Physiology and Endocrinology of the Ovarian Cycle in Macaques

Abstract

Macaques provide excellent models for preclinical testing and safety assessment of female reproductive toxicants. Currently, cynomolgus monkeys are the predominant species for (reproductive) toxicity testing. Marmosets and rhesus monkeys are being used occasionally. The authors provide a brief review on physiology and endocrinology of the cynomolgus monkey ovarian cycle, practical guidance on assessment and monitoring of ovarian cyclicity, and new data on effects of social housing on ovarian cyclicity in toxicological studies. In macaques, cycle monitoring is achieved using daily vaginal smears for menstruation combined with cycle-timed frequent sampling for steroid and peptide hormone analysis. Owing to requirements of frequent and timed blood sampling, it is not recommended to incorporate these special evaluations into a general toxicity study design. Marmosets lack external signs of ovarian cyclicity, and cycle monitoring is done by regular determinations of progesterone. Cynomolgus and marmoset monkeys do not exhibit seasonal variations in ovarian activity, whereas such annual rhythm is pronounced in rhesus monkeys. Studies on pair- and group-housed cynomolgus monkeys revealed transient alterations in the duration and endocrinology of the ovarian cycle followed by return to normal cyclicity after approximately six months. This effect is avoided if the animals had contact with each other prior to mingling. These experiments also demonstrated that synchronization of ovarian cycles did not occur.

FIGURE 1

Regulation of hypothalamic GnRH system and pituitary gonadotropin synthesis and release. Kisspeptin, the ligand of the GPR54 receptor located on GnRH neurons, is a key regulator of GnRH production. Steroids, leptin, and Ghrelin are believed to modulate GnRH secretion via the kisspeptin/GPR54 system. GnRH I and II molecules are present in primates, and both molecules act via GnRH I receptors in gonadotropes. Synthesis and secretion of LH and FSH are locally modulated by a variety of paracrine factors.

FIGURE 2

Endocrine profiles of estradiol (E₂), LH, FSH, and progesterone (P) during the ovarian cycle in the cynomolgus monkey and presumed feedback actions of ovarian hormones. The day of ovulation is denoted as zero, and days for follicular and luteal phases are counted relative to ovulation time point. Endocrine data represent M ± SD of forty-four animals except for FSH with data from fourteen animals. During ovulation, ovarian steroids are assumed to exert primarily a direct and positive feedback effect on pituitary gonadotropin secretion, whereas negative feedback actions occur during the follicular and luteal phases. Studies in ovariectomized monkeys also suggested a hypothalamic site of E₂ action. Whether progesterone also has a direct positive effect on hypothalamus is unclear.

FIGURE 3

Duration of the ovarian cycle in the cynomolgus monkey based on daily vaginal smears. Data represent 926 spontaneous cycles. Average cycle (± SD) duration is 30.4 ± 4.7 days with a range of 19 to 69 days. Median cycle duration is 30 days, indicating a normal distribution of cycle duration.

FIGURE 4

Schematic representation of the concept of follicle selection in macaques (modified from Zeleznik 2001). For explanation, the reader is referred to section “Follicular Phase and Follicle Selection.”

FIGURE 5

Graphic representation of daily bleeding records from vaginal smears that are used to determine ovarian cycle duration in the cynomolgus monkey. The data correspond to those shown in Figure 15 (animals la and 2a). X-axis denotes days, and y-axis denotes the number of cycles. Grey boxes indicate bleeding. Black boxes aid in cycle determination by denoting the number of days for every month. Cycle duration is determined by the interval between first day of bleeding and last day prior to onset of next bleeding. Note that animal la maintained a rather constant cycle duration (26–31 days), whereas its companion, animal 2a, showed cycle irregularity (increased cycle length up to 52 days or lack of cyclicity) during pair housing. Cyc. = cycle duration in days; P = shift from single to pair housing; SD = start of dosing; DS = dosing stopped.

FIGURE 6

Perineal sex-skin color changes and swelling in adult cynomolgus monkeys during the follicular, ovulatory, and luteal phases of the ovarian cycle. Paired photographs of the sex skin and ovaries from individual animals were taken at the time of experimental ovariectomy. Note that these changes occur only in some animals and cannot be used to generally monitor ovarian cyclicity in this species.

FIGURE 7

Progesterone-based ovarian cycle monitoring in the marmoset. Note that luteal-phase duration is about twice as long as follicular phase. To initiate cycle monitoring, prostaglandin F2-alpha is used for induction of luteolysis in the first cycle (arrow).

FIGURE 8

Group mean cycle length (left panel) and individual menstrual-cycle length (right) from eight animals under individual and group-housing conditions. Data are M ± SD and are displayed in Figure 13 for every animal. The reference line (blue circles) in the left panel depicts the mean cycle duration in seventy-eight single-housed animals (M ± SD). Dotted lines indicate the normal variation (30.4 ± 4.7 days) of menstrual cycle length. Note the absence of seasonal variation in cycle duration. The right panel depicts the individual data for the eight individual animals. Note the increased variability and cycle duration after transition from single to group housing and the fact that these cycle irregularities were transient. Following approximately six cycles, ovarian cycle duration had returned to normal. The x-axis denotes the numbers of successive ovarian cycles. Negative cycles numbers are those prior to having all eight animals transferred into one group.

FIGURE 9

Group mean (± SD) progesterone (left panel) and estradiol (right panel) levels during ovarian cycles from eight animals under individual and group-housing conditions (cycle duration is shown in Figure 8). Data for cycle 1 were collected during single housing, whereas subsequent cycle data are derived under group-housing conditions (cycles 3, 5, 7, 9, and 11). The asterisks indicate statistical significances. Note that during cycle 3—within the period of cycle irregularity and prolongation—the ovulatory estradiol peak and luteal progesterone peak were suppressed, suggesting a transient lack of ovulation.

FIGURE 10

Individual ovarian cycle pattern (double plot) of pair-housed cynomolgus monkeys with separate cycle patterns and not influenced by the cage mate. The y-axis denotes the grading of vaginal smear evaluation.

FIGURE 11

Individual ovarian cycle pattern (double plot) of pair-housed cynomolgus monkeys with parallel cycle patterns and not influenced by the cage mate. The y-axis denotes the grading of vaginal smear evaluation.

FIGURE 12

Individual ovarian cycle pattern (double plot) of pair-housed cynomolgus monkeys with irregular cycle patterns and not influenced by the cage mate. The y-axis denotes the grading of vaginal smear evaluation.

FIGURE 13

Individual cyclicity pattern of eight cynomolgus monkeys throughout a period of approximately seventeen months (data were replotted from Figure 8). Animals were initially single housed (shaded area) and then co-housed as a group. Every plot represents one animal. The y-axis denotes the grading of vaginal smear evaluation. Note the absence of synchronization and seasonality and the transient irregularity/absence of cycling in some animals during group housing.

FIGURE 14

Cycle duration before and during treatment in six cynomolgus monkeys housed in pairs (animal 1/animal 2, animal 3/animal 4, animal 5/animal 6). Prior to formation of pairs, the animals knew each other. Vertical arrow represents the onset of oral daily dosing of vehicle in a toxicity study. Dashed horizontal lines represent the upper and lower limits of spontaneous ovarian cycle duration. Note that cycle duration remained within normal limits for all animals.

FIGURE 15

Cycle duration before and during treatment in four cynomolgus monkeys housed in pairs (upper panel: animal la/animal 2a; lower panel: animal 3a/animal 4a). Prior to formation of pairs (long vertical arrow), the animals were single housed and did not know each other. Short vertical arrow represents the onset of oral daily dosing of vehicle in a toxicity study. Dashed horizontal lines represent the upper and lower limits of spontaneous ovarian cycle duration. Note that cycle duration remained within normal limits in one animal of each pair, whereas in the other animal, cycle duration transiently was prolonged or menstrual bleeding was absent.