concentra-Effects of additional gonadotropin-releasing hormone and prostaglandin F 2α treatment to an estradiol/progesterone-based embryo transfer protocol for recipient lactating dairy cows

This study was designed to evaluate whether the utilization of a second PGF 2α treatment at the end of an estradiol/progesterone (E2/P4)-based protocol with or without GnRH at the beginning of the protocol would improve pregnancy rates of lactating Holstein cows assigned to timed embryo transfer. A total of 501 lactating Holstein cows in 5 farms were enrolled in the experiment. Within farm, cows were blocked by parity and, within block, were assigned randomly to (1) insertion of an intravaginal P4 device (controlled internal drug-releasing device; CIDR) and estradiol benzoate on d −11, PGF 2α on d −4, CIDR withdrawal and an injection of estradiol cypionate on d −2, and timed embryo transfer on d 7 (1-PGF; n = 164); (2) the same treatments as 1-PGF, but with PGF 2α administered on d −4 and −2 (2-PGF; n = 171); and (3) 2-PGF with the addition of a GnRH treatment on d −11 (GnRH + 2-PGF; n = 166). Ovaries were scanned by transrectal ultrasonography on d −11, −4, and 7, and blood samples were collected on d −11, −4, 0, and 7 for P4 determination. Treatment comparisons were performed using contrasts. The proportion of cows with a new corpus luteum on d −4 was greater in GnRH + 2-PGF cows. Cows in 1-PGF had a greater P4 concentration on d 0 but lesser P4 on d 7 compared with cows in the other groups. Cows assigned to receive 2-PGF (2-PGF and GnRH + 2-PGF) had greater estrus expression, and a greater proportion of cows ovulated to estradiol cypionate. No further contrast effects were observed for follicle diameter, double ovulation rate, pregnancy per embryo transfer (P/ET) on d 32 and 60, or pregnancy loss. As P4 concentration on d −4 increased, P/ET on d 60 tended to increase. Cows with P4 ≥3.66 ng/mL on d −4 had greater P/ET on d 32 and 60 than those with P4 below that threshold. Regardless of treatment, cows with P4 concentration ≥3.66 ng/mL also had a greater pregnancy per synchronized protocol (P/SP) on d 60. Also, a P4 concentration on d −4 (low or high) × follicle diameter (continuous) interaction tendency was observed when evaluating P/ET. Although P/ET did not differ among cows with different follicles sizes with reduced P4 concentration on d −4 (<3.66 ng/mL), it increased in cows with larger follicles exposed to increased P4 concentration (≥3.66 ng/mL). When P4 on d 0 was evaluated, P/ET on d 32 and 60 was greater for cows with low (≤0.09 ng/mL) versus high (>0.21 ng/ mL) P4; as P4 concentration on d 0 increased, P/ET linearly decreased. In summary, cows with increased P4 concentrations during growth of the ovulatory follicular wave had improved P/ET. Administering a second PGF 2α dose reduced P4 concentration on d 0 and increased ovulatory response to the protocol, but no benefits were observed on P/ET or P/SP.


INTRODUCTION
Timed embryo transfer (TET) is an effective tool to increase reproductive efficiency in lactating dairy cows (Vasconcelos et al., , 2011Demetrio et al., 2007). Timed ET protocols may bypass the negative effects of extrinsic and intrinsic factors, such as heat stress or prolonged follicle growth on oocyte and early embryo development, by allowing all cows to receive a goodquality embryo. Moreover, in vitro-produced embryos allow the use of ET on a large scale, but it is not clear whether the fertility of recipient lactating dairy cows is dependent on factors previously reported for timed AI (TAI) programs, such as progesterone (P4) concentra-Effects of additional gonadotropin-releasing hormone and prostaglandin F 2α treatment to an estradiol/progesterone-based embryo transfer protocol for recipient lactating dairy cows tion during ovulatory follicle development (Bisinotto et al., 2010;Martins et al., 2011;Wiltbank et al., 2012), proestrus length (Peters and Pursley, 2003;Pereira et al., 2013bPereira et al., , 2014, circulating P4 concentration on d 0 of the estrous cycle (Pereira et al., 2013b(Pereira et al., , 2015, and increased P4 concentrations on the day subsequent to TAI (Demetrio et al., 2007).
Circulating P4 concentration (Pereira et al., 2015(Pereira et al., , 2017b or the presence of a corpus luteum (CL; Pereira et al., 2013a) during follicular growth has been associated with improved fertility in TAI protocols, which in turn is often associated with improved embryo quality (Rivera et al., 2011;Wiltbank et al., 2014). However, few studies have evaluated the effects of circulating P4 concentrations in TET protocols for recipient cows. Recently, our group demonstrated that supplementing P4 with 2 intravaginal inserts at the beginning of an estradiol (E2)/P4-based protocol in cows not in diestrus (P4 <1.0 ng/mL) increased pregnancy per AI (P/AI), whereas no benefit was observed with a similar treatment in cows receiving TET (Pereira et al., 2017a). Slight elevations in circulating P4 concentrations near AI (≥0.4 ng/mL) are associated with a reduction in P/AI (Brusveen et al., 2009;Martins et al., 2011) or pregnancy per embryo transfer (P/ET; Pereira et al., 2013b). Experiments have investigated the effect of additional treatments or amount of PGF 2α on luteolysis and P/AI in Ovsynch-type protocols Giordano et al., 2013b;Pereira et al., 2015); however, the same data are not available for the use of additional PGF 2α in E2/P4-based TET protocols.
Previously, cows that did not show estrus had reduced pregnancy per TAI or TET compared with cows detected in estrus (Pereira et al., 2016). Moreover, cows with a CL on the day of the final PGF 2α administration of a TAI protocol had greater P/AI, which was associated with increased follicle diameter, although no association was observed between follicle diameter and P/AI in cows without a CL on the day of PGF 2α treatment (Pereira et al., 2015). Pereira et al. (2020) demonstrated that P/AI linearly decreased as follicle diameter on the day of TAI increased in cows with a P4 concentration <3.66 ng/mL on the day of PGF 2α administration. The authors speculated that the development of larger follicles might benefit P/AI when growing under high concentrations of P4 . Moreover, GnRH administration at the beginning of a TET protocol might increase the proportion of cows with a CL at PGF 2α administration, potentially favoring P/ET in cows with follicles of large diameter. Based on this rationale, we hypothesized that (1) a second PGF 2α treatment at the end of the protocol would improve luteolysis and cause further reductions in P4 concentration near ovulation, benefiting pregnancy rates; and (2) the addition of a GnRH treatment at the beginning of an E2/P4-based TET protocol would induce the formation of a new CL in the cycle preceding TET. Therefore, the objectives of this experiment were to evaluate whether (1) the addition of a second PGF 2α treatment at the end and (2) an additional GnRH treatment at the beginning of an E2/P4 TET protocol would modify P4 concentrations during the growth of the preovulatory follicle, ovarian dynamics, and affect P/ET in lactating dairy cows.

MATERIALS AND METHODS
This experiment was conducted in 5 commercial dairy farms located in Paraná, Brazil, and 1 commercial dairy farm located in Minas Gerais, Brazil, from July 2012 to January 2013. All procedures with cows followed the recommendations of the Guide for the Care and Use of Agricultural Animals in Research and Teaching (FASS, 2010) and have been approved by the São Paulo State University Institutional Animal Care and Use Committee.
In all farms and throughout the experiment, cows were housed in freestall barns, with access to an adjoining sod-based paddock, and were milked 3 times daily. All procedures including injections, ovarian ultrasonography, pregnancy diagnosis, blood collection, and TET were performed while cows were restrained in self-locking head gates at the feedline. Cows had ad libitum access to a TMR based on corn silage and Tifton-85 (Cynodon dactylon) hay as forage sources and a concentrate composed of corn, soybean meal, citrus pulp, whole cottonseed, minerals, and vitamins. The TMR was balanced to meet or exceed the nutritional requirements of lactating dairy cows (NRC, 2001).

Ovarian Ultrasonography
Ovaries were evaluated by transrectal ultrasonography (Aloka SSD-500 with a 7.5-MHz linear-array transducer, Aloka) on d −11, −4, and 7 to determine the presence of a CL, as well as on d 0 to measure the diameter of the largest follicle. A new CL on d −4 was defined as the presence of a CL on d −4 that was not visualized on d −11. The diameter of the ovulatory follicle was determined by the largest follicle present on the ovary on d 0 that corresponded to an observed CL on the same ovary on d 7. Cows with follicles <8 mm on d 0 but with a CL on d 7 were defined as "early ovulatory cows" and were not used in the analyses of ovulatory follicle diameters.

Estrus Detection
Estrus detection patches (Estrotect; Rockway Inc.) were placed on the tail head of the animals. The estrus detection patches were not used to alter the timing of AI because all cows were bred to TET by study personnel during the experimental period. The patches were visually scored on d 0 by research personnel, and cows with a ≥50% activated patch were considered to be expressing estrus between d −2 and 0 of the protocol (Franco et al., 2018).

Ovulation to ECP, P/ET, and Pregnancy Loss
Cows with a CL present on d 7 were defined as having ovulated to the ECP and received TET. Therefore, P/ET was calculated by dividing the number of pregnant cows at the pregnancy diagnosis on d 32 and 60 of gestation by the number of cows that received TET. Pregnancy loss was calculated by dividing the number of cows that lost their pregnancy between 32 and 60 d of gestation by the number of pregnant cows on d 32. Pregnancy per synchronization protocol (P/SP) was calculated by dividing the number of pregnant cows by the number of cows enrolled to each treatment.

Sample Collection
Milk production was measured daily from d 0 to 7, and mean daily production in this period was used for further analysis. Blood samples were collected on d −11 Pereira et al.: ADDING GnRH AND PGF 2α TO AN EMBRYO TRANSFER PROTOCOL (n = 463), −4 (n = 481), 0 (n = 494), and 7 (n = 499) by coccygeal venipuncture into commercial blood collection tubes (Vacutainer; Becton Dickinson). Following blood sample collection, tubes were immediately placed on ice, maintained at 4°C for 12 h, and centrifuged at 1,500 × g for 15 min at room temperature for serum collection. Serum was stored at −20°C for subsequent P4 analysis. Serum concentrations of P4 were quantified by RIA using the Coat-A-Count solid phase 125 I radioimmunoassay kit (Diagnostic Products Inc.) that had previously been validated for bovine samples . The intra-and interassay coefficients of variation were 5.6 and 10.2%, respectively, and the sensitivity of the assay was 0.01 ng/mL, calculated as 2 standard deviations below the mean counts per minute at maximum binding, as described previously Pereira et al., 2013a).

Statistical Analysis
A power calculation was performed using P/AI on d 60 as the primary goal of the present trial; 137 cows per group would be required to detect a 10-percentagepoint difference in this variable. For all analyses performed herein, cow was used as the experimental unit, analyzed as a completely randomized block design, and the Satterthwaite approximation was used to determine the denominator degrees of freedom for tests of fixed effects.
The binomial variables (new CL at PGF 2α on d −4, proportion of cows with a CL at PGF 2α on d −4, proportion of cows detected in estrus, ovulation to ECP, double ovulation, P/ET on d 32 and 60, and pregnancy loss between d 32 and 60) were analyzed using the GLIMMIX procedures of SAS (version 9.4; SAS Institute Inc.) with farm as a random effect. The initial model contained the effects of treatment, parity, and the resulting interaction, whereas DIM, BCS, number of previous uses of the intravaginal P4 device, and milk yield were included as covariables in all analyses. The continuous dependent variables for each specific day (P4 concentrations on d −11, −4, and 7, and follicular diameter) were analyzed separately using the GLIM-MIX procedures of SAS (version 9.4; SAS Institute Inc.), and variables included in the models were treatment, parity, and the resulting interaction; DIM, BCS, and milk yield were included as covariables, and farm was included as a random effect. In all treatment comparisons, orthogonal contrasts were performed as follows: (1) effect of second PGF 2α administration: 1-PGF versus 2-PGF (2-PGF and GnRH + PGF), and (2) effect of GnRH administration: 2-PGF versus GnRH + PGF.
In a previous experiment, the circulating P4 concentration on d −4 of the synchronization protocol that resulted in the greatest combined sensitivity and specificity for fertility responses was ≥3.66 ng/mL (Pereira et al., 2015). Thus, additional statistical analyses were performed with cows classified as having a low (<3.66 ng/mL) or high (≥3.66 ng/mL) P4 concentration on d −4 to evaluate the associations between P4 concentration and P/ET. The effects of P4 concentration on d −4 (above or below 3.66 ng/mL) were also analyzed as interactions with treatments to evaluate the resulting interaction. The data were analyzed using the GLIM-MIX procedure of SAS, with farm as the random effect; the means were adjusted using the TUKEY statement. Similarly, cows were categorized using P4 concentration on d 0 based on results from previous experiments that observed the greatest combined sensitivity and specificity for P/AI (P4 ≤0.09 ng/mL) and P/ET (P4 ≤0.21 ng/mL) (Pereira et al., 2013b). Hence, P4 concentration on d 0 was classified as low (≤0.09 ng/mL), medium (0.10-0.21 ng/mL), or high (>0.21 ng/mL) to investigate the association with P/ET.
The GLIMMIX procedure of SAS (version 9.4; SAS Institute Inc.) was used to determine whether each individual measurement (follicle diameter on d 0, estrus expression, and P4 concentration on d −4 of synchronization protocol) influenced P/ET in a linear, quadratic, or cubic manner. The LOGISTIC procedure of SAS (version 9.4; SAS Institute Inc.) was used to determine the intercept and slope(s) value(s) according to maximum likelihood estimates from each significant continuous order effect, and the probability of pregnancy was determined. Nonlinear regression curves were constructed using the minimum and maximum values detected for each individual measurement. Covariables with P > 0.10 were removed from the statistical models. For all continuous variables analyzed, results are expressed as least squares means ± standard error of the mean. In all analyses, differences were considered significant when P ≤ 0.05 and tendencies when 0.05 < P ≤ 0.10.

RESULTS
Treatment with GnRH increased (P < 0.01) the proportion of cows with a new CL on the day of first PGF 2α administration (d −4), regardless of whether the analysis contained all cows with CL or only those that had new CL on d −4 of the protocol (Table 1). In both analyses, a greater proportion of cows assigned to receive the treatment that contained GnRH at the beginning of the protocol had a new CL on d −4 compared with those assigned to either 1-PGF (P < 0.01) or 2-PGF (P < 0.01; Table 1). Additionally, no treatment effects were observed on P4 concentrations analyzed on d −11 (P = 0.52) or −4 (P = 0.48) of the synchronization protocol (Table 1). On d 0, cows assigned to 1-PGF had a greater P4 concentration compared with cows in 2-PGF (P < 0.01) and GnRH + 2-PGF (P = 0.02), whereas opposite results were observed on d 7 of the experiment when all cows were included in the analysis (P < 0.01) or when only cows that ovulated to the protocol were considered (P = 0.07; Table 1).
Cows assigned to receive 2-PGF (2-PGF and GnRH + 2-PGF) had greater estrus expression than those assigned to 1-PGF (P = 0.04), whereas no differences were observed between 2-PGF and GnRH + 2-PGF (P = 0.60). Similarly, a greater proportion of cows ovulated to ECP following 2-PGF administration (P < 0.01), but the effect of GnRH was not significant (P = 0.48; Table 1). No further contrast effects were observed on follicle diameter (P = 0.52), double ovulation rate (P ≥ 0.14), P/ET on d 32 (P ≥ 0.52) and 60 (P ≥ 0.49), or on pregnancy loss (P ≥ 0.48; Table 1). Regardless of treatment, as follicle diameter measured on d 0 increased, P/ET on d 60 tended to decrease (P = 0.09). A P4 concentration on d −4 (low or high) × follicle diameter (continuous) interaction tendency (P = 0.10) was observed on P/ET at d 60 ( Figure 2). Pregnancy per ET on d 60 did not differ among P4 on d −4 in cows having follicles <16 mm on d 0, but P/ET increased in cows having a follicle diameter ≥16 mm as P4 concentration increase on d −4 (linearly; Figure 2). Table 2 reports the effect of estrus expression on various parameters of reproductive physiology. Circulating P4 concentration on d −4 did not differ (P = 0.78) between cows that expressed estrus or not. Cows detected in estrus had a larger ovulatory follicle (P = 0.02), reduced P4 concentration on d 0 (P = 0.01), greater P4 concentration on d 7 (P < 0.01), ovulation rate (P < 0.01), and P/ET on d 32 and 60 (P ≤ 0.01), whereas pregnancy loss was not affected (P = 0.88; Table 2). As follicle diameter on d 0 increased in cows that did not express estrus, P/ET linearly decreased (P = 0.02); the same effect was not observed in cows that expressed estrus (P = 0.30; Figure 3). Pregnancy loss between d 32 and 60 increased as follicle diameter on d 0 increased in cows that did not express estrus (linear effect; P = 0.04), and the same effect tended to be observed in cows that expressed estrus (P = 0.06; Figure 4). Table 3 reports the effects of d −4 P4 concentration (< or ≥3.66 ng/mL) on reproductive variables of lactating dairy cows. Progesterone concentration on d −4 did not affect circulating P4 concentration on d 0 (P = 0.38) or d 7 (P = 0.18). Nonetheless, dairy cows with reduced P4 concentration on d −4 (<3.66 ng/mL) had a larger (P < 0.01) ovulatory follicle. Regardless of treatment, P4 concentration on d −4 did not affect the incidence of ovulation and pregnancy loss (P ≥ 0.36), but greater P4 concentration on d −4 resulted in Pereira et al.: ADDING GnRH AND PGF 2α TO AN EMBRYO TRANSFER PROTOCOL  greater P/ET on d 32 and 60 (P ≤ 0.05), and greater P/SP on d 60 (P = 0.04; Table 3). Similarly, as P4 concentration on d −4 increased, P/ET and P/SP on d 60 tended to increase in a linear manner (P = 0.08; Figure  5). Cows that expressed estrus and had P4 ≥3.66 ng/ mL on d −4 had a greater P/ET on d 60 than low-P4 cows expressing estrus (35.0 vs. 49.8%, respectively; P = 0.03), but no effects of P4 concentration difference were observed on P/ET of cows that did not express estrus (19.6 vs. 19.8%, respectively; P = 0.98). In the same analysis, treatment did not affect the distribution of cows within each P4 class, whereas ovulation rate was affected by treatment (P < 0.01). Cows assigned to 2-PGF (P < 0.01) and GnRH + 2-PGF (P = 0.04) had a greater ovulation rate than cows in 1-PGF, but no differences were observed between 2-PGF and GnRH + 2-PGF (P = 0.48; Table 4). Last, P/ET tended to be greater (P = 0.06) and P/SP was greater (P = Pereira et al.: ADDING GnRH AND PGF 2α TO AN EMBRYO TRANSFER PROTOCOL Figure 2. Logistic regression analysis of the relationship between progesterone (P4) concentration on d −4 (time of first PGF 2α administration) and pregnancy per embryo transfer (P/ET) for cows that had a follicle <16 or ≥16 mm on d 0. A P4 concentration on d −4 × follicle diameter interaction tendency (P = 0.10) was observed.  (Table 4). A similar approach was performed using P4 concentration on d 0 of the protocol, in which cows were classified as having low (≤0.09 ng/mL), medium (0.10-0.21 ng/mL), and high (>0.21 ng/mL) P4 concentration. A greater proportion of cows assigned to 1-PGF was classified into the high-P4 group compared with cows assigned to 2 PGF 2α doses (2-PGF and GnRH + 2-PGF; P < 0.01), but no further differences were observed when cows classified into the low-and medium-P4 groups were analyzed (P ≥ 0.13; Figure  6). Animals in the low-P4 (P < 0.01) and medium-P4 (P = 0.02) groups had greater estrus expression rates than those having a high P4 concentration on d 0 of the protocol (Table 5). Additionally, P4 concentration on d −4 was lower for cows having low (P < 0.01) P4 on d 0 compared with cows classified into the medium-and high-P4 groups, but P4 on d 7 was not affected by previous P4 concentration on d 0 (P = 0.26). The P/ET on d 32 and 60 was greater for lowversus high-P4 (P = 0.02) cows, whereas medium-P4 cows were intermediate and did not differ from low-P4 cows (P = 0.25) or high-P4 cows (P = 0.20; Table 5). In agreement, as P4 concentration on d 0 increased, P/ET on d 60 decreased linearly (P = 0.02; Figure  7), but no further effects (linear, quadratic, or cubic) were observed on P/ET when P4 concentration on d 7 was analyzed (P = 0.77; Figure 8).

DISCUSSION
The goal of the present study was to evaluate how (1) the addition of a second dose of PGF 2α and (2) the use of a GnRH treatment at the beginning of the protocol would affect the reproductive function of lactating dairy cows. Our results provide practical knowledge into how physiological and reproductive performance of embryo recipients lactating dairy cows might be managed to improve reproductive performance under specific TET protocols. One of the major practical insights is that the provision of 2 PGF 2α doses, on d −4 and −2 of the synchronization protocol, increased estrus expression and ovulation rate to ECP, which will benefit TET protocols, with more recipient cows being able to receive an embryo at TET. Another interesting result was that dairy cows with increased circulating P4 concentration during the estrous cycle preceding ET had greater P/ ET. Important physiological insights were obtained when fertility data were analyzed in relation to specific measures of reproductive physiology, including mean P4 concentration on d −4, 0, and 7, estrus expression, and diameter of the ovulatory follicle.
Circulating P4 concentration on d 0 had a dramatic effect on P/ET, as cows with a greater P4 concentration on d 0 had a reduced P/ET. In a previous experiment, Pereira et al. (2013a) used different circulating P4 values to define complete luteolysis and reported that cows bred by TAI had greater fertility if they had  Madureira et al. (2022) demonstrated that dairy cows with reduced P4 on day of AI had greater estrus expression and P/AI compared with cows having greater P4 at AI. In the present experiment, when P4 concentration on d 0 was categorized, cows with low P4 (≤0.09 ng/mL) showed more estrus and had greater pregnancy success on d 32 and 60 than cows with P4 >0.21 ng/mL. Therefore, based on previous and current data, P4 concentration during follicular development is important for embryo recipients, suggesting that the factors determining high fertility in lactating dairy cows with higher P4 are associated not only with higher embryo quality, but also with the uterine environment during follicular development, leading to higher embryonic maintenance in lactating cows submitted to AI or ET.
Based on a meta-analysis of 6 experiments, Borchardt et al. (2018) evaluated the effects of an additional PGF 2α treatment during an Ovsynch protocol on luteolysis and P/AI of lactating dairy cows. The addition of a PGF 2α treatment positively affected luteal regression and improved P/AI by 4.6 percentage points (Borchardt et al., 2018). It is worth mentioning that in the present experiment, P/ET on d 0 was increased by 4.4 percentage points when 1-PGF and 2-PGF (2-PGF and GnRH + 2-PGF) were compared (34.1 vs. 38.5%, respectively). Nevertheless, there was a clear effect of Pereira et al.: ADDING GnRH AND PGF 2α TO AN EMBRYO TRANSFER PROTOCOL Figure 4. Logistic regression analysis of the relationship between follicle diameter (d 0) and pregnancy loss for cows that expressed or did not express estrus. the second PGF 2α treatment on ovulation rate that increased this rate by 11.5 percentage points (1-PGF: 80.9 vs. 2-PGF: 92.4%); ovulation rate is a very important parameter to be evaluated in synchronization protocols, as it strongly correlates with the proportion of cows that are able to receive an embryo on d 7. Lopes et al. (2020)   . Those authors reported that the second PGF 2α treatment increased P/AI, altered the duration of the LH surge, and increased circulating concentration and total follicular estradiol, suggesting that the benefits of PGF 2α on fertility are not related to its known luteolytic effects, but might be related to changes in the release of GnRH and the endocrine milieu of the preovulatory follicle. Moreover, we can speculate that physical activity also plays a role in the reproductive performance of lactating dairy, as it has been correlated with spontaneous estrus intensity (Madureira et al., 2015;Burnett et al., 2017Burnett et al., , 2018Silper et al., 2017) and ovulation rate (Madureira et al., 2015). However, the relationships, if any, among physi-Pereira et al.: ADDING GnRH AND PGF 2α TO AN EMBRYO TRANSFER PROTOCOL Figure 6. Effects of treatments on distribution of cows according to progesterone (P4) concentration on d 0 of the protocol in lactating dairy cows with a corpus luteum (CL) at PGF administration that were assigned to receive 1 of 3 treatments: (1) controlled internal drug-releasing device (CIDR) and 2.0 mg of estradiol benzoate (EB), PGF 2α on d −4, CIDR removal and 1.0 mg of estradiol cypionate (ECP) on d −2 (1-PGF); (2) 1-PG protocol + another PGF 2α dose on d −2 (2-PGF); and (3) 2-PGF protocol + GnRH administration on d −11 (GnRH + 2-PGF). For all treatments, timed embryo transfer was performed on d 7. cal activity, estrus intensity, and P4 concentration on d 0 and 7 in TET protocols have not been reported and warrant further investigation. Therefore, P4 concentration on d 0 should be reduced upon the second PGF 2α treatment, and it becomes clear with the current and previous data (Pereira et al., 2015) that the benefits of second PGF 2α treatment involve such a reduction. The use of GnRH is a strategy to synchronize the follicular wave at the beginning of the protocol and to synchronize the time of ovulation near TAI (Pereira et al., 2015(Pereira et al., , 2017b. Nonetheless, in countries such as Brazil, E2 and P4 are broadly used to synchronize the follicular wave and time of ovulation in lactating dairy cows (Pereira et al., 2013a(Pereira et al., , 2015(Pereira et al., , 2017b(Pereira et al., , 2021Vasconcelos et al., 2018). Moreover, E2/P4-based protocols for TAI are associated with a greater circulating P4 concentration at time of PGF 2α administration, low circulating P4 concentration at TAI, and greater estrus expression (Pereira et al., 2013b(Pereira et al., , 2015(Pereira et al., , 2016. However, few studies have evaluated strategies to synchronize the estrus cycle of lactating embryo dairy recipients during a TET protocol (Pereira et al., 2013b(Pereira et al., , 2017a. In a similar experimental design, Pereira et al. (2015) reported that addition of a GnRH treatment at the beginning of a TAI protocol increased the proportion of cows with a new CL on d −4 and, consequently, yielded different P4 concentrations on d −4 of the protocol and improvements in the reproductive performance of lactating dairy cows. It was speculated that the pres-ence of a CL during the synchronization protocol (d −4) would increase circulating P4 concentration during the period of follicle development, which has been reported in Ovsynch-based protocols with concurrent improvements in P/AI (Bisinotto et al., 2010(Bisinotto et al., , 2013Denicol et al., 2012) and greater embryo quality (Cerri et al., 2011a;Rivera et al., 2011). Nonetheless, in the present experiment, P/ET on d 32 and 60 did not differ between treatments. However, an additional analysis demonstrated that, regardless of treatment, cows with greater P4 concentration on d −4 had greater P/ET on d 32 and 60, and greater P/SP on d 60. In TAI protocols, the effects of increased circulating P4 concentration at the moment of PGF 2α administration have been associated with an improved condition of the ovulated oocyte and subsequent benefits in embryo quality during the first week of embryo development . Specifically, for TET, Cerri et al. (2011b) collected embryos 6 d after AI in cows submitted to protocols with low or high P4 and reported no effect of P4 on fertilization rate, but a tendency for greater embryo quality was observed in cows receiving high-P4 protocols during preovulatory follicle development. Similarly, Rivera et al. (2011) reported an increased proportion of transferrable embryos (78.6 vs. 55.9%) when donor cows that were superovulated during the first follicular wave (low P4) received 2 intravaginal P4 devices during the synchronization protocol. In contrast, Pereira et al. (2017b)  tion of 2 intravaginal P4 devices in an E2/P4-based protocol in cows with low P4 concentrations (<1.0 ng/ mL) at the beginning of the protocol resulted in increased P4 concentrations (1-CIDR = 1.77 ± 0.23 vs. 2-CIDR = 2.18 ± 0.24 ng/mL), whereas no effects were observed in TET (Pereira et al., 2017a). These results suggest that P4 concentration during follicle development must be greater than that achieved (≥2 ng/mL) herein and in previous studies using TAI (Pereira et al., 2015), indicating that P4 concentration ≥3.66 ng/mL at PGF 2α results in greater P/AI and P/ET. A direct fertility analysis was consistent with this statement because cows with circulating P4 concentration ≥3.66 ng/mL had a d 60 P/AI 11.3% greater than cows with P4 <3.66 ng/mL (42.9 vs. 31.6%). In agreement with our data, Pereira et al. (2017b) reported that, regardless of treatment and heat stress occurrence, cows with P4 ≥3.66 ng/mL had 50% more P/AI than cows with a P4 concentration below this threshold. Nonetheless, GnRH treatment did not benefit P/ET herein and it is likely that strategies to increase P4 to higher levels are necessary to improve P/ET, such as the utilization of 2 P4 devices or presynchronization protocols .
The fact that cows with larger follicles on d 0 also tended to have a reduced P/ET on d 60 has been reported by others when TAI and TET protocols based on E2 and P4 were evaluated (Pereira et al., 2016. In the present experiment, an interaction of P4 on d −4 × follicle diameter tended to be detected, such that in cows with P4 <3.66 ng/mL at PGF 2α , follicle diameter had no effect on P/ET, whereas smaller fol-licles were ovulated in cows with P4 ≥ 3.66 ng/mL, resulting in a reduced P/ET compared with cows that ovulated larger follicles. Similarly, dairy cows that were synchronized with an E2/P4 protocol and had a CL at the time of PGF 2α had greater P/AI as ovulatory follicle diameter increased (linear effect), but this effect of follicle diameter was not observed in cows without a CL at PGF 2α (Pereira et al., 2015). These results provide further evidence that the growth of a large follicle in a low-P4 environment negatively affects fertility, but might have an opposite effect under a high-P4 environment.
Moreover, P/ET and pregnancy loss were affected by estrus expression and follicle diameter, corroborating previous studies from our research group (Pereira et al., 2015). In agreement with our data, cows ovulating a larger follicle at TAI might have greater pregnancy loss due to ovulation of an oocyte from a somewhat persistent follicle (Pereira et al., 2015. Persistent follicles induce premature maturation of the oocyte (Mihm et al., 1994;Revah and Butler, 1996) and alterations in the oviductal environment (Binelli et al., 1999), which might explain the negative effects on TAI but not TET.
It has been suggested that the diameter of the preovulatory follicle during the estrus cycle may be a key factor contributing to P4 concentrations post-AI, as larger follicles generate larger CL with greater endogenous P4 production (Vasconcelos et al., 2001;Mussard et al., 2007;Cooke et al., 2014). The increase in fertility with increasing P4 is expected due to the fact that greater P4 during the early luteal phase increases embryo length and IFN-tau production (Mann et al., 1999;Carter et al., 2008;Clemente et al., 2009;Cooke et al., 2014) and might indicate a CL being reflective of enhanced follicle and oocyte health or physiological function (Pereira et al., 2017a,b). Nonetheless, in agreement with our data, greater P4 production on d 7 is not always associated with higher fertility in TET (Mantovani et al., 2005;Demetrio et al., 2007;Pereira et al., 2021) protocols, suggesting that P4 effects might be mediated through actions on the embryo before but not after d 7.
Reduced circulating E2 concentrations during the periovulatory period have been implicated in inefficient sperm transport, suboptimal oviductal/uterine environment, and impaired oocyte fertilization (Hawk and Cooper, 1975;Ryan et al., 1993;Buhi, 2002). Nevertheless, expression of estrus is always associated with increased P/AI (Pereira et al., 2017a;Madureira et al., 2022) and P/ET (Pereira et al., 2017a), suggesting that improvements in fertilization rates cannot be the sole mechanism affecting reproductive function of dairy cows. In fact, estrus expression was associated with 49 and 66% improvements in P/ET and P/AI, respectively (Pereira et al., 2015), indicating that much of this improvement in fertility is independent of potential effects of estrus on oocyte development, fertilization, or early embryo development. Thus, enhancement of the uterine environment by factors associated with estrus expression affected embryo development after d 7. Estrus expression requires low circulating P4 concentrations in the presence of an elevated E2 level for a determined time period (Allrich, 1994). Therefore, it is logical to speculate that strategies to increase estrus expression in TAI or TET protocols could be achieved by adding E2 (Souza et al., 2007;Brusveen et al., 2009;Pereira et al., 2013b), being an advantage of ECP protocols, since estrus detection is a predictor of greater P/AI and reduced pregnancy loss (Pereira et al., 2013b(Pereira et al., , 2014(Pereira et al., , 2016. In addition to ECP, increasing protocol length resulted in a greater estrus detection when 8-versus 9-d protocols were compared, likely due to greater endogenous E2 production or reduced circulating P4 concentration near TAI (Pereira et al., 2014). The longer duration protocol would also allow an extended period for follicle growth, likely increasing circulating E2 concentration, allowing greater time for CL regression, and reducing circulating P4 concentration (Pereira et al., 2013a). In the current experiment, the second PGF 2α dose increased estrus expression by 9.4 percentage points (67.7 vs. 77.1%, respectively), highlighting that even in protocols that use E2 to synchronize ovulation, an improvement in estrus expression was detected. Likewise, Madureira et al. (2021) observed that lower concentrations of P4 on d 0 were tightly correlated with estrus expression and pregnancy rates, suggesting an association between complete CL regression and the behavioral trigger of estrus in the hypothalamus. In fact, concentration of P4 in the preceding estrus cycle has a significant effect on estrus expression and P/AI, demonstrating that the whole profile of P4, not only at individual time points, is likely important to determine fertility.
In summary, similar to what has been observed in TAI protocols, results from the present experiment support the concept that dairy cows with increased circulating P4 concentration during the estrous cycle preceding ET have greater P/ET, and highlight the fact that P4 is also associated with embryo maintenance after d 7. Administration of the second PGF 2α dose was important in reducing P4 concentration at the expected time of AI on d 0 and increased ovulatory response to the protocol. Additional research efforts that promote an increase in mean P4 concentration to a greater extent than GnRH, such as 2 P4-releasing devices or presynchronization, are warranted as potential alternatives to improve fertility of recipient lactating dairy cows.