In vivo antral follicle wall biopsy: a new research technique to study ovarian function at the cellular and molecular levels.

BACKGROUND: In vivo studies involving molecular markers of the follicle wall associated with follicular fluid (FF) milieu are crucial for a better understanding of follicle dynamics. The inability to obtain in vivo samples of antral follicle wall (granulosa and theca cells) without jeopardizing ovarian function has restricted advancement in knowledge of folliculogenesis in several species. The purpose of this study in mares was to develop and validate a novel, minimally invasive in vivo technique for simultaneous collection of follicle wall biopsy (FWB) and FF samples, and repeated collection from the same individual, during different stages of antral follicle development. We hypothesized that the in vivo FWB technique provides samples that maintain the normal histological tissue structure of the follicle wall layers, offers sufficient material for various cellular and molecular techniques, and allows simultaneous retrieval of FF. METHODS: In Experiment 1 (ex vivo), each follicle was sampled using two techniques: biopsy forceps and scalpel blade (control). In Experiment 2 (in vivo), FWB and FF samples from 10-, 20-, and 30-mm follicles were repeatedly and simultaneously obtained through transvaginal ultrasound-guided technique. RESULTS: In Experiment 1, the thickness of granulosa, theca interna, and theca externa layers was not influenced (P > 0.05) by the harvesting techniques. In Experiment 2, the overall recovery rates of FWB and FF samples were 97 and 100%, respectively. However, the success rate of obtaining samples with all layers of the follicle wall and clear FF varied according to follicle size. The expression of luteinizing hormone receptor (LHR) was mostly confined in the theca interna layer, with the estradiol-related receptor alpha (ERRα) in the granulosa and theca interna layers. The 30-mm follicle group had greater (P < 0.05) LHR expression in the theca interna and ERRα in the granulosa layer compared to the other groups. The overall expression of LHR and ERRα, and the intrafollicular estradiol were higher (P < 0.05 - P < 0.0001) in the 30-mm follicle group. CONCLUSION: The in vivo technique developed in this study can be repeatedly and simultaneously used to provide sufficient FWB and FF samples for various cellular and molecular studies without jeopardizing the ovarian function, and has the potential to be translated to other species, including humans.

Follicle growth and endocrine dynamics in women with spontaneous luteinized unruptured follicles versus ovulation.

STUDY QUESTION: Do growth patterns and endocrine profiles differ between ovulatory follicles (OvFs) and luteinized unruptured follicles (LUFs) in women? SUMMARY ANSWER: Growth rates, diameters and associated endocrine profiles differed between OvFs and LUFs in unstimulated cycles. WHAT IS KNOWN ALREADY: Two-three waves of antral follicles develop during the menstrual cycle in ovulatory women of reproductive age, with the second or third wave terminating in ovulation. In contrast, some women can develop LUFs, where a preovulatory follicle fails to rupture and there is subsequent luteinization of the follicle wall. However, no study has compared OvFs and LUFs in unstimulated cycles. STUDY DESIGN, SIZE, DURATION: This retrospective observational study was conducted in 56 healthy women of reproductive age (range: 19-41 years) and with a history of regular menstrual cycles. PARTICIPANTS/MATERIALS, SETTING, METHODS: Participants who met inclusion criteria were enrolled, as previously reported. Daily transvaginal ultrasonography was performed for one interovulatory interval (IOI) to measure the diameters of all follicles >2 mm. Blood samples were collected every 3 days during the IOI to measure serum concentrations of FSH, LH, estradiol and progesterone. MAIN RESULTS AND THE ROLE OF CHANCE: The interval from emergence to deviation (i.e. follicle selection) was shorter (P < 0.05) for LUFs compared to OvFs. However, the intervals from emergence to maximum diameter and deviation to maximum diameter were longer (P < 0.05) for LUFs compared to OvFs. Follicle deviation in LUFs occurred at a larger diameter (P < 0.05) compared to OvFs, and LUFs grew to larger (P < 0.0001) diameters compared to OvFs. Moreover, LUFs grew faster (P < 0.05) from emergence to deviation and from deviation to maximum diameter, compared to OvFs. LUFs were associated with low (P < 0.05) systemic LH levels at emergence and maximum diameter compared to OvFs. LUFs were also associated with low (P < 0.05) systemic FSH and high (P < 0.05) systemic progesterone at deviation and maximum diameter, respectively. Estradiol was higher (P < 0.05) at deviation and lower (P < 0.05) at maximum diameter for LUFs compared to OvFs. LIMITATIONS, REASONS FOR CAUTION: A 3-day interval of blood sampling for hormonal analyses was conducted, as a more frequent sampling interval was not considered acceptable by the study volunteers. A 3-day sampling interval did not allow characterization of acute changes in hormone production during the IOI. In addition, study visits were less frequent when LUFs persisted long after the expected day of the second ovulation of the IOI. WIDER IMPLICATIONS OF THE FINDINGS: Information about the growth and endocrine dynamics of OvFs and LUFs developing in unstimulated cycles in women may be applied to the early detection of LUF-associated anovulatory infertility and clinical management of women with this condition. STUDY FUNDING/COMPETING INTEREST(S): No external funding sources were used for this study. The authors have no conflicts of interest in publishing this manuscript. TRIAL REGISTRATION NUMBER: ClinicalTrials.gov Identifier: NCT01389141.

Pre-ovulatory follicle affects corpus luteum diameter, blood flow, and progesterone production in mares.

Color Doppler ultrasonography was used to study the temporal relationships between pre-ovulatory follicle (POF) and corpus luteum (CL) diameter and blood flow, with systemic progesterone (P4) concentration during two transitional ovulatory seasons in mares. Variables of POF and CL/P4 were evaluated for 6days before and 17days after ovulation, respectively. Evaluations were performed during two consecutive estrous cycles in spring and fall seasons, and during the last estrous cycle of the season. There were significant correlations among POF and CL variables, and P4 concentration that ranged from 0.24 to 0.95, and among the ratios of different variables that ranged from 0.39 to 0.92. There were linear regressions (P<0.01-0.001) for all comparisons among different variables. The POF diameter before the first ovulation of the season was larger (P<0.05), and POF vascularity was less (P<0.05), than in the last estrous cycle during the season. The CL blood flow was less (P<0.01) during the last compared with first pre-ovulatory period of the season. The POF diameters were positively correlated (r=0.67) during the two pre-ovulatory periods of spring and fall. Results provide evidence that the POF affects CL diameter and blood flow, and subsequently P4 production, and that POF diameter is repeatable within the same individual during different seasons.

Follicle vascularity coordinates corpus luteum blood flow and progesterone production.

Colour Doppler ultrasonography was used to compare the ability of preovulatory follicle (POF) blood flow and its dimensions to predict the size, blood flow and progesterone production capability of the subsequent corpus luteum (CL). Cows (n=30) were submitted to a synchronisation protocol. Follicles ≥7mm were measured and follicular wall blood flow evaluated every 12h for approximately 3.5 days until ovulation. After ovulation, cows were scanned daily for 8 days and similar parameters were evaluated for the CL. Blood samples were collected and plasma progesterone concentrations quantified. All parameters were positively correlated. Correlation values ranged from 0.26 to 0.74 on data normalised to ovulation and from 0.31 to 0.74 on data normalised to maximum values. Correlations between calculated ratios of both POF and CL in data normalised to ovulation and to maximum values ranged from moderate (0.57) to strong (0.87). Significant (P<0.0001) linear regression analyses were seen in all comparisons. In conclusion, higher correlations were observed between the dimensions of POF and/or CL and blood flow of both structures, as well as POF and/or CL blood flow with plasma progesterone concentrations of the resultant CL. These findings indicate that follicle vascularity coordinates CL blood flow and progesterone production in synchronised beef cows.

Changes in intrafollicular concentrations of free IGF-1, activin A, inhibin A, VEGF, estradiol, and prolactin before ovulation in mares.

Changes in intrafollicular growth factors and hormones were evaluated in vivo in postdeviation and impending ovulation follicles. Mares (n = 30) were randomly assigned to five experimental groups based on target diameters of 25, 30, 35, 40 mm, and impending signs of ovulation. Furthermore, data belonging to two or more proximal diameter groups that were not different were combined and regrouped for each factor separately. Follicular fluid-free insulin-like growth factor 1 was highest (P < 0.003) in 35-mm follicles, followed by the 40-mm and impending ovulation follicle group, and the 25- to 30-mm follicle group. However, concentrations of insulin-like growth factor binding protein 2 in follicular fluid did not differ (P > 0.05) among groups. Additionally, follicular fluid activin A tended (P < 0.06) to be higher in impending ovulation follicles when compared with the 25- to 40-mm follicle group. Concentrations of intrafollicular estradiol were higher (P < 0.0001) in 40-mm and impending ovulation follicles than in the other follicle groups. Follicular fluid concentrations of inhibin A and vascular endothelial growth factor were lower (P < 0.05) in the 40-mm and the impending ovulation follicle group when compared with the 25- to 35-mm follicle group. Systemic and intrafollicular prolactin levels were lower (P < 0.05) in the impending ovulation group when compared with the 25- to 40-mm follicle group. Prolactin concentrations were higher (P < 0.05) in the follicular fluid than in the plasma. The novel findings of this study, a decrease in intrafollicular-free insulin-like growth factor 1, inhibin A, vascular endothelial growth factor, and prolactin during the final stages of follicular growth, document for the first time the occurrence of dynamic changes among intrafollicular factors and hormones during the stages of follicle dominance and as ovulation approaches.

Mechanism for greater frequency of contralateral than ipsilateral relationships between corpus luteum and ovulatory follicle for wave 3 in heifers.

During the last wave of the interovulatory interval (IOI), the permutations of the relationship between the ovulatory follicle and the CL (ipsilateral vs. contralateral) and the number of follicular waves (two vs. three) per IOI differ in frequency of occurrence as follows: ipsilateral relationship and two waves (34%), contralateral relationship and two waves (34%), ipsilateral relationship and three waves (8%), and contralateral relationship and three waves (24%). Deviation or the continuation in growth rate of the future ovulatory follicle and a decrease in growth rate of the future subordinate follicles begin well before luteolysis in two-wave IOIs and during luteolysis in three-wave IOIs. The largest follicle decreases in diameter and loses its dominant status before completion of deviation when it is ipsilateral and adjacent to the regressing CL during wave 3. Dominant status switches from the largest follicle in the ipsilateral ovary to the next-largest follicle which may be in either ovary. Switching accounts for the greater frequency of a contralateral follicle-CL relationship than for ipsilateral follicle-CL relationship during the ovulatory wave in three-wave IOIs. It is proposed that the phenomenon results from commonality in angioarchitecture so that the decrease in blood flow to the regressing CL is associated with a decrease in blood flow to adjacent follicles.

The mare as a model for luteinized unruptured follicle syndrome: intrafollicular endocrine milieu.

Luteinized unruptured follicle (LUF) syndrome is a recurrent anovulatory dysfunction that affects up to 23% of women with normal menstrual cycles and up to 73% with endometriosis. Mechanisms underlying the development of LUF syndrome in mares were studied to provide a potential model for human anovulation. The effect of extended increase in circulating LH achieved by administration of recombinant equine LH (reLH) or a short surge of LH and decrease in progesterone induced by prostaglandin F2α (PGF2α) on LUF formation (Experiment 1), identification of an optimal dose of COX-2 inhibitor (flunixin meglumine, FM; to block the effect of prostaglandins) for inducing LUFs (Experiment 2), and evaluation of intrafollicular endocrine milieu in LUFs (Experiment 3) were investigated. In Experiment 1, mares were treated with reLH from Day 7 to Day 15 (Day 0=ovulation), PGF2α on Day 7, or in combination. In Experiment 2, FM at doses of 2.0 or 3.0 mg/kg every 12 h and human chorionic gonadotropin (hCG) (1500 IU) were administered after a follicle ≥32 mm was detected. In Experiment 3, FM at a dose of 2.0 mg/kg every 12 h plus hCG was used to induce LUFs and investigate the intrafollicular endocrine milieu. No LUFs were induced by reLH or PGF2α treatment; however, LUFs were induced in 100% of mares using FM. Intrafollicular PGF2α metabolite, PGF2α, and PGE2 were lower and the ratio of PGE2:PGF2α was higher in the induced LUF group. Higher levels of intrafollicular E2 and total primary sex steroids were observed in the induced LUF group along with a tendency for higher levels of GH, cortisol, and T; however, LH, PRL, VEGF-A, and NO did not differ between groups. In conclusion, this study reveals part of the intrafollicular endocrine milieu and the association of prostaglandins in LUF formation, and indicates that the mare might be an appropriate model for studying the poorly understood LUF syndrome.

Relationship between days of the luteolytic period and locations of the preovulatory follicle and CL in interovulatory intervals with two or three follicular waves in heifers.

Diameter of follicles was determined every 12 hours and progesterone (P4), FSH, and LH concentrations were determined every 6 hours from Day 12 (Day 0 = ovulation) to the ovulation at the end of the interovulatory interval (IOI). Groups were assigned on the basis of an ipsilateral (Ipsi) versus contralateral (Contra) relationship between the preovulatory follicle and CL and two follicular waves (2W) versus three waves (3W) per IOI. Numbers of IOIs were Ipsi-2W (n = 6), Ipsi-3W (n = 6), and Contra-3W (n = 8). Normalization to the end of luteolysis (day that P4 was closest to 1.0 ng/mL) indicated for the first time that concentrations of P4 and FSH were greater (P < 0.05) in 3W IOIs than in 2W IOIs for the 3 days before the beginning of a P4 decrease. The beginning of a P4 decrease occurred about 5 days and 6 hours after emergence of the preovulatory wave at 6 mm in 2W and 3W IOIs, respectively. On the day of diameter deviation between the future dominant and largest subordinate follicles in wave 3 of 3W IOIs, the future dominant follicle had the following characteristics: (1) distribution of diameters differed (P < 0.01) from unimodality; (2) diameter was greater (P < 0.05) in the Contra-3W group (9.8 ± 0.4 mm) than in the Ipsi-3W group (8.8 ± 0.3 mm); (3) diameter was similar to the diameter at the beginning of the P4 decrease (9.6 ± 0.9 mm); and (4) diameter was as small or smaller than diameter of the largest subordinate in seven of 14 heifers compared with zero of seven heifers in wave 2 of 2W IOIs. The differences involving deviation may be related to a reported greater frequency of the Contra-3W group than Ipsi-3W group. Results supported the hypothesis that emergence of the ovulatory wave occurs well before the beginning of luteolysis in 2W IOIs and near the beginning of luteolysis in 3W IOIs.

Minor FSH surge, minor follicular wave, and resurgence of preovulatory follicle several days before ovulation in heifers.

Blood samples were collected and follicle diameters were determined daily beginning on Day 12 (Day 0 = ovulation) in 35 interovulatory intervals (IOIs) in heifers. A minor follicular wave with maximal diameter (6.0 ± 0.3 mm) on Day -4 was detected in six of seven IOIs that were scanned for follicles 4 mm or greater. The number of IOIs with a CV-identified minor FSH surge toward the end of the IOI was greater (P < 0.03) in two-wave IOIs (10/17) than in three-wave IOIs (4/18). The 17 two-wave IOIs were used for study of the temporal relationships among preovulatory follicle, FSH, LH, and estradiol. Daily growth rate of the preovulatory follicle was maximum on Days -11 to -7, minimum (P < 0.05) on Days -7 to -4, and increased (resurged, P < 0.05) on Days -4 to -3. A transient increase in FSH was maximum on mean Day -4, and the peak of a minor FSH surge occurred on Day -4.5 ± 0.2. Concentration of LH and estradiol increased between Days -5 and -4. Results demonstrated resurgence of the preovulatory follicle apparently for the first time in any species. Resurgence seemed more related temporally to the minor FSH surge than to the LH increase, but further study is needed. Results supported the novel hypotheses that a minor FSH surge near the end of the IOI is temporally associated with (1) the emergence of a minor follicular wave and (2) the resurgence in growth rate of the preovulatory follicle.

Hormone concentrations temporally associated with contralateral and ipsilateral relationships between the CL and preovulatory follicle during the third follicular wave in heifers.

Concentrations of circulating hormones after Day 14 (Day 0 = ovulation) were determined daily in 87 interovulatory intervals (IOIs) in heifers. The IOIs were grouped into four permutations according to an ipsilateral (Ipsi) or contralateral (Contra) relationship between the CL and the preovulatory follicle and two (2W) or three (3W) follicular waves per IOI. The number of IOIs per group differed (P < 0.005) from equality among the Ipsi-2W (n = 27), Contra-2W (n = 31), Ipsi-3W (n = 9), and Contra-3W (n = 20) groups. A continuous decrease in progesterone (luteolysis) began later (P < 0.05) in the Contra-3W group (Day 18.0 ± 0.4) than in each of the Ipsi-2W (15.4 ± 0.2), Contra-2W (15.6 ± 0.2), and Ipsi-3W (16.2 ± 0.5) groups. Concentrations of LH and estradiol began to increase near the beginning of luteolysis in each group. A minor FSH surge that did not stimulate a major follicular wave developed in about 50% of the IOIs in each group, except that none were detected in the Ipsi-3W group. The minor FSH surge reached a peak about 4 days before ovulation and several days after wave 3 had emerged. The hypothesis that luteolysis begins earliest in two-wave IOIs, intermediate in three-wave IOIs with an ipsilateral CL/follicle relationship, and latest in three-wave IOIs with a contralateral relationship was supported. The hypothesis that a minor FSH surge occurs most frequently in association with three follicular waves was not supported.

Two-way coupling between FSH and the dominant follicle in heifers.

Follicular Wave 1 and 2 and the associated FSH Surge 1 and 2 were used to designate the first two waves and surges of the interovulatory interval in two experiments in heifers. In experiment 1, a group with early (group E, N = 9) and late (group L, N = 5) development of the dominant follicle of Wave 1 were used as natural models to study FSH/follicle coupling. The day of wave emergence and the day of deviation in diameters between the two largest follicles were not different between groups. Emergence of Wave 2 and maximal FSH concentration in Surge 2 was approximately 1 day later (P < 0.03) in group L. Diameter of the dominant follicle of wave 1 (13.8 ± 0.3 mm vs. 12.0 ± 0.3 mm) and FSH concentrations in Surge 2 (0.29 ± 0.02 ng/mL vs. 0.21 ± 0.03 ng/mL) were first greater (P < 0.05) in group E than in group L at 4 and 5 days, respectively, after wave emergence. In experiment 2, treatment with estradiol (N = 8) when the dominant follicle of Wave 1 was ≥ 11 mm (Hour 0) resulted in a decrease (P < 0.02) in FSH and slower (P < 0.05) growth rate of the follicle between Hours 0 and 4. Results supported the following hypotheses: (1) the FSH surge that stimulates emergence of a follicular wave is associated with final growth of the dominant follicle of the previous anovulatory wave; and (2) suppression of FSH Surge 2 when the dominant follicle of Wave 1 is ≥ 11 mm is associated with a decrease in diameter. It is concluded for the first time that two-way FSH/follicle coupling in heifers continues during final growth of the dominant follicle of Wave 1 and that Surge 2 is the FSH source.

Endocrinology of number of follicular waves per estrous cycle and contralateral or ipsilateral relationship between corpus luteum and preovulatory follicle in heifers.

A 3-d extension of the luteal phase occurs in interovulatory intervals (IOIs) with a contralateral relationship between the corpus luteum (CL) and preovulatory follicle with 3 follicular waves (Contra-3W group). Concentrations of FSH, progesterone, LH, and estradiol-17ß for the ipsilateral versus contralateral CL and/or follicle relationship and 2 versus 3 waves per IOI were studied in 14 heifers. Follicular waves and FSH surges were designated 1, 2, or 3, according to order of occurrence in the IOI. The day (day 0 = ovulation) of the FSH peak in surge 2 occurred earlier (P < 0.02) in 3-wave IOIs (day 6.3 ± 0.5) than in 2-wave IOIs (day 8.5 ± 0.5). Mean FSH was higher in 3-wave than in 2-wave IOI on 82% of the days in the IOI. Repeatability or individuality in FSH concentration was indicated by a correlation (r = 0.54, P < 0.04) in FSH concentrations between ovulations at the beginning and at the end of the IOI. Concentrations of LH and estradiol increased (P < 0.05) near the beginning of the luteolytic period in 2-wave IOI regardless of the CL and/or follicle relationship. In the Contra-3W group, LH and estradiol remained at basal concentrations concurrently with FSH surge 3 and extension of the luteal phase. The hypotheses were supported that FSH surge 2 occurs earlier in 3-wave IOIs than in 2-wave IOIs and that the development of 3-wave IOIs occurs in individuals with greater FSH concentrations. Extension of the luteal phase in the Contra-3W group was temporally associated with lower concentrations of LH and estradiol.

Length of the luteal phase and frequency of the four permutations between two or three follicular waves and ipsilateral or contralateral locations of the corpus luteum and preovulatory follicle in heifers.

In Survey 1, the records for 196 interovulatory intervals (IOIs) from 24 heifers were used to study the frequency and repeatability for number of follicular waves per IOI and to study the ipsilateral and contralateral relationships between the extant corpus luteum and ovulatory follicle. In Survey 2, 96 IOIs were used from the controls of the previous experiments that included the day of the end of the luteolytic period (progesterone <1.0 ng/mL). In Survey 1, the percentage of two-wave IOIs (63%) was greater (P < 0.0002) than for three-wave IOIs (37%). The percentage of ovulatory periods with a contralateral relationship (59%) was greater (P < 0.003) than with the ipsilateral relationship (41%). There were more repeats (66%) than reverses (34%) between adjacent IOIs in number of waves per IOI (P < 0.004), but there was no difference in number of corpus luteum/follicle relationships between the ovulatory periods at the beginning and end of an IOI. For the four permutations of ipsilateral and two waves, contralateral and two waves, ipsilateral and three waves, and contralateral and three waves in Survey 2, the interval (days) from ovulation to the day the progesterone was <1.0 ng/mL was 17.8 ± 0.2(a), 17.6 ± 0.2(a), 20.0 ± 0.3(b), and 21.4 ± 0.3(c), respectively, and the number of IOIs was 33 (34%)(a), 34 (35%)(a), 8 (8%)(b), and 22 (23%)(a), respectively; means with different superscripts are different (P < 0.05). The luteal phase was longer for the contralateral relationship than for the ipsilateral relationship for three-wave IOIs but not for two-wave IOIs. The hypothesis was supported that the frequency of the ipsilateral and three-wave permutation was less than for each of the other three permutations.

Interrelationships among progesterone, LH, and luteal blood flow during a pulse of a PGF2α metabolite and functional role of LH in the progesterone rebound in heifers.

On Day 16 (Day 0 = ovulation) or before the expected transition into the luteolytic period, heifers were not treated (control group, N = 7) or were treated with a single 0.1-mg dose of estradiol (E2) (E2 group, N = 6) or E2 combined with the GnRH antagonist acyline (E2/Ac group, N = 5). Hourly blood samples were collected from hour of treatment (Hour 0) to Hour 20. Estradiol induced a pulse of PGFM, but the peak of the pulse occurred 2 hours later (P < 0.05) and mean PGFM concentrations during the descending portion of the pulse were lower (P < 0.05) in the E2/Ac group than in the E2 group. In the E2 group, concentration of progesterone (P4) decreased (P < 0.05) during the ascending portion of the PGFM pulse, and increased (rebounded; P < 0.05) along with an LH increase during the descending portion. In the E2/Ac group, a rebound in P4 and an increase in LH were not detected during the descending portion of the PGFM pulse. The percentage of CL with color Doppler signals of blood flow increased (P < 0.04) concurrently with the PGFM increase during Hours 0 to 5 and during the ascending portion of the PGFM pulse. Blood flow and PGFM decreased concurrently. The following hypotheses were supported: (1) LH has a positive effect on PGFM pulses; (2) the rebound in P4 and the increase in LH during the descending portion of a PGFM pulse are functionally related; and (3) the increase in luteal blood flow in association with a PGFM pulse represents a direct effect of PGF2α rather than an effect of P4 or LH.

Contralateral ovarian location between the future ovulatory follicle and extant corpus luteum increases the length of the luteal phase and number of follicular waves in heifers.

Location of the future ovulatory follicle and CL in the same ovary (ipsilateral) or opposite ovaries (contralateral) and number of major follicular waves (two or three) per interovulatory interval (IOI) was studied in 14 heifers. Follicle diameter and a blood sample for progesterone (P4) assay were obtained each day throughout an IOI. Heifers were partitioned into three groups: ipsilateral follicle/CL relationship and two follicular waves (Ipsi-2W, N = 5), contralateral relationship and two follicular waves (Contra-2W, N = 5), and contralateral relationship and three waves (Contra-3W, N = 4). Only one heifer had an ipsilateral relationship and three waves and was not included in the analyses. An unexpected observation was slower growth of the dominant follicle of Wave 1 in the Ipsi-2W group than in the Contra-2W and Contra-3W groups. Increased P4 production in the Contra-3W group compared with the Ipsi-2W and Contra-2W groups was indicated by significantly greater P4 concentration averaged over Days 0 to 20 (Day 0 = ovulation), longer interval from ovulation to the beginning of a decrease in P4 and to the beginning of postluteolysis (P4 <1 ng/mL), and longer IOI. The interval from the beginning of postluteolysis to ovulation was not different among groups, indicating that the prolonged IOI reflected the prolonged luteal phase. An effect of the follicle/CL relationship on length of the IOI was not detected in mares. Results supported the hypothesis that the prolonged luteal phase of the contralateral follicle/CL relationship favors the development of three follicular waves/IOI in heifers.