and 3000 mg/L group F0 males, but none of the values were statistically significant compared to the concurrent control rats. Moreover, a similar rise in TSH concentration was not detectable in the females. Curiously, the T3 concentration was actually increased in the 3000 mg/L group males when compared to controls. Although some PND 21 group means occasionally achieved statistical significance relative to controls (higher TSH for F1 males at 360 and 3000 mg/L and slightly lower T4 for F2 females at 1000 mg/L), there were no apparent biological effects with treatment-related trends. Overall the changes were therefore not attributed to parental resorcinol exposure.
Necropsy and Histopathology
F0 and F1 Adults
One F0 female (1000 mg/L) and single F1 males in the control, 360, and 1000 mg/L groups died or were euthanized in extremis (due to mechanical injury) during the study. However, necropsy findings in the resorcinol-treated animals that died were suggestive of underlying disease conditions that led to pathological changes. In light of the finding that no parental mortality occurred at the 3000 mg/L exposure level, none of these deaths were attributed to resorcinol consumption.
No resorcinol-related parental organ weight differences or adverse histopathologic alterations were observed for either sex in either generation (TG data presented in Table 8). No micromorphologic differences were noted upon histopathologic examination of male or female reproductive organs in the animals that had displayed reduced fertility, such as failure to mate, conceive, or sire a litter. Ovarian primordial follicle counts were unaffected in the F1 females that had consumed drinking water containing 3000 mg resorcinol/L (Table 3).
Reduced thyroid follicle colloid content occurred with increased incidence in 3000 mg/L group F0 males and was considered a test article-related microscopic finding. Decreased colloid content, characterized by small follicles with little or no colloid present within the follicular lumen, was diagnosed when that change diffusely affected both TG lobes. Using these criteria, the incidence in males was 2/30, 2/30, and 7/30 in the control, 1000, and 3000 mg/L groups, respectively. In females, the incidence of this change was 3/30, 6/30, and 4/30 in these same treatment groups. It appeared as if only the highest exposure concentration of 3000 mg/L had affected follicular colloid content in F0 males. Subsequently, a more sensitive stereoscopic micromorphometry method was applied to the TG histological sections of F0 control and 3000 mg/L group males and females as well as to 1000 mg/L males. This approach yields objective numbers that can be statistically analyzed. It was thus confirmed that the difference in follicular colloid content was significantly lower in F0 males from the 3000 mg/L group, but not in females (Figure 3A and B ) and not in F0 males from the 1000 mg/L group (data not shown). Figure 3A shows the results from the 3000 mg/L males and revealed that the follicular colloid content was statistically significantly different from concurrent controls. By contrast, Figure 3B reveals that in 3000 mg/L F0 females there were no significant differences in colloid content between resorcinol-exposed rats and controls.
A single F0 male in the 3000 mg/L group had markedly increased TG weights, both in absolute terms and relative to the final body and brain weights. Histopathological examination disclosed bilateral macrofollicular cell adenomas. Neither any male of the F1 3000 mg/L group nor any female rat of either the F0 or F1 generation had any significant TG weight or histopathological alterations compatible with adenomatous changes.
F1 and F2 Offspring
No macroscopic anomalies attributable to resorcinol exposure were observed upon necropsy of F1 and F2 offspring that were found dead or examined at the scheduled PND 21 study terminations. There were no changes in brain, spleen, thymus gland, and TG weights among F1 and F2 offspring, compared to the concurrent control animals, at the scheduled PND 21 necropsies (data not shown).
DISCUSSION
The results obtained in the present test guideline–compliant two-generation reproduction study conducted in rats, at maximum palatable drinking water concentration levels, have filled a major gap in the toxicology database of resorcinol. The evaluations performed went beyond the testing guidelines because they included a comprehensive assessment of the histological and functional status of the TG. Preceding experimental studies in animals and human clinical medicine case reports had identified that endocrine gland as the critical target organ. Resorcinol exposure of rats, at present the preferred test animal species for reproductive toxicity hazard assessments and in addition for agents with potential TG toxicity, occurred by drinking water. The oral route, as well as the protracted resorcinol intake via water, were most appropriate with respect to the anticipated potential exposure of the human population at large from contaminated drinking water at concentrations that would most likely be much lower that those evaluated in the present study. Pilot pharmacokinetic measurements indicated that resorcinol was readily bioavailable, as was anticipated based on preceding metabolic disposition studies following oral administration (Kim and Matthews 1987). However, rapid metabolism and removal of free resorcinol from the systemic circulation prevented sustained delivery of thyrotoxic concentrations to the TG. Thus, the ingestion of high daily doses of resorcinol (= administered dose) did not cause TG toxicity because of efficient metabolism such that the dose delivered to the target organ was neither high nor sustained enough to elicit adverse effects. This interpretation of the data obtained in the present study is supported by the observations that administration of slow release formulations of resorcinol to rats (oily solutions or dosing with resorcinol diacetate) allows TG toxicity and goitrogenic histologic effects to manifest themselves (Doniach and Logothetopoulos, 1953). Regrettably, those studies did not include any pharmacokinetic measurements of plasma resorcinol concentrations. When high doses are administered by either SC injection or oral bolus, they cause nonspecific signs of acute and transient toxic effects on the CNS at doses in the range of 110 to 154 mg/kg (Doniach and Logothetopoulos 1953; NTP 1992). In contrast, the protracted ingestion of high daily doses of resorcinol via the drinking water did not elicit any detectable signs of CNS stimulation. Our present pilot bioanalytical determinations revealed that very low concentrations of free resorcinol in blood were detectable in only a few animals of either sex from the 3000 mg/L groups. These data, obtained with a highly sensitive analytical method, showed that resorcinol was rapidly metabolized and removed from the systemic circulation.
When evaluated across both generations, the data obtained in this study have demonstrated that even at very high daily resorcinol intake, the reproductive performance in rats of either sex was unaffected. There were no detectable adverse effects on either F1 or F2 offspring growth, survival, or general health and well-being during the preweaning period or on sexual maturation in the F1 generation. Furthermore, it is notable that despite extended exposure to high daily doses, no resorcinol-related gross pathology, organ weight changes, or adverse microscopic target organ effects were observed in either the F0 or F1 parental animals. This no detectable changes outcome extended to all biological end points recorded for F1 or F2 offspring.
The study design provided resorcinol exposure levels up to the maximum palatable concentration. Although there were modest adverse effects on water consumption and cumulative body weight gains in both parental generations in the 3000 mg/L group, there was no evidence of additive effects on mean body weights or body weight gains across two generations. Furthermore, the reproduction data revealed no indications of enhanced sensitivity of females during gestation and lactation. Less pronounced reduced water consumption was also observed in F0 males and females in the 1000 mg/L groups during the pre-mating period. Resorcinol-related decreases in water intake were not considered adverse because there were no associated effects on feed intake and feed utilization, indicating that metabolic homeostasis was uncompromised.
In spite of the high and extended duration exposure levels, there were no significant resorcinol-induced changes in the serum concentrations of T3, T4, or TSH in either the F0 or F1 parental animals at the termination of the study or in the F1 or F2 offspring on PND 4 or 21. The histologic appearance of TGs from the relatively young rats in both controls and those administered resorcinol via drinking water showed considerable variability within individual thyroid lobes, between the thyroid lobes of the same rat and between individual rats. This inherent physiologic variability made microscopic detection of subtle structural alterations difficult. The histologic changes of thyroid follicular cells were minimal and were consistent with a non-adverse adaptive response of the rat TG to a goitrogen with a short plasma half-life. It must be emphasized that all TG end points in the F0 and F1 parents were assessed only once at the very end of the study when the animals had been exposed to resorcinol for several months. Functionally, the normal rat TG is more active than that of most other animal species and humans. This is reflected in the physiologic histological structure of the gland and witnessed by smaller follicles with less stored colloid, which are lined by more columnar follicular cells. This “active” histologic structure of the normal rat TG makes the detection of minimal to mild increases in functional activity difficult in this species (Capen 2001).
Significantly reduced TG follicular colloid content was observed only in F0 males of the 3000 mg/L group. It is known from numerous other studies that male rats are more sensitive than female rats to agents that act on the TG (Capen 2001). Colloid content in the TGs of F0 females and F1 males and females in the 3000 mg/L group was indistinguishable from controls, as was colloid content in the 1000 mg/L F0 male and female rats. The lower colloid content in 3000 mg/L F0 males was not considered to be an adverse effect for the following reasons:
There were no resorcinol-related changes in mean absolute and/or relative TG weights in any of the four treatment groups of either sex.
There were no alterations in other TG end points in either generation. One might anticipate more pronounced TG effects in the F1 generation due to the prolonged exposure, which began prenatally via parental treatment and continued postnatally without interruption. Yet histology revealed that the colloid content in both the 3000 mg/L group F1 males and females was indistinguishable from concurrent controls. Although the follicular lumenal colloid content of male F0 rats was reduced at the 3000 mg/L exposure level, there was apparently an adaptive cellular response to a new homeostatic function level with no adverse physiological consequences.
A dose-related trend towards higher TSH levels was observed only in F0 males, but statistically the values were not different from the controls, and there was no such effect in F0 females or F1 rats of either sex. Treatment-related effects that do not alter function and are not associated with effects on other related end points are less likely to be adverse effects than if the converse were true (Lewis 2002).
In our study, there were no functional sequelae associated with the decreased colloid content in F0 males. Serum T3, T4, and TSH were not consistently statistically different from controls. Furthermore, there were no alterations in any of the reproductive parameters.
Therefore, the present observations indicate overall that sufficient internal exposure to resorcinol was achieved for a critical evaluation of potential adverse thyroid and reproductive effects of the chemical.
The presence of bilateral TG adenomas in one F0 male rat in the 3000 mg/L group was most likely a spontaneous occurrence. Follicular cell adenomas in the TG of rats are not uncommon tumors and occur as low incidence spontaneous background lesions. Control rats of the Sprague-Dawley strain at comparable age can occasionally develop a thyroid tumor without any chemical exposure (National Research Council 2005).
Among more than 300 chemicals evaluated in rat lifetime carcinogenesis bioassays by the US National Cancer Institute and National Toxicology Program, only 18 have been associated with the development of follicular cell neoplasms. Even when the neoplastic response was restricted to a single sex, the opposing one had exposure-induced follicular cell hyperplasia (Hardisty and Boorman 1990). In the present study neither males of the F1 generation nor females in either generation displayed any TG histological changes. The National Toxicology Program has extensively evaluated resorcinol toxicity and carcinogenicity in rats and mice. No neoplasms or preneoplastic lesions in the TG follicles were noted in 17-day, 13-week, or 2-year studies in either species (NTP 1992).
The daily resorcinol exposure levels given via gavage on five days each week in the NTP testing were of similar magnitude to the daily doses delivered 7 days each week by the administration via drinking water in the present study. Therefore, the data collected in our comprehensive evaluation with expanded TG end points (qualitative histology and quantitative stereomicroscopy; TSH, T3, and T4), including pilot pharmacokinetic measurements, should be considered in the context of all other chemical testing outcomes that included the TG.
Based on the results presented it was concluded that the 3000 mg resorcinol/L concentration may be designated as the no-observed-adverse-effect level (NOAEL), whereas the 1000 mg/L level was the no-observed-effect level (NOEL). Neither the TG nor any of the reproductive end points evaluated over two generations were adversely affected by high daily resorcinol doses ingested via drinking water.
Footnotes
Figures and Tables
Acknowledgments
The authors thank Professor Charles C. Capen, DVM, MSc, PhD, Department of Veterinary Biosciences, The Ohio State University, for his critical review of the thyroid histology microscopic slides.