Effect of 200 µg of gonadorelin at the first GnRH of the Resynch-25 on ovarian dynamics and fertility in lactating Holstein cows

Our objective was to determine the effect of a 200 µg dose of gonadorelin hydrochloride 25 d after previous artificial insemination ( AI ) in a Resynch-25 resynchronization program on ovulatory response, circulating progesterone ( P4 ) concentrations before and after treatment, and pregnancy per AI ( P/AI ) compared with a 100 µg dose in lactating Holstein cows. Experimental d 0 was considered the day of the previous AI. Lactating dairy cows (n = 3,240) with an average of 126 d in milk ( DIM ) and between 1 to 6 services were randomly assigned to receive 100 µg or 200 µg of go-nadorelin hydrochloride on d 25 ( GnRH25 ). On d 32 post-AI, cows diagnosed nonpregnant with the presence of a corpus luteum ( CL ) detected by ultrasound (n = 1,249) received PGF 2α treatments on d 32 and 33, followed by a GnRH 32 h later and AI 16 h after this last GnRH. Blood samples were collected on d 25, 32, and 34 to evaluate serum P4 concentrations. Transrectal ultrasonographic examination was performed on d 25 and 27 to assess ovulatory response to GnRH25. Cows were checked for pregnancy on d 32, 46, and 88 after AI. The larger dose of GnRH increased the overall proportion of cows that ovulated to the GnRH25 (100 µg: 25.0% vs. 200 µg: 32.5%). However, when cows were evaluated separately according to the pregnancy status on d 32 post-AI, there was no treatment effect within cows pregnant and nonpregnant. Even though treat-ment increased the proportion of cows with serum P4 ≤ 0.42 ng/mL at G2 (100µg: 86.2% vs. 200µg: 93.0%), it did not affect P/AI on d 32, 46, and 88. Furthermore, a greater proportion of cows without a functional CL at GnRH25 had circulating P4 concentrations ≥1.00 ng/ mL on d 32 and lower than 0.42 ng/mL on G2. These cows also had a greater P/AI on d 32, 46, and 88. In summary, the larger dose of GnRH on d 25 post-AI did not increase the ovulatory response in nonpregnant cows and P/AI on d 32, 46, and 88 after AI after the Resynch-25 program. Additionally, nonpregnant cows without a functional CL at GnRH25 were better synchronized after the Resynch-25 protocol and had greater P/AI on d 32, 46, and 88 after timed-AI.


INTRODUCTION
Pregnancy per artificial insemination (P/AI) and service rate are the 2 primary drivers of the reproduction performance of dairy herds (Overton and Cabrera, 2017).The service rate is affected by the interval between inseminations (IBI).A currently common strategy in US dairy farms to reduce IBI is to initiate the resynchronization Ovsynch protocol with the first GnRH 25 d after previous AI (GnRH25), before the pregnancy diagnosis (Resynch-25;Silva et al., 2009).This strategy is even more impactful in dairy farms that rely only on timed-AI programs without detection of estrus.However, poor ovulatory response to GnRH25 may compromise the reproductive success of this Resynch-25 program (Lopes et al., 2013).Some studies reported no differences in the reproductive performance of lactating dairy cows when GnRH25 was removed from the protocol, and estrous detection was performed alongside timed-AI (TAI) for second and subsequent services (Wijma et al., 2017(Wijma et al., , 2018;;Pérez et al., 2020).One of the reasons for this might be an inadequate ovulatory response to the GnRH25.Therefore, a practical strategy that leads to a greater ovulatory response to this treatment may improve the reproductive success of this resynchronization program.
Several studies have shown that ovulation to the first GnRH of Ovsynch (G1) is associated with an increase in the proportion of cows synchronized to the program and fertility outcomes to the first service (Bello et al., 2006;Chebel et al., 2006;Giordano et al., 2013) and subsequent services (Silva et al., 2007).Still, promoting the ovulation of a dominant follicle in a random stage of the estrous cycle can be challenging.High serum progesterone (P4) concentrations or the absence of a dominant follicle (Sartori et al., 2001) and, consequently, low estradiol (E2) concentrations can negatively impact the ovulatory response to exogenous GnRH (Motta et al., 2020).
One of the factors that has a significant influence on ovulatory response to exogenous GnRH is the stage of follicular development.Inducing ovulation with G1 in synchronization of ovulation programs on d 6 or 7 of the estrous cycle promotes the best synchronization response (Vasconcelos et al., 1999;Bello et al., 2006;Bello et al., 2007).Thus, pre-synchronization strategies for the first service were developed to ensure that cows start the Ovsynch program at these desired days of the estrous cycle (Bello et al., 2006;Galvão et al., 2007;Souza et al., 2008).These pre-synchronization programs combined with Ovsynch have promoted high P/AI for first (Herlihy et al., 2012) and subsequent services (Giordano et al., 2012b(Giordano et al., , 2012c)).Still, they are usually long and use PGF 2α , which makes their application as a resynchronization strategy challenging.For the first service, it is possible to start the pre-synchronization before the voluntary waiting period (VWP); therefore, the time to first AI postpartum is not delayed.Nonetheless, PGF 2α could only be used in a resynchronization program after cows are diagnosed as nonpregnant due to its abortifacient properties.Since these programs are long and most US dairy farms perform pregnancy diagnosis 32 d post-AI or later, these more elaborated programs could increase the average IBI and decrease the herd service rate (Mendonça et al., 2012).
Another strategy developed to improve synchronization rate and fertility outcomes of resynchronization programs was to use inducers of ovulation (GnRH or human chorionic gonadotropin -hCG) 7 d before the start of Ovsynch and pregnancy diagnosis.These strategies aimed to improve fertility outcomes by optimizing ovulatory response to G1 due to better control of the onset of the follicular wave.Studies also tested this strategy with pregnancy diagnosis performed on different days after AI.For instance, GnRH used 7 d before the initiation of Ovsynch at nonpregnancy diagnosis (NPD) on d 32 (Lopes et al., 2013) or 39 post-AI (Dewey et al., 2010;Lopes et al., 2013) increased P/ AI compared with Ovsynch without presynchronization with GnRH.Moreover, presynchronization with hCG 7 d before the initiation of Ovsynch, 7 d before NPD increased P/AI without increasing IBI.However, the same effect was not observed when hCG was replaced by GnRH (Giordano et al., 2012b).Even though these strategies have improved P/AI, they involve an extra hormonal treatment, increasing animal handling, labor, and cost.In addition, it may alter the expression of estrus in nonpregnant cows after AI, which may reduce the service rate in farms performing AI after detection of estrus.
Another factor that may impair the ovulatory response to GnRH is the analog or product used.Different GnRH products and analogs may induce different magnitudes of the luteinizing hormone (LH) surge (Picard-Hagen et al., 2015), ovulatory responses to the first GnRH of Ovsynch (Souza et al., 2009), and P/ AI (Luchterhand et al., 2019).Furthermore, increasing the dose of the same GnRH product induced a greater LH surge regardless of circulating P4 concentration (Giordano et al., 2012a), increasing ovulatory response to this treatment (Giordano et al., 2013).Our laboratory has also demonstrated that increasing the dose of gonadorelin hydrochloride from 100 to 200 µg increased ovulatory response to G1 and P/AI 32 post TAI in firstservice lactating Holstein cows treated with the Double Ovsynch program (Valdés-Arciniega et al., 2023).
Thus, in the present study, we aimed to determine the effect of 200 µg of GnRH (gonadorelin hydrochloride) on ovulatory response, serum P4 concentrations during the protocol, and P/AI compared with 100 µg GnRH at the GnRH25 for lactating dairy cows in a dairy farm that relies mainly on TAI for reinseminations.We hypothesized that the 200-µg GnRH dose would lead to a greater ovulatory response to GnRH25 in nonpregnant cows at d 32 post-TAI and consequently increase P/AI for Resynch-25 compared with the 100-µg GnRH dose.

MATERIALS AND METHODS
All procedures involving animals were previously approved by the Institutional Animal Care and Use Committee of the University of Wisconsin-Madison (IACUC, protocol #V006207-R01).

Animals, Housing, and Farm Reproductive Management
This experiment was conducted from March 2020 to April 2021 on a commercial dairy farm in southcentral Wisconsin (Evansville, WI).Lactating dairy cows were housed in a regular freestall barn and a freestall barn with a cross-ventilated system.Stalls in both barns were bedded with recycled sand.Animals were grouped in pens according to their parity (primiparous or multiparous), milk production, and days in milk (DIM).Cows were milked thrice daily and fed a TMR diet formulated to meet or exceed the nutrient requirements according to their milk production (NRC, 2001).The average daily milk production during the week before treatment was 39.9 ± 0.2 kg/d for primiparous and 54.7 ± 0.12 kg/d for multiparous cows.Animals had free access to clean water and feed throughout the day.
First service cows were submitted to a Double Ovsynch program (GnRH, 7 d later PGF 2α , 3 d later GnRH, 7 d later GnRH, 7 d later PGF 2α , 1 d late PGF 2α , approximately 32 h later GnRH, and approximately 16 h later TAI; Souza et al., 2008).For the first 18 weeks of the study, primiparous and multiparous cows had their first service postpartum at 82 ± 3 DIM and 72 ± 3 DIM, respectively.After that, all cows received their first AI at 77 ± 3 DIM.The farm relied almost entirely on TAI, with only 3.8% of the total AI of the farm being after detection of estrus during the experimental period.

Treatment Administration and Experimental Design
Trained farm personnel administered all injections using single-dose syringes and 3.5-cm needles in semimembranosus or semitendinosus muscles of cows.Weekly cohorts of lactating Holstein cows (n = 3,239; primiparous, n = 1,113; multiparous, n = 2,126) that had been previously bred 1 to 6 times (1 time, n = 1,976; 2 times, n = 791, 3 times, n = 302; ≥ 4 times, n = 170) were assigned to receive either 100 µg or 200 µg of GnRH (GnRH25; gonadorelin hydrochloride; Factrel; Zoetis, Parsippany-Troy Hills, NJ) 25 d after previous AI or 19 d after embryo transfer (ET), according to their ear tag number (odd numbers -100 µg, even numbers -200 µg; Figure 1).The assignment of cows to each treatment was made by the on-farm management software (DairyComp 305, Valley Ag Software).At the time of treatment administration, electronic identification ear tags (Allflex) were scanned by the farm manager in front of the cow using an electronic identification reader and a hand-held computer with Pocket CowCard software (DairyComp 305, Valley Ag Software) that specified the dose that should be administered to each cow by a farm staff in the back of the cow.The farm personnel and veterinarians did not know the criteria adopted by the software for treatment assignment.
Cows diagnosed as pregnant did not receive further treatment (n = 1,668).Cows diagnosed as nonpregnant (n = 1,571) were then checked for the presence of a corpus luteum (CL).Nonpregnant cows with a CL (n = 1,249) received 2 PGF 2α (25 mg dinoprost, Lutalyse HighCon; Zoetis, Parsippany-Troy Hills, NJ) one day apart, a GnRH (100 µg) ~32 h after the second PGF 2α and were TAI ~16 h later (Figure 1).When no CL was detected, cows were enrolled in an Ovsynch + CIDR protocol (n = 188; Stevenson et al., 2006).This protocol consisted of the administration of GnRH (100 µg) and the insertion of an intravaginal P4 device (Eazi-Breed CIDR Cattle Inserts; Zoetis, Parsippany-Troy Hills, NJ) on d 32.Seven days later, the CIDR was removed, and 2 doses of PGF 2α one day apart were administered.Thirty-two hours after the last PGF 2α , cows received a second GnRH (100 µg) treatment and TAI ~16 h later.Eighty cows that finished the Resynch-25 did not receive AI and were used as recipients for embryo transfer.These cows were excluded from the P/AI analysis only.Artificial inseminations following the Resynch-25 protocol used Holstein sexed semen (n = 8), Holstein conventional semen (n = 844), and Angus conventional semen (n = 317).In the case of the Ovsynch + CIDR protocol, the AI procedure used Holstein sexed semen (n = 2), Holstein conventional semen (n = 127), and Angus conventional semen (n = 55).Additionally, 4 cows did not receive AI after the Ovsynch + CIDR protocol because of unknown reasons.Artificial inseminations were performed by 4 AI technicians blinded to treatments using semen from multiple sires purchased by the farm.Farm veterinarians blinded to treatments performed pregnancy diagnosis on d 32, 46, 88, and 200 post-TAI by transrectal ultrasonography examination of the reproductive tract (Easi-Scan, BCF Technology Ltd.).Pregnancy was confirmed by embryo heartbeat.
Nine cows (1 primiparous and 8 multiparous) were considered nonpregnant at the NPD because they were detected in estrus between GnRH25 and pregnancy diagnosis.In addition, 4 multiparous cows were inseminated after estrous detection between NPD and completion of the Resynch-25 protocol.Of the nonpregnant cows at NPD, 51 were designated as "do not breed."In addition, 36 cows were not bred after completing the Resynch-25 protocol.Moreover, 35 cows were identified as having an embryo without heartbeat and did not complete the protocol.Data from these cows were collected and analyzed until the point of estrus detection or until NPD for those designated as "do not breed."

Ultrasonographic Examinations of Ovaries
Transrectal ultrasonographic examination of the ovaries was performed on a subset of cows on d 25 (n = 1,073) and 27 (n = 1,042) post-TAI and subsequently on d 34 post-TAI only for previously scanned cows diagnosed nonpregnant on d 32 (n = 499).A 7.5-MHz linear transducer (Easi-Scan:Go, BCF Technology Ltd.) connected to a mini-iPad generation 4 (Apple Inc., Cupertino, CA) was utilized to record the sonographic videos.Ovarian structures were measured using a software program for video metrics analysis (Kinovea 0.8.15;Kinovea.org).Ovulation to GnRH25 was determined when one or more large antral follicles (≥9 mm in diameter) were detected on d of GnRH25 and disappeared 2 d later, and a luteal structure appeared on the same site.Only nonpregnant cows with a CL on d 32 were evaluated on d 34.
A frozen image of the sonographic videos was used to determine the apparent maximal ovarian structure size, using height (H) and width (W) at a 90° angle.The display background gridlines of 10 mm length were utilized to calibrate the software calipers.The mean diameter of follicles, corpora lutea (CLs), and luteal cavities were calculated as the average between the H and the W. Corpus luteum and luteal cavity area were determined by the equation 0.5 H × 0.5 W × π (Kastelic et al., 1990;Martins et al., 2011).The total luteal area (TLA) of CLs with a cavity was calculated by subtracting the luteal cavity area from the CL's area.The TLA consists of the sum of the areas when the cows had more than 1 CL.

Blood Sampling and P4 Assays
Blood samples were collected on d 25, 32, and 34 post-TAI by coccygeal vein or artery puncture.Only the serum P4 concentration of cows diagnosed nonpregnant with a CL on d 32 that continued the Resynch-25 protocol were considered on d 34.Vacuum tubes of 9 mL (Vacuette Z serum clot activator, Greiner Bio-One International GmbH, Kremsmünster, Austria) and 20-gauge, 3.8 cm needles (Vacuette Multi-Sample Blood Collection Needle, Greiner Bio-One International GmbH, Kremsmünster, Austria) were used.After being collected, blood samples were stored in a cooler with ice and transported to the laboratory within ~6 h.Once in the laboratory, blood samples were centrifuged at 2,000 × g for 10 min at 4°C for serum separation.Serum was then transferred to 2 mL microcentrifuge tubes (Premium MCT Graduated Microcentrifuge Tubes, Fisher Scientific, USA) and stored at −20°C until RIA assays to measure progesterone (P4) concentration.
The circulating P4 concentration was measured using a solid-phase RIA kit (ImmuChem coated tube Progesterone; MP Biomedicals, Costa Mesa, CA).A total of 9 assays were performed.The average assay sensitivity, calculated as 2 SD less than the mean counts per minute of maximum binding, was 0.02 ng/mL.Intra-and inter-assay coefficients of variation were 7.0 and 8.7%, respectively.Cows were considered to have a functional CL if serum P4 was greater than 1.00 ng/mL on that day.Cows were considered to have had complete luteolysis when serum P4 was ≤0.42 ng/mL 48 h after the first PGF 2α injection (d 34).This cutoff for complete luteolysis was determined by a receiver operating characteristic (ROC) analysis described in the statistical analysis section.

Statistical Analysis
The determination of the sample size was based on an expected increase in ovulatory response of nonpregnant cows to GnRH25 from 40% to 65%, when using the larger dose of GnRH.This increase in ovulatory response was expected to translate to an 8-percentage point overall improvement in P/AI after timed-AI for cows treated with the larger dose of GnRH (42% vs. 50%).A priori power analysis performed in G*Power (version 3.1.9;Faul F et al., 2007) using the options of z test, logistic regression, one tail, odds ratio of 1.38 (H0: 42% vs. H1: 50%), α = 0.05 and β = 0.80, and binomial distribution indicated that a total sample size of 965 cows was required.
Continuous variables such as ovulatory follicle size, TLA, and circulating P4 concentrations were analyzed by ANOVA using the MIXED procedure of SAS (version 9.4; SAS Institute Inc., Cary, NC).The residual option of the MIXED procedure was used to assume normality and homoscedasticity of residuals after visual inspection of Studentized residual plots for each variable after fitting the model.Variables that did not fulfill assumptions were square-root-transformed or logtransformed for the analysis.All continuous variables are presented as actual means ± SEM obtained using the MEANS procedure of SAS.Furthermore, continuous variables that were transformed (t) to fulfill normality assumptions had their estimated least squared means (LSM) and standard error (SE) back transformed (Bt), and the results are presented in the tables.To back transform variables that were square-root-transformed, we used the formulas: LSM Bt = (LSM t ) 2 and SE Bt = 2 × LSM t × SE t .To back transform variables that were logtransformed we used the formulas: Binary variables such as ovulatory response to G1, proportion of cows with a functional CL on d 32 post-TAI, and P/AI were analyzed by logistic regression using the GLIMMIX procedure of SAS.Binary data are presented as proportions that were obtained using the MEANS procedure.To avoid confounding, due to the small numbers, cows bred with sexed semen (n = 20; 100 µg, n = 10; 200 µg, n = 10) and used as recipients (n = 78; 100 µg, n = 35; 200 µg, n = 43) after the synchronization with the Resynch-25 protocol were excluded from the analyses of P/AI on d 32, 46, 88, and pregnancy losses.Two cows enrolled in the Ovsynch + CIDR protocol bred with sexed semen were also excluded from the P/AI analysis.
The analysis of conception rate of previous breeding included the fixed effect of treatment (100 µg vs. 200 µg), parity (primiparous or multiparous), semen type (sexed or conventional), service number (first or subsequent services), and reproductive technology used for previous breeding (AI or ET).The 2-way interactions between treatment and the other fixed effects were initially included in the model and subsequently excluded from the model by backward elimination if P > 0.20.
The serum P4 cutoff on d 34 to determine complete luteolysis was stablished after performing a ROC analysis using RStudio version 1.3.1073(RStudio Team, 2020) to identify the optimal P4 value for predicting P/AI on d 32.The P4 concentration of 0.42 ng/mL exhibited the best specificity (85%) and sensitivity (94.8%), with an area under the curve of 0.53 (95% CI = 0.47, 0.58).Cows were then classified based on serum P4 levels equal to or below 0.42 ng/mL or greater than 0.42 ng/mL for further analyses described below.
To check differences between cows diagnosed pregnant and nonpregnant on d 32 post-TAI, the models considered the fixed effects of pregnancy status on d 32, parity, service (2nd, 3rd, and ≥4th) and the interaction pregnancy status × parity.The models to determine the effect of treatment on continuous and binary response variables considered the fixed effects of treatment, parity, service and the interaction treatment × parity.The effect of CL presence on GnRH25 d was determined considering the fixed effects of CL presence, treatment, parity, service and all possible 2-and 3-way interactions.The last model aimed to identify the effect of ovulatory response to GnRH25 and considered the fixed effects of ovulatory response to GnRH25, parity, service and the interaction ovulatory response to GnRH25 × parity.Considering our hypothesis that ovulatory response to GnRH25 enhances fertility, a one-tailed test was used in this model.Interactions with P-values greater than 0.20 were backward eliminated from the model unless used to evaluate specific treatment effects, such as the interaction between treatment and pregnancy diagnosis at d 32 on ovulatory response to treatment.
Besides the analysis of the effect of ovulatory response to the GnRH25, all other models used a 2-tailed test.Probability values were considered different when P ≤ 0.05 and trending toward significance when P > 0.05 and ≤0.10.

Effect of GnRH25 on the P/AI of the Inseminations Before Treatment
Among the 3,239 cows that received GnRH25, 1,668 were confirmed pregnant on d 32 after the previous breeding.Treatment did not affect (P = 0.75) P/AI for the inseminations before the GnRH25 treatment [100 µg GnRH: 51.3% ( 850 ET: 55.1% (234/425), P = 0.72) also did not affect P/ AI 32 d after previous breeding.However, primiparous had greater (P < 0.01) P/AI for the breeding before the GnRH25 compared with multiparous cows [primiparous: 56.0% (623/1,113) vs. multiparous: 49.2% (1,045/2,126)].No significant 2-way interactions between treatment and any of the other fixed effects were detected in the observed data.
Total luteal area (TLA) on d 25 was greater (P < 0.01) in pregnant cows compared with cows that were not pregnant on d 32 post-TAI, agreeing with serum P4 concentrations.Moreover, a tendency (P = 0.06) for interaction between parity and pregnancy status on TLA at treatment was observed (Table 1).There was no difference in d 25 TLA between primiparous and multiparous in pregnant cows, whereas nonpregnant multiparous cows had greater d 25 TLA than primiparous nonpregnant cows.Additionally, a greater proportion of pregnant cows had more than one CL at GnRH25 regardless of parity (Table 1).
Ovulatory response to GnRH25 was affected (P < 0.01) by pregnancy status (Table 1).A smaller proportion of pregnant cows ovulated in response to GnRH25 than nonpregnant (13.1% vs. 42.6%,respectively).Furthermore, the proportion of cows with at least one follicle ≥9 mm in diameter was greater for pregnant than nonpregnant cows and for multiparous than primiparous (Table 1).Pregnancy status and parity also affected pre-ovulatory follicle diameter (Table 1).Preovulatory follicle was smaller for pregnant cows compared with nonpregnant (14.72 ± 0.31 mm vs. 15.61 ± 0.19 mm, respectively) and primiparous compared with multiparous (14.35 ± 0.26 mm vs. 15.85 ± 0.20 mm, respectively).

Effects of Treatment on Ovarian Parameters, Ovulatory Response, and P/AI
The larger dose of gonadorelin hydrochloride (200 µg) increased the overall ovulatory response to GnRH25 (Table 2).However, when treatment was evaluated on cows within each pregnancy status, treatment did not affect ovulatory response to GnRH25 for pregnant or nonpregnant cows (Table 2).In addition, treatment did not affect the proportion of nonpregnant cows with functional CL, CL detected by ultrasound, and circulating P4 concentrations on d 32 post-TAI (Table 3).In contrast, a greater proportion (P = 0.04) of multiparous cows (14.3%) were enrolled in Ovsynch + CIDR compared with primiparous cows (10.2%).
Interestingly, a greater proportion of nonpregnant cows enrolled in the Resynch-25 receiving the 200 µg dose at GnRH25 had circulating P4 concentrations lower than 0.42 ng/mL on the day of the last GnRH of the Ovsynch (G2, d 34; Table 3).The overall P/ AI of cows enrolled in the Resynch-25 in the present study was 48.0%, 44.7%, and 43.4% at d 32, 46 and 88 post-AI, respectively.For the cows enrolled in the Resynch-25, treatment did not affect any of the fertility parameters evaluated in the study (Table 3).However,  Mean values in the same row with different superscripts differ (P ≤ 0.05) for the interaction of pregnancy status on d 32 post-AI and parity. A,B Mean values in the same row with different capital superscripts tended to differ (P-values >0.05 and ≤0.10) for the interaction between pregnancy status on d 32 and parity.
2 Fixed effect of pregnancy status (pregnant vs. nonpregnant) 32 d after previous AI.
3 Fixed effect of parity (primiparous vs. multiparous). 4 Effect of the interaction between pregnancy status on d 32 post-AI and parity.The interaction was removed from the final model if P > 0.20.Considering serum P4 cutoff of 1.00 ng/mL to classify a CL as functional, the utilization of a B-mode ultrasound examination at NPD resulted in the accurate classification of 86.9% (423/487) of nonpregnant cows with available serum P4 analysis.Still, 6.8% (33/487) of cows classified as not having a CL and enrolled in the Ovsynch + CIDR protocol had serum P4 ≥ 1.00 ng/mL.On the other hand, 6.4% (31/487) actually had serum P4 < 1.00 ng/mL but were identified as having a CL by the farm veterinarian.These cows (n = 31) were inseminated after the Resynch-25 protocol, but only 4/31 (12.9%) were pregnant 32 d after TAI, and 3/31 (9.7%) were pregnant by 46 d post-TAI.On the contrary, 19 out of 27 (70.4%)nonpregnant cows that were accurately identified as not having a functional CL and enrolled in the Ovsynch + CIDR protocol were pregnant 32 d after TAI.When considering all cows enrolled in the Ovsynch + CIDR protocol, neither treatment nor parity affected P/AI on d 32, 46 and 88 post-TAI (Table 3).Additionally, there was no effect of treatment nor parity on pregnancy losses of these cows (Table 3).

Association Between Functional CL Presence at GnRH25 and Ovarian Parameters, Ovulatory Response, and P/AI
Table 4 describes the association between functional CL presence at GnRH25 and its interaction with treatment with ovarian and fertility parameters in nonpregnant cows at d 32.The larger dose of GnRH25 effectively increased the ovulatory response in cows with a functional CL at GnRH25.In contrast, treatment did not affect the ovulatory response in cows without a functional CL at GnRH25 (Table 4).In addition, the proportion of cows with a functional CL on d 32 was greater (P < 0.01) for cows without a functional CL at GnRH25 than for cows with a functional CL (95.5% vs. 78.3%,respectively).
We observed an interaction between functional CL presence at GnRH25 and parity in the proportion of nonpregnant cows with a functional CL at d 32 (Table 4).A greater proportion (P = 0.05) of nonpregnant multiparous cows with a functional CL at GnRH25 also had a functional CL on d 32 compared with primiparous cows (81.8% vs. 69.2%,respectively).On the other hand, in nonpregnant cows without a functional CL at GnRH25, there was no parity effect on the proportion of cows with a functional CL on d 32 (98.8% vs. 94.1%,P = 0.13, for primiparous and multiparous cows, respectively).An interaction between the presence of a functional CL at GnRH25 and parity was also observed in the proportion of nonpregnant cows detected without a CL by the farm veterinarians on d 32 (Table 4).A greater proportion (P = 0.01) of nonpregnant primiparous cows with a functional CL  A greater (P < 0.01) proportion of nonpregnant cows with functional CL at GnRH25 had serum P4 > 0.42 ng/mL at G2 compared with cows without a functional CL at GnRH25 (17.5 vs. 5.2%, respectively).Finally, cows without functional CL at GnRH25 had greater P/ AI on d 32 (53.8% vs. 39.7%,P < 0.01), 46 (51.9% vs. 36.3%,P < 0.01), and 88 (50.4% vs. 36.3%,P < 0.01) post-AI than cows with functional CL at GnRH25, but pregnancy loss did not differ between them (Table 4).

Association between Ovulatory Response to GnRH25 and Ovarian Parameters and P/AI
Nonpregnant cows that did not ovulate to GnRH25 had greater (P < 0.01) serum P4 concentrations on d 25 than cows that ovulated (Table 5).Additionally, cows that did not ovulate to GnRH25 tended to have more than one CL on d 25 (Table 5).Even though cows that did not ovulate on d 25 had greater serum P4 concentration on d 32, a decreased proportion of these cows had a functional CL on this day than cows that ovulated to GnRH25 (Table 5).Consequently, this difference in serum P4 concentration was even greater when only considering serum P4 concentrations of cows with a functional CL on d 32 (Table 5).Lastly, ovulation to GnRH25 improved P/AI 32, 46, and 88 d post-TAI in cows that continued the Resynch-25 but did not alter pregnancy losses (Table 5).Mean values in the same row with different superscripts differ (P ≤ 0.05) for the interaction between CL presence and parity. A,B Mean values in the same row with different capital superscripts tended to differ (P-values >0.05 and ≤0.10) for the interaction between CL presence and parity.

DISCUSSION
The present study determined the effect of increasing the dose of the GnRH25 from 100 to 200 µg of gonadorelin hydrochloride on ovarian parameters and fertility of lactating dairy cows enrolled in a Resynch-25 program not combined with insemination after detection of estrus.The larger GnRH dose increased the overall ovulatory response in all cows, including pregnant and nonpregnant cows.Nevertheless, when we evaluated only nonpregnant cows, the larger GnRH dose did not improve ovulatory response to GnRH25 and P/AI for cows submitted to a Resynch-25, contrary to our hypothesis.This absence of treatment effect on ovulatory response in nonpregnant cows seems to be due to the wide variability in the stage of the estrous cycle and follicle development at the time of the GnRH25 treatment.Ricci et al. ( 2017) indicated that the day of CL regression after TAI varies widely among nonpregnant cows, and about 21% of nonpregnant cows do not undergo CL regression by d 32 post-AI.
A recent study from our laboratory also compared 100 vs. 200 of GnRH, using the same GnRH product, but at G1 of the breeding Ovsynch in a Double Ovsynch program for first-service TAI in lactating dairy cows.In that study, the 200 µg GnRH dose increased ovulation compared with a 100 µg GnRH dose (81.3% vs. 65.1%;Valdes-Arciniega et al., 2023).A presynchronization with an Ovsynch protocol, as in the Double Ovsynch, increases the chance of the presence of a GnRH-responsive dominant follicle at the time of G1 (Bello et al., 2006).Also, pre-synchronization with inducers of ovulation (GnRH or hCG) 7 d before the initiation of the Ovsynch for resynchronization of ovulation has been efficient in increasing P/AI (Giordano et al., 2012b;Lopes et al., 2013).The present study did not use a pre-synchronization strategy, decreasing the chance of cows having a dominant follicle at the time of GnRH25.The lack of treatment effect in cows without a dominant follicle may not have allowed us to find an overall treatment effect in ovulatory response to GnRH25 of cows diagnosed nonpregnant on d 32.
Treatment with 200 µg of GnRH 25 d post-AI only increased ovulatory response in nonpregnant cows when a functional CL was present, which may be associated with the presence of a responsive ovulatory follicle and the interference of progesterone.This result agrees with other studies that indicate that a 200 µg dose of GnRH can partially overcome the negative effect of circulating P4 concentrations on ovulatory response to exogenous GnRH induced-LH surge (Giordano et al., 2012a).Cows with a functional CL on d 25 could be in late metestrus or diestrus stages of the estrous cycle.In this scenario, the negative effect of high circulating P4 concentrations on the LH surge induced by GnRH may have led to ovulation failure (Motta et al., 2020).In these cows, the larger dose of the GnRH might result in an LH peak with greater magnitude (Giordano et al., 2012a), partially overcoming the inhibitory effect of elevated serum P4 concentrations on the anterior pituitary and resulting in a greater ovulatory response than the smaller GnRH dose.
The poor ovulatory response in pregnant cows (13.1%) is another evidence that a combination of high circulating P4 concentrations and inappropriate stage of antral follicle development impair ovulatory response to exogenous GnRH.As expected, serum P4 concentrations were extremely high in pregnant cows (primiparous: 10.11 ± 0.29 ng/mL; and multiparous: 8.59 ± 0.17 ng/ mL), inducing a negative effect in the magnitude of the LH surge (Colazo et al., 2008;Giordano et al., 2012a).In addition, if pregnant cows had a new follicular wave onset on d 20 to 22 post-AI, they would be on d 5 to 3 of the follicular wave, which have a poor ovulatory response to an induced-LH surge (Sartori et al., 2001).These 2 factors may explain the absence of a treatment effect on ovulatory response in pregnant cows.
Although the larger dose of GnRH improved ovulatory response in nonpregnant cows with functional CL at GnRH25, it did not improve P/AI after TAI with Resynch-25.The increase in ovulatory response in these cows may not have improved the proportion of cows synchronized after the Resynch-25.In fact, nonpregnant cows with functional CL at GnRH25 had decreased P/AI than cows without a functional CL at GnRH25 (39.9% vs. 54.0%,respectively), and no interaction between functional CL at GnRH25 and treatment was observed.A greater proportion of nonpregnant cows with functional CL at GnRH25 would be in late diestrus than those without CL at GnRH25.They also may undergo luteolysis between d 25 and 32 post-AI (Wijma et al., 2017;Pérez et al., 2020).As indicated in our results, the presence of a functional CL at GnRH25 decreased the proportion of cows with functional CL on d 32 post-AI.These cows without a functional CL (serum P4 < 1.00 ng/mL) on d 32 post-AI were most likely in proestrus, estrus, or early metestrus.Cows in early metestrus on d 32 post-AI have an immature CL that likely does not respond to PGF 2α treatments on d 32 and 33 post-TAI, having P4 concentrations >0.42 ng/mL at G2 and reduced P/AI after the Resynch-25 TAI.Cows in estrus on d 32 post-AI would most likely ovulate before TAI.Thus, cows with functional CL at GnRH25 had a smaller probability of synchronizing and reduced P/AI to Resynch-25 than cows without a functional CL at GnRH25.
On the contrary, cows without a functional CL at GnRH25 were likely to be in proestrus, estrus, or early  Fixed effect of ovulatory response to GnRH25.
3 Fixed effect of parity (multiparous or primiparous). 4 Last GnRH treatment of the Ovsynch protocol.
metestrus stages of the estrous cycle 25 d post-AI.Cows in early metestrus are in the initial stages of follicle and CL development.Hence, they would not ovulate independently of the magnitude of the LH surge induced by the exogenous GnRH (Xu et al., 1995).Cows in proestrus and estrus periods are close to ovulation and would most likely ovulate independently of the administration of exogenous GnRH (Sartori et al., This interpretation was based on the results that even with only 45.3% of ovulation to GnRH25 for cows without functional CL on d 25, 95.5% had a functional CL on d 32 post-TAI.In addition, most of these cows (94.8%) had a CL responsive to the PGF 2α administrations on d 32 and 33, decreasing serum P4 to less than 0.42 ng/ mL at G2. Thus, a large proportion of cows without a functional CL at GnRH25 were synchronized at the end of the Resynch-25 and had a greater P/AI than those without a functional CL at GnRH25.These results were surprising and contradictory since previous studies indicated that cows starting Ovsynch without a functional CL have decreased P/AI than cows with a CL (Denicol et al., 2012;Bisinotto et al., 2013Bisinotto et al., , 2015)).However, those studies were performed with first service (Denicol et al., 2012) or a combination of first and subsequent service cows (Bisinotto et al., 2013(Bisinotto et al., , 2015)).Close to 70% of first service cows presynchronized with PGF 2α that lack a CL at the initiation of the synchronization protocol are anovular (Bisinotto et al., 2010).Anovular cows submitted to Ovsynch have reduced P/AI (Denicol et al., 2012;Bisinotto et al., 2013Bisinotto et al., , 2015)).On the contrary, our study only included cows receiving second or greater services submitted to Double Ovsynch for the first service before enrollment.Therefore, these cows had a small chance of being anovular at the time of GnRH-25 (Gümen et al., 2003).
Furthermore, Bisinotto et al. ( 2013) reported that cows without a CL at G1 of Ovsynch but with a new CL at the time of PGF 2α had a similar P/AI compared with cows with a CL at G1.However, in their study, only 64.4% of cows without a CL at G1 had a new CL at PGF 2α of the breeding Ovsynch.On the other hand, in the present study, most cows (95.5%) that did not have a CL at GnRH-25 had a functional CL at the time of PGF 2α .Also, even though ovulation to the last GnRH of the Resynch-25 protocol was not evaluated in the present study, a greater proportion of cows that started the Resynch-25 without a CL had a functional CL at the time of PGF 2α of the Resynch-25 and underwent complete luteolysis than cows with a CL at the initiation of the protocol.
The overall P/AI at d 32 post-TAI in cows enrolled in Resynch-25 was 48.0% and seemed to be greater than other studies using a Resynch-25 combined [39.6% by (Wijma et al., 2017) and 43% by (Pérez et al., 2020)] or not combined [28% and 25.5% by (Giordano et al., 2012c)] with AI after observed estrus.Contrary to our hypothesis, treatment did not affect P/AI, which may be explained by the lack of treatment effect on ovulatory response to GnRH25 and the proportion of functional CL present on d 32 post-AI.
Ovulatory response to GnRH was associated with an improvement in P/AI (17.3% increase in P/AI 32 d post-TAI), which seems to be related to improved synchrony to the Resynch-25 protocol and better control of the size and age of the ovulatory follicle at the end of the protocol.Ovulation to GnRH25 ensures better synchrony to the Resynch-25, increasing the proportion of cows with a functional CL 7 d later on d 32 post-AI (day of the PGF 2α of the Ovsynch; Bello et al., 2006).Nevertheless, nonpregnant cows that do not ovulate to the GnRH25 have a greater chance of undergoing natural luteolysis between d 25 and 32 post-AI and do not have a CL at pregnancy diagnosis (Pérez et al., 2020).In addition, cows that ovulate to the first GnRH of the Ovsynch start a new follicular wave approximately 1 to 2 d after treatment, controlling the size of the ovulatory follicle and the ovulatory follicular wave duration (~10 d).Cows that do not ovulate to G1 may have an ovulatory follicle that is too young or old.Ovulatory follicular wave duration has been associated with fertility (Day, 2015), and more than 6 d of follicle dominance was related to poor embryo quality (Cerri et al., 2009).
The improvements in fertility associated with the ovulation to the first GnRH of an Ovsynch have been reported in several studies in timed-AI programs for first service (Bello et al., 2006;Chebel et al., 2006;Cerri et al., 2009;Giordano et al., 2013;Bisinotto et al., 2015).However, when Ovsynch is used for resynchronization, ovulatory response to the first GnRH is relatively reduced compared with ovulatory responses after G1 of Double Ovsynch or G6G protocols [71.8% (Souza et al., 2008) and 85.0% (Bello et al., 2006), for Double Ovsynch and G6G, respectively].Previous studies reported ovulatory response to the G1 of Ovsynch in resynchronization programs ranging from 40% to 62% (Lopes et al., 2013;Bruno et al., 2014;Wijma et al., 2017;Sauls-Hiesterman et al., 2020).The 42.6% ovulatory response to GnRH25 in the present study is within this range.Considering the 17.3% increase observed in P/AI 32 d post-AI of cows that ovulate to GnRH25 compared with the ones that did not ovulate, a considerable rise in ovulatory response to GnRH25 would be necessary to produce a significant increase in P/AI of the Resynch-25 protocol.
Some authors have suggested the removal of the GnRH25 of the Resynch-25 protocol for a short resynchronization program in cows detected with a CL at NPD 32 d after AI (Wijma et al., 2017(Wijma et al., , 2018;;Pérez Leão et al.: LARGER DOSE OF THE FIRST GnRH IN A RESYNCH-25 et al., 2020;Sauls-hiesterman et al., 2020).Unlike our study design, these studies combined TAI with estrus detection for reinsemination of nonpregnant cows.When evaluating only cows submitted to TAI, removing the GnRH25 either decreased P/AI (Wijma et al., 2017;Pérez et al., 2020;Sauls-hiesterman et al., 2020) or tended to increase pregnancy loss (Wijma et al., 2018).Data observed in our study suggests that the presence of a newly formed CL close to d 25 before or after) may be a better predictor of pregnancy for cows submitted to TAI using Resynch-25 than only using the presence of a CL (≥15 mm) and a large follicle (≥10 mm) at NPD 32 d after AI.Strategies that identify cows with a newly formed CL close to d 25 may help select the most appropriate resynchronization program to optimize P/AI in second and greater services.The use of technologies such as a cow-side blood P4 test or a color Doppler ultrasonography of the ovaries on d 32 may help identify CL functionality, increasing the accuracy of identification of cows without functional CL.These strategies could reduce the administration of unnecessary treatments and potentially increase fertility in systems that rely on resynchronization with TAI for reinsemination.
A limitation of our study was that not all nonpregnant cows received TAI after treatment at GnRH25.Cows without the visual presence of a CL, determined by the farm veterinarian using a B-mode ultrasound at NPD (32 d post-TAI), were enrolled in an Ovsynch + CIDR program at NPD and received TAI on d 42 after the previous TAI.The farm had used this strategy before this experiment, and it was based on previous studies that determined that cows without a CL at the PGF 2α of the Ovsynch or Resynch have a reduced probability of pregnancy (Sterry et al., 2006;Giordano et al., 2016;Wijma et al., 2017Wijma et al., , 2018;;Pérez et al., 2020).In fact, despite the small numbers of cows enrolled in this protocol, we observed an increased P/AI for them and a reduced P/AI for cows without a functional CL at NPD that continued the Resynch-25.Based on these results, it seems worth implementing this strategy for cows without a CL at NPD to optimize P/AI.
Another limitation of our study was the determination of CL presence at NPD by farm veterinarians using ultrasound.A discrepancy between the proportions of cows with functional CL determined by serum P4 concentration and visual CL determined by ultrasound has been previously reported (Stevenson et al., 2008;Giordano et al., 2016;Martins et al., 2023).Structural CL regression takes longer than functional luteolysis (Acosta et al., 2002;Cunha et al., 2022;Umaña Sedó et al., 2022).A recent study indicated that circulating P4 concentrations reached basal levels 2 d after the d of luteolysis onset (Cunha et al., 2022).However, the same cows took 3 d to reduce CL volume <45% after non-induced luteolysis onset in lactating dairy cows (Cunha et al., 2022).Thus, cows may still have a visible CL by ultrasonography from 1 to 3 d after luteolysis onset, but the CL may not produce enough P4 to reach circulating P4 concentrations >1.00 ng/mL.On the other hand, some cows that ovulated within 2 d after GnRH25 may have a small visible CL at d 32 that may produce enough P4 to achieve serum P4 concentrations slightly above 1.00 ng/mL.

CONCLUSION
The 200 µg dose of gonadorelin hydrochloride did not increase the ovulatory response to the first GnRH of the Resynch-25 program in nonpregnant cows compared with the 100 µg GnRH dose.This lack of effect led to no differences in the proportion of synchronized cows, P/AI on d 32, 46, or 88 post-TAI, and pregnancy losses in our study.Furthermore, cows with no functional CL at the initiation of the Resynch-25 were better synchronized and had greater P/AI on d 32, 46, and 88 post-AI than cows with functional CL.Finally, the ovulatory response to GnRH-25 increased P/AI at 32, 46, and 88 d after AI.Data provided in the present study increased the knowledge related to ovarian parameters and responses in cows submitted to a Resynch-25 program in farms without detection of estrus.This may allow the development of improved strategies to resynchronize and reinseminate nonpregnant cows after ovarian evaluation using ultrasonography.
Leão et al.: LARGER DOSE OF THE FIRST GnRH IN A RESYNCH-25 Leão et al.: LARGER DOSE OF THE FIRST GnRH IN A RESYNCH-25

Figure 1 .
Figure 1.Schematic diagram of the experimental design.Cows were randomly assigned to receive 100µg or 200µg of GnRH 25 d after the previous artificial insemination (AI).A pregnancy diagnosis was performed on d 32 post-AI.Pregnant cows did not receive any further treatment.Nonpregnant cows were checked for the presence of a corpus luteum (CL) by ultrasound during pregnancy diagnosis.Cows without CL were enrolled in an Ovsynch + CIDR protocol (GnRH + CIDR d 32, PGF 2α on d 39 and 40 post-AI, GnRH 32 h after the last PGF 2α , and TAI 16 h later).Nonpregnant cows with a CL finish the Resynch-25 protocol by receiving PGF 2α on d 32 and 33 and GnRH 32 h after the last PGF 2α treatment, followed by TAI 16 h later.A subset of cows had blood collected (B) on d 25, 32, and 34 post-AI and ultrasonographic (US) examination on d 25, 27, and 34 post-AI.
Leão et al.: LARGER DOSE OF THE FIRST GnRH IN A RESYNCH-25 Table 1.Effect of parity (primiparous vs. multiparous) and pregnancy status (pregnant vs. nonpregnant) on d 32 after previous artificial insemination (AI) on ovulatory response to GnRH25 1 , circulating progesterone (P4) concentrations, and ovarian structures on d 25 and 27 after previous AI in lactating Holstein cows Pregnancy Status on d 32 Mean total luteal area on d 25 post-AI, mm 2 Leão et al.: LARGER DOSE OF THE FIRST GnRH IN A RESYNCH-25 Leão et al.: LARGER DOSE OF THE FIRST GnRH IN A RESYNCH-25

1
GnRH-25 = gonadorelin hydrochloride given 25 d after previous AI. 2 Fixed effect of CL presence 25 d after previous AI. 3 Fixed effect of parity (multiparous or primiparous).4 Fixed effect of treatment (100 or 200 µg of gonadorelin hydrochloride).5 Effect of the interaction between CL presence on d 25 and parity.6 Effect of the interaction between CL presence on d 25 and treatment.7 Last GnRH treatment of the Ovsynch protocol.
Leão et al.: LARGER DOSE OF THE FIRST GnRH IN A RESYNCH-25 Table 5.Effect of ovulatory response to GnRH-25 1 on total luteal area, circulating progesterone (P4) concentrations, pregnancies per artificial insemination (P/AI), and pregnancy losses in lactating Holstein cows diagnosed nonpregnant 32 d after previous AI Ovulatory response serum P4 concentration on d 25 of cows with P4 ≥ 1.00 ng/ml on d 25, ng/mL ± serum P4 concentration on d 25 of cows with P4 ≥ 1.00 ng/ml on d 25, ng/mL ± SEM 4.45 ± 0.31 6.10 ± 0.30 Mean total luteal area on d25, mm 2

Table 3 .
Effect of treatment (TRT) 1 and parity on total luteal area, circulating progesterone (P4) concentrations, pregnancies per artificial insemination (P/AI), and pregnancy losses in lactating Holstein cows diagnosed nonpregnant 32 d after previous AI

Table 4 .
Effect of presence or absence of a functional corpus luteum (CL) 25 d after previous artificial insemination (AI), treatment with a larger (200 µg) or regular (100 µg) dose, and parity on ovulatory response to GnRH-25 1 , circulating progesterone (P4) concentrations, pregnancies per artificial insemination (P/AI), and pregnancy losses in lactating a,b,c