Available online at Theriogenology xx (2012) xxx

Transcription

1 Available online at Theriogenology xx (2012) xxx Fertility in dairy cows following presynchronization and administering twice the luteolytic dose of prostaglandin F 2 as one or two injections in the 5-day timed artificial insemination protocol E.S. Ribeiro a, R.S. Bisinotto a, M.G. Favoreto a, L.T. Martins a, R.L.A. Cerri a,b, F.T. Silvestre a,c, L.F. Greco a, W.W. Thatcher a, J.E.P. Santos a, * a Department of Animal Sciences, University of Florida, Gainesville, Florida 32611, USA b Department of Land and Food Systems, University of British Columbia, Vancouver, British Columbia V6T 1Z4, Canada c Pfizer Animal Health, Bakersfield, California 93101, USA Received 15 August 2011; received in revised form 24 December 2011; accepted 15 January 2012 Abstract The objectives were to evaluate pregnancy per AI (P/AI) of dairy cows subjected to the 5-day timed AI protocol under various synchronization and luteolytic treatments. Cows were either presynchronized or received supplemental progesterone during the synchronization protocol, and received a double luteolytic dose of PGF 2, either as one or two injections. In Experiment 1, dairy cows (n 737; Holstein 250, Jersey 80, and crossbred 407) in two seasonal grazing dairy farms were randomly assigned to one of four treatments in a 2 2 factorial arrangement. The day of AI was considered study Day 0. Half of the cows were presynchronized (G6G: PGF 2 on Day 16 and GnRH on Day 14) and received the 5-day timed AI protocol using 1 mg of cloprostenol, either as a single injection (G6G-S: GnRH on Day 8, PGF 2 on Day 3, and GnRH AI on Day 0) or divided into two injections of 0.5 mg each (G6G-T: GnRH on Day 8, PGF 2 on Day 3 and 2, and GnRH AI on Day 0). The remaining cows were not presynchronized and received a controlled internal drug-release (CIDR) insert containing progesterone from GnRH to the first PGF 2 injection of the 5-day timed AI protocol, and 1 mg of cloprostenol either as a single injection on Day -3 (CIDR-S) or divided into two injections of 0.5 mg each on Days -3 and -2 (CIDR-T). Ovaries were examined by ultrasonography on Days 8 and 3 and plasma progesterone concentrations were determined on Days 3 and 0. In Experiment 2, 655 high-producing Holstein cows had their estrous cycle presynchronized with PGF 2 at 46 3 and 60 3 days postpartum and were randomly assigned to receive 50 mg of dinoprost during the 5-day timed AI protocol, either as a single injection or divided into two injections of 25 mg each. Pregnancies per AI were determined on Days 35 and 64 after AI in both experiments. In Experiment 1, presynchronization with G6G increased the proportion of cows with a CL on Day 8 (80.6 vs. 58.8%), ovulation to the first GnRH of the protocol (64.2 vs. 50.2%), and the presence (95.6 vs. 88.4%) and number (1.79 vs. 1.30) of CL at PGF 2 compared with CIDR cows. Luteolysis was greater for two injections compared to a single PGF 2 injection (two PGF vs. single PGF %), especially in presynchronized cows (G6G-T 96.2 vs. G6G-S 61.7%). For cows not presynchronized, two PGF 2 injections had no effect on P/AI (CIDR-S 30.2 vs. CIDR-T 34.3%), whereas for presynchronized cows, it improved P/AI (G6G-S 28.7 vs. G6G-T 45.4%). In Experiment 2, the two-pgf 2 injection increased P/AI on Days 35 (two PGF vs. single PGF %) and 64 (two PGF % vs. single PGF %) after AI. Presynchronization and dividing the dose of PGF 2 (either cloprostenol or dinoprost) into two injections increased P/AI in lactating dairy cows subjected to the 5-day timed AI protocol Elsevier Inc. All rights reserved. Keywords: Dairy cow; Luteolysis; Presynchronization; Progesterone; 5-Day timed AI protocol * Corresponding author. Tel.: ; fax: address: jepsantos@ufl.edu (J.E.P. Santos) X/$ see front matter 2012 Elsevier Inc. All rights reserved. doi: /j.theriogenology

2 2 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx 1. Introduction Presynchronizing the estrous cycle of dairy cows with either two injections of PGF 2 [1,2] or a combination of GnRH and PGF 2 [3] increases the proportion of cows initiating the timed AI protocol in early diestrus. Cows at this stage of the estrous cycle have an increased probability to ovulate in response to exogenous GnRH [3,4] and have reduced occurrence of spontaneous luteolysis before the end of the timed AI protocol [1,2]. In addition to better synchrony of the estrous cycle [4], such responses assure that the ovulatory follicle develops under high systemic concentrations of progesterone and limit the duration of follicle dominance, all of which are important factors affecting fertility [5 7]. Furthermore, incorporating GnRH into presynchronization protocols induces estrous cyclicity in anovular cows [8]. Nevertheless, presynchronization increases labor, animal handling, and extends the length of the program until AI, which may constitute a nuisance for some dairy producers. An alternative to improve synchronization without lengthening timed AI programs is progesterone supplementation during the protocol. The use of intravaginal devices for controlled release of progesterone from the GnRH to the PGF 2 injections maintains blood progesterone concentrations that prevent premature estrous behavior, LH surge, and ovulation. These devices have been used during timed AI protocols to improve fertility of dairy cows when the estrous cycle is not presynchronized [2,9], or when cows are selectively inseminated following detection of estrus between presynchronization and initiation of a timed AI protocol [10,11]. Additionally, supplementing progesterone might benefit cows with low endogenous concentrations during development of the ovulatory follicle. The use of supplemental progesterone in lieu of presynchronization might be particularly attractive for protocols with a reduced interval of follicular dominance, as follicle turnover and aging of the oocyte might be less critical in these programs. Reducing the interval from the initial GnRH to induction of luteolysis from 7 to 5 days increased pregnancy per AI (P/AI) [7], although two injections of PGF 2 given on Days 5 and 6 after the first GnRH were needed for adequate CL regression. The second PGF 2 injection may be particularly important in presynchronized cows because of increased ovulation to the first GnRH injection and, consequently, increased occurrence of newly formed CL at PGF 2, which may not consistently regress after a single PGF 2 injection on Day 5 [7]. Nevertheless, a larger dose of PGF 2 administered as a single injection on Day 5 might enhance the luteolytic response, which would facilitate management of cows subjected to the 5-day timed AI program. The objectives of the present study were to compare fertility responses of lactating dairy cows subjected to the 5-day timed AI protocol, either presynchronized or supplemented with progesterone, and receiving twice the luteolytic dose of PGF 2, either as a single injection or divided into two injections. The hypotheses tested were that non-presynchronized cows subjected to the 5-day timed AI protocol with supplemental progesterone would have similar P/AI compared with presynchronized cows subjected to the same protocol without progesterone supplementation. Additionally, luteolysis and P/AI in cows subjected to the 5-day timed AI protocol would be similar if twice the luteolytic dose of PGF 2 is used, either as a single treatment on Day 5 of the program, or divided into two treatments on Days 5 and Materials and methods 2.1. Experiment Cows, pasture and management Experiment 1 was conducted in two commercial grazing dairy farms located in Florida, USA. Both were fall/winter calving dairy herds with similar genetics and management practices. The average milk production per cow was approximately 6000 kg/lactation. A total of 737 lactating dairy cows (250 Holsteins, 80 Jerseys, and 407 crossbreds) were enrolled in this study. The crossbred cows were mostly composed by F1 (50/50) and F2 (25/75) generations of crossbreeding between Holstein and Jersey genetics. Different genetic groups were managed together in a pasture-based production system in both farms. Cows were maintained under irrigated pastures of 2.7 ha and managed in a daily rotational program with a stocking rate of approximately 10 cows/ha. The pasture was composed by annual ryegrass (Lolium multiflorum Lam.) during winter and early spring, and Tifton 85 Bermuda grass (Cynodon spp.) during late spring and summer. Cows were offered variable amounts of concentrate (7 13 kg/ cow/d) during and immediately after each of the two daily milkings, according to forage availability and stage of lactation. The concentrate was based on soybean hulls, citrus pulp, whole lined cottonseed, cottonseed hulls, corn gluten feed, corn meal, soybean meal, molasses, and a mineral-vitamin premix. This concentrate was designed to contain, on a dry matter basis, approximately 15% crude protein, 4.5% fat, and 28% neutral detergent fiber.

3 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx 3 G6G Single PGF injection Cloprostenol (1 mg) GnRH + AI Presynchronization PGF GnRH GnRH No progesterone supplementation 3 d N = d 6 d 5 d US US 1 d 2 d (0.5 mg) (0.5 mg) Cloprostenol Two PGF injections N = 187 GnRH + AI CIDR Single PGF injection Cloprostenol (1 mg) GnRH + AI N = 190 No presynchronization GnRH CIDR 3 d Study Day US d US 1 d 2 d (0.5 mg) (0.5 mg) Cloprostenol Two PGF injections N = 181 GnRH + AI Fig. 1. Diagram of Experiment 1. AI, artificial insemination; BS, blood sampling for progesterone; CIDR, controlled internal drug-release insert containing 1.38 g of progesterone; GnRH, 100 g of gonadotropin-releasing hormone; G6G, presynchronization with 0.5 mg of PGF 2 as cloprostenol, followed 2 days later by 100 g GnRH; PGF 2, 1 mg of cloprostenol as a single injection or divided into two injections of 0.5 mg each; US, ultrasonography of the ovaries Experimental design The experimental design was completely randomized with blocks. Within farm, cows were blocked in groups of four according to breed (Holsteins, Jerseys and crossbreds), parity (primiparous and multiparous), and days in milk (DIM). Within each block, cows were randomly assigned to one of four treatments arranged asa2 2 factorial. There were two synchronization and two luteolytic treatments (Fig. 1) Synchronization and luteolytic treatments All cows in the study were subjected to the 5-day timed AI protocol [7], but with variants in synchronization and luteolytic treatments according to the experimental design. The day of AI was considered study day 0. Half of the cows were presynchronized using a GnRH/PGF 2 -based protocol G6G [3] with the timed AI protocol starting 6 days later, whereas the other half of the cows were not presynchronized and received a controlled internal drug-release (CIDR) insert (Eazi- Breed CIDR Cattle Insert; Pfizer Animal Health, Madison, NJ, USA) containing 1.38 g of progesterone between study Days 8 and 3. On study Day 3, cows in each synchronization treatment received one of two luteolytic treatments with 1 mg of a PGF 2 analog (cloprostenol sodium; Estrumate, Merck Animal Health, Summit, NJ, USA) either as a single injection on Day 3, or divided into two injections of 500 g each on Days 3 and 2. All cows were timed inseminated on study Day

4 4 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx 0 concurrent with the final injection of GnRH (Cystorelin, 2 ml containing 50 g/ml of gonadorelin diacetate tetrahydrate, Merial, Ltd., Inselin, NJ, USA) Ovarian ultrasonography and ovulatory responses Ultrasonographic examinations of the ovaries were performed with a 7.5 MHz linear transducer (Easy Scan, BCF Technology, Rochester, MN, USA) on study Days 8 and 3. Presence, location, and number of CL and follicles 8 mm were recorded. Ovulation to the first GnRH of timed AI protocol injection was characterized by appearance of a new CL on Day 3in an ovary containing at least one follicle 8mmonDay 8. A new CL at PGF 2 was characterized by the presence of a new CL, independent of the presence of follicles on Day Progesterone analysis and luteolysis On study Days 3 and 0, concurrent with the injection of PGF 2 and AI, respectively, a subgroup of 212 cows in one of the farms had blood sampled by puncture of the coccygeal vein or artery into evacuated tubes containing K 2 EDTA (Becton Dickinson, Franklin Lakes, NJ, USA). For cows receiving a CIDR insert, blood was sampled on Day 3 at least 15 min after insert removal; 0.6 ng/ml were deducted from the concentration, in accordance with the expected contribution of progesterone from the intravaginal insert [12]. Samples were immediately placed on ice and transported to the laboratory within 5 h after collection. Tubes were centrifuged at 2000g for 15 min. Plasma was separated, frozen at 20 C, and later analyzed for concentration of progesterone by RIA using a commercial kit (Coat-A-Count; Siemens Healthcare Diagnostics, Los Angeles, CA, USA). Samples were analyzed in a single assay with a sensitivity of 0.05 ng/ml, calculated at two SD below the mean counts per min at maximum binding. The intra-assay CV was calculated for each duplicate and it averaged 4.7%. Plasma samples of a cow on estrous cycle Days 4 (0.9 ng/ml) and 8 (4.3 ng/ml) were repeated several times in the assay and their CVs were 15.6 and 8.8%, respectively. Seventeen cows with progesterone concentration 1 ng/ml on Day 3 were not used for analysis of luteolysis, since it was assumed they did not have a functional CL to be regressed. Cows with progesterone concentration 1 ng/ml on Day 3 and 1 ng/ml on Day 0 were considered as having CL regression. Receiver operator characteristic (ROC) curves were used to determine concentrations of progesterone at PGF 2 and AI that resulted in highest accuracy to predict pregnancy on Day 35 after AI. Cows having both of these established criteria of progesterone concentrations were considered as having an ideal progesterone profile Detection of estrus, body condition score and DIM On study Day 2, tailheads of all cows were painted using paintsticks (All-Weather Paintstick; LA-CO Industries, Chicago, Illinois, USA) for detection of estrus on the day of AI based on removal of tailpaint. Cows were scored for body condition (BCS) in a 1 to 5 scale (1 emaciated, 5 obese [13]) at AI and 35 days later. For statistical analyses, cows were categorized as BCS 2.50, BCS 2.75 to 3.00, or BCS Similarly, cows were categorized according to body condition change from AI to pregnancy diagnosis as having lost, no change, or gained BCS. Cows were also categorized according to their DIM at AI as low ( 50 DIM; median of 40, range of 28 49), medium (50 90 DIM; median of 71, range of 50 90), or high ( 90 DIM; median of 109, range of ) Pregnancy diagnoses and pregnancy loss At 35 days after timed AI, pregnancy was diagnosed by ultrasonographic examination of the uterus and its contents, and was characterized by visualization of a live embryo by the presence of heartbeat. Cows diagnosed pregnant were re-examined by transrectal palpation of the uterus 64 days after timed AI. Pregnancies per AI were calculated as the number of pregnant cows at 35 and 64 days after AI, divided by the total number of cows inseminated. Pregnancy loss was calculated as the difference in the number of cows pregnant at 35 and 64 d, divided by the number of pregnant cows at 35 d Experiment Cows, housing and diets Experiment 2 was conducted in a 5000-cow commercial dairy farm located in Florida with rolling herd average of kg/yr. Lactating Holstein cows, 254 primiparous and 401 multiparous, were assigned to the study. Primiparous and multiparous cows were housed separately in free-stall barns equipped with sprinklers and fans. Both categories received the same total mixed ration to meet or exceed the nutrient requirements of a lactating Holstein cow producing 45 kg of milk/d with 3.5% fat and 3.2% true protein [14]. Cows were fed twice daily and milked thrice daily. Diet consisted of corn silage, rye grass silage, finely ground corn, dried citrus pulp, whole lined cottonseed, dried corn distiller s grains, wet brewer s grains, soybean meal, molas-

5 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx 5 Single PGF injection Presynchronization Dinoprost (50 mg) GnRH + AI N = 327 PGF PGF GnRH 3 d Study Day 14 d 12 d 5 d 1 d 2 d (25 mg) (25 mg) Dinoprost Two PGF injections GnRH + AI Days in milk (± 3) N = 328 Fig. 2. Diagram of Experiment 2. AI, artificial insemination; GnRH, 100 g of gonadotropin-releasing hormone; PGF 2, 25 mg of dinoprost for presynchronization or 50 mg of dinoprost, either as a single injection or divided into two injections of 25 mg each during the 5-d timed AI protocol. ses, minerals and vitamins. The diet was designed to contain, on a dry matter basis, approximately 15.9% crude protein, 5.1% fat, 24% starch, and 33% neutral detergent fiber Experimental design The experimental design was completely randomized with blocks. Weekly, a cohort of cows from 43 to 49 DIM were blocked by parity (primiparous or multiparous) and randomly assigned to one of two luteolytic treatments (Fig. 2). As in Experiment 1, study Day 0 was considered the day of AI Synchronization protocol and luteolytic treatments Estrous cycles were presynchronized with two injections of 25 mg of PGF 2 (dinoprost tromethamine; lutalyse sterile solution, Pfizer Animal Health) administered at 46 3 and 60 3 DIM [1]. The 5-day timed AI protocol was initiated at 72 3 DIM and cows were assigned to receive 50 mg of PGF 2 administered either as a single injection (n 327) or divided into two injections (n 328) of 25 mg each Pregnancy diagnoses, milk yield and BCS Pregnancy was determined 35 and 64 days after AI as described in Experiment 1. Yields of milk were recorded for individual cows once monthly, and the average for the first 3 mo postpartum was used for statistical analysis. Cows were categorized according to milk production above or below the mean value within parity. The BCS at AI was recorded and categorized as in Experiment Statistical analyses Experiment 1 All statistical models included the effects of synchronization treatment, luteolytic treatment, and interaction between them. The effects of farm, parity, breed, DIM, BCS at AI, change in BCS, sire, and AI technician were included when P Binary data were analyzed by multivariate logistic regression models using the LOGISTIC procedure of SAS Version 9.2 (SAS/STAT, SAS Institute, Inc., Cary, North Carolina, USA). A stepwise backward elimination was used and covariates were removed based on the Wald statistics criterion when P Synchronization and luteolytic treatments were forced into the final models. The probability of detecting a CL on study Day 8 was modeled using the logistic function with the intercept and the coefficient estimates from the logistic regression analyses according to synchronization treatment and DIM or BCS of cows using the formula: P 1 1 e (a b 1X 1 a b 2 X 2...).

6 6 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx Concentrations of progesterone on study Days 3 and 0 were analyzed by ANOVA using the GLM procedure of SAS. The models included the effects of synchronization treatment, luteolytic treatment, the interaction between the two treatment factors, BCS at AI, breed, parity, presence of CL on Day 8, ovulation to the first GnRH, and number of CL on Day 3. The ROC analysis option of Med-Calc Version (MedCalc Software, Mariakerke, Belgium) was used to determine the optimal concentration of progesterone on study Days 3 and 0 that predicted P/AI on Day 35 with greatest combined sensitivity and specificity Experiment 2 All statistical models included the effect of treatment, which was forced into the final models. The effects of parity, BCS at AI, change in BCS, milk yield, AI technician, and sire were included when P Data with binomial distribution were analyzed by multivariate logistic regression models using the LOGIS- TIC procedure of SAS. A stepwise backward elimination was used and covariates were removed based on the Wald statistics criterion when P In both experiments, differences with P 0.05 were considered significant and 0.05 P 0.10 were considered as a tendency. 3. Results 3.1. Experiment Effects of treatments The BCS of cows at AI and Day 35 did not differ (P 0.35) among treatments and they averaged and , respectively. Similarly, DIM at AI did not differ among treatments and they averaged for G6G and for CIDR and for single PGF 2 and for two PGF 2. According to the ROC, the concentrations of progesterone that resulted in the highest combined sensitivity and specificity to predict pregnancy on Day 35 after AI were 3.78 ng/ml at the injection of PGF 2 (study Day 3) and 0.32 ng/ml at AI (study Day 0). Therefore, cows with progesterone concentrations 3.78 ng/ml and 0.32 ng/ml on study Days 3 and 0, respectively, were considered as having the ideal progesterone profile. Presynchronization with G6G increased (P 0.01) the proportion of cows with CL on the day of the first GnRH of the timed AI protocol, ovulation to the first GnRH (study Day 8), and presence and number of CL and progesterone concentration on the day of the PGF 2 injection of the protocol (study Day 3) compared with CIDR cows (Table 1). The proportion of cows with CL on study Day 8 increased (P 0.01) with DIM and BCS of cows (Fig. 3). Interestingly, the benefit of G6G on the proportion of cows with CL was influenced by DIM and BCS. The G6G increased (P 0.05) the proportion of cows with CL when they had fewer than 90 DIM (Fig. 3A) and when the BCS was 3.50 (Fig. 3B). Both synchronization and luteolytic treatments influenced luteolysis. In general, G6G resulted in a smaller (P 0.05) proportion of cows undergoing luteolysis than did CIDR. Similarly, cows receiving the single injection of PGF 2 had a lower (P 0.01) incidence of luteolysis than those receiving the two injections of PGF 2, particularly when presynchronized with G6G (Table 1). As a consequence, cows in the single PGF 2 group had greater (P 0.01) concentrations of progesterone at AI and a smaller (P 0.01) proportion of cows with ideal progesterone profile than cows in two injections of PGF 2. Additionally, synchronization and luteolytic treatments affected detection of estrus at AI, following the same pattern as that for CL regression. Fewer (P 0.01) cows receiving G6G were detected in estrus compared with cows receiving CIDR. Similarly, treatment with twice the luteolytic dose of PGF 2 as a single injection reduced (P 0.04) detection of estrus compared to dividing the dose into two injections. There was an interaction (P 0.05) between synchronization and luteolytic treatments for P/AI on Day 35 after AI. For CIDR cows, method of PGF 2 administration had no effect on P/AI, whereas for G6G cows, dividing the dose into two injections improved (P 0.01) P/AI (Table 1). On Day 64 after AI, there was no interaction (P 0.17) between treatments. However, presynchronization tended (P 0.10) to increase P/AI compared with CIDR, and two injections of PGF 2 increased (P 0.02) P/AI compared with a single injection. There were no differences in pregnancy loss among treatments. As expected, cows having the ideal progesterone profile had greater (P 0.01) P/AI on Days 35 and 64 than those not having it, and they also had less (P 0.04) pregnancy loss between gestational Days 35 and 64 (Fig. 4). Cows with a CL at the first GnRH injection of the timed AI protocol had greater (P 0.01) P/AI than cows without CL (39.2 vs. 24.3%); this effect occurred despite synchronization and luteolytic treatments. Cows with low DIM had smaller (P 0.01) P/AI than those with moderate and high DIM (23.5 vs vs. 40.8%). Similarly, cows with BCS 2.50 had smaller (P 0.01) P/AI than cows with BCS of 2.75 to 3.00 or BCS 3.25 (28.7 vs vs. 40.9%). Additionally,

7 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx 7 Table 1 Effects of synchronization and luteolytic treatments on reproductive responses of grazing dairy cows (Experiment 1). Item Treatment a P b G6G CIDR Single PGF 2 Two PGF 2 Single PGF 2 Two PGF 2 S L S L [% (no./no.) or LSM SEM] CL on Day (147/179) 79.1 (148/187) 57.9 (110/190) 59.7 (108/181) Ovulation to 1st GnRH c 65.4 (117/179) 63.1 (118/187) 54.7 (104/190) 46.7 (84/180) New CL on Day 3 d 78.2 (140/179) 77.0 (144/187) 64.2 (122/190) 65.6 (118/180) CL on Day (173/179) 94.7 (177/187) 86.3 (164/190) 90.6 (163/180) Number CL on Day Progesterone (ng/ml) Day Day Ideal progesterone profile e 26.0 (13/50) 58.2 (32/55) 28.1 (16/57) 44.0 (22/50) Luteolysis f 61.7 (29/47) 96.2 (50/52) 82.0 (41/50) 95.7 (44/46) Estrus at AI 20.8 (37/178) 32.1 (60/187) 33.7 (64/190) 37.6 (68/181) Pregnant Day (51/178) 45.4 (84/185) 30.2 (57/189) 34.3 (61/178) Day (49/177) 40.0 (74/185) 26.1 (49/188) 29.2 (52/178) Pregnancy loss 3.9 (2/51) 11.9 (10/84) 14.0 (8/57) 14.8 (9/61) a All cows were subjected to the 5-d timed AI protocol. G6G injection of PGF 2 on study Day 16, followed by an injection of GnRH on study Day 14, and starting timed AI protocol on study Day 8; CIDR controlled internal drug-release insert containing 1.38 g of progesterone used from study Days 8 to 3; Single PGF 2 injection of 1 mg of cloprostenol as a single injection on study Day 3; two PGF 2 injection of 1 mg of cloprostenol divided into two injections of 0.5 mg on study Days 3 and 2. b S, effect of synchronization treatment (G6G vs. CIDR); L, effect of luteolytic treatment (single PGF 2 vs. 2 PGF 2 ); S L, interaction between S and L treatments. c Proportion of cows with a new CL on study Day 3 that had a follicle 8 mm on study Day 8. d Proportion of cows with a new CL on study Day 3 independent of the presence of follicle 8 mm on study Day 8. e Cows with concentration of progesterone in plasma 3.78 ng/ml on study Day 3 and 0.32 ng/ml on study Day 0. f Proportion of cows with concentration of progesterone in plasma 1 ng/ml on study Day 3 and 1 ng/ml on study Day 0. cows that experienced loss of BCS from AI to study Day 35, compared with those maintaining or gaining BCS, had lower (P 0.01) P/AI on Days 35 (30.5 vs vs. 39.4%) and 64 (27.1 vs vs. 35.1%) after AI. The BCS and DIM did not affect pregnancy loss between Days 35 and 64 of gestation Effects of breeds Body condition score did not differ (P 0.10) among breeds and averaged , and at AI, and , 2.74, 0.04 and on Day 35 for Holsteins, Jerseys, and crossbreds, respectively. Days in milk at AI differed (P 0.01) among breeds and were greater for Holsteins, followed by Jerseys and then crossbreds (Holstein vs. Jersey vs. crossbred ). Breed tended (P 0.07) to influence the proportion of cows with CL on study Day 8, and it was greater (P 0.03) for crossbred than Jersey cows (Table 2). However, there were no statistical differences among breeds for ovulation to the first GnRH, new CL, presence of CL, and number of CL on study Day 3, luteolysis, and estrus at AI (Table 2). Holstein cows had a lower (P 0.02) concentration of progesterone on study Day 3 than Jersey and crossbred cows, but these differences had disappeared at AI. The proportion of cows with ideal progesterone profile did not differ among breeds. Pregnancies per AI on Days 35 and 64 differed among breeds (P 0.01) because they were less for Holstein and Jersey than crossbred cows (Table 2). However, there was an interaction (P 0.01) between synchronization treatment and breed, and the lower P/AI for Holsteins than crossbreds only occurred when they received CIDR (Holstein 18.7 vs. Jersey 34.3 vs. crossbred 40.9%), but not when they were presynchronized with G6G (Holstein 39.1 vs. Jersey 31.0 vs. crossbred 37.4%) Ovarian responses and P/AI Ovulation to the first GnRH of the 5-day timed AI protocol did not affect P/AI (no ovulation 33.4 vs. ovulation 35.6%). Because luteolysis was compromised when twice the luteolytic dose of PGF 2 was administered as a single injection, further analyzes of the impact of ovulation to the first GnRH of the timed AI protocol were performed only for cows receiving twice

8 8 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx Probability of CL (%) Probability of CL (%) P < A G6G 40.0 CIDR the luteolytic dose divided into two injections. For the latter cows, ovulation to the first GnRH of the timed AI increased (P 0.01) P/AI (no ovulation 36.4 vs. ovulation 43.0%), and an interaction (P 0.04) between ovulation and breed was detected. Fertility was compromised in Holstein cows that did not ovulate to the first GnRH injection (no ovulation 20.0 vs. ovulation 42.4%), whereas there was no effect of ovulation to the first GnRH on P/AI for Jersey (no ovulation 36.0 vs. ovulation 36.4%) and crossbred cows (no ovulation 47.6 vs. ovulation 44.6%) Experiment 2 Postpartum interval (d) 100 P < B G6G 40 CIDR BCS at AI Fig. 3. Probability of cows having a CL at the first GnRH injection of a timed AI protocol (study Day 8) according to treatment and days postpartum (Panel A) or body condition score (Panel B) in Experiment 1. G6G, PGF 2 and GnRH on study Days 16 and 14, respectively, and starting the timed AI protocol on study Day 8. CIDR, no presynchronization and use of controlled internal drugrelease insert containing progesterone between study Days 8 and 3. Treatment influenced (P 0.05) the probability of CL in cows with fewer than 90 days postpartum or in cows with BCS at AI Days in milk and BCS at AI did not differ between treatments and averaged days and , respectively. Luteolytic treatments did not influence the expression of estrus at AI (two injections 34.1% vs. single injection 29.7%). Dividing the dose of PGF 2 into two treatments in the 5-day timed AI protocol increased P/AI on Days 35 (44.5 vs. 36.4%; P 0.02) and 64 (40.3 vs. 32.6%; P 0.04) after AI, but it did not alter pregnancy loss (Fig. 5). No interactions were detected between treatment and parity, BCS, or milk production on detection of estrus at AI, P/AI, and pregnancy loss. 4. Discussion Pregnancy per AI of dairy cows improved with presynchronization and dividing the dose of PGF 2 into two injections when cows were subjected to the 5-day timed AI protocol and received twice the luteolytic dose of cloprostenol (Experiment 1) or dinoprost (Experiment 2). The 5-day timed AI program has been used to reduce the period of follicle dominance and to improve P/AI in dairy cows [7] and heifers [15] and beef cows [16]. The limiting factor to reduce the period of follicle dominance in timed AI programs is regression of the newly formed CL resultant from ovulation to the first GnRH injection of the protocol. The newly formed CL is less responsive to luteolytic actions of PGF 2 in the first 5 days of development [17]. Ithas been proposed that the early CL has distinct molecular responses to PGF 2 compared with the midcycle CL that is typically sensitive to PGF 2 [17]. Previous studies indicated that the standard luteolytic dose of PGF 2 was not sufficient to optimize luteal % P < P < Ideal progesterone profile Not ideal progesterone profile P < Pregnant 35 d Pregnant 64 d Pregnancy loss Fig. 4. Pregnancy per AI on Days 35 and 64 after timed insemination and pregnancy loss for cows with or without ideal progesterone profiles on study Days 3 and 0 in Experiment 1. Cows with progesterone concentrations 3.78 ng/ml on study Day 3 and 0.32 ng/ml on study Day 0 were considered as having the ideal progesterone profile, based on receiver operator characteristic curves.

9 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx 9 Table 2 Effect of breed on reproductive responses of grazing dairy cows (Experiment 1). Item Breed P Holstein Jersey Crossbred [% (no./no.) or LSM SEM] CL on Day (181/250) ab 65.0 (52/80) a 68.8 (280/407) b 0.07 Ovulation to 1st GnRH c 56.4 (141/250) 50.0 (40/80) 59.6 (242/406) 0.15 New CL on Day 3 d 73.6 (184/250) 66.3 (53/80) 70.7 (287/406) 0.35 CL on Day (229/250) 90.0 (72/80) 92.6 (376/406) 0.10 Number CL on Day Progesterone (ng/ml) Day b a a 0.02 Day Ideal progesterone profile e 39.6 (38/96) 27.3 (9/33) 43.4 (36/83) 0.15 Luteolysis f 87.1 (74/85) 80.7 (25/31) 82.3 (65/79) 0.65 Estrus at timed AI 31.6 (79/250) 33.8 (27/80) 30.3 (123/406) 0.90 Pregnant Day b (70/249) 32.5 b (25/77) 39.1 a (158/404) 0.01 Day b (60/249) 27.3 b (21/77) 35.6 a (143/402) 0.01 Pregnancy loss 14.3 (10/70) 16.0 (4/25) 9.5 (15/158) 0.45 a,b Values on the same row with different superscript differ (P 0.05). c Proportion of cows with a new CL on study Day 3 that had a follicle 8 mm on study Day 8. d Proportion of cows with a new CL on study Day 3 independent of the presence of follicle 8 mm on study Day 8. e Cows with concentration of progesterone in plasma 3.78 ng/ml on study Day 3 and 0.32 ng/ml on study Day 0. f Proportion of cows with concentration of progesterone in plasma 1 ng/ml on study Day 3 and 1 ng/ml on study Day 0. regression and P/AI in the 5-day timed AI [7,18]. To solve this problem, the administration of a second luteolytic dose of PGF 2 7 to 24 h after the first dose resulted in improved luteolysis and fertility of dairy [7] and beef cows [18]. The disadvantage of this approach is that cows have to be handled an additional day. Furthermore, the mechanism by which the second dose % P = P = Pregnant 35 d Pregnant 64 d Pregnancy loss Single PGF 2 injection Two PGF 2 injections P = Fig. 5. Pregnancy per AI on Days 35 and 64 after timed insemination and pregnancy loss in lactating dairy cows subjected to 5-d timed AI protocol. Cows received 50 mg of dinoprost administered either as a single injection (n 327) on study Day 3 or divided into two injections (n 328) administered on study Days 3 and 2 (Experiment 2). enhances luteolysis is not clear. Because an interval as short as 7 h between injections was sufficient to improve fertility of beef cows [18], we hypothesized that increasing the dose as a single injection would result in similar CL regression and P/AI compared to the same amount of PGF 2 administered divided into two injections. The results of Experiments 1 and 2 clearly indicated that, for presynchronized cows, twice the luteolytic dose of PGF 2 either as cloprostenol or dinoprost administered as a single injection 5 days after GnRH was not sufficient to optimize CL regression and P/AI in the 5-day timed AI protocol. In Experiment 1, for cows that ovulated to the first GnRH and received a single PGF 2 injection, only 45.7% had the ideal progesterone concentration at AI ( 0.32 ng/ml), whereas for cows that received PGF 2 divided into two injections, this proportion was 78.1%. In fact, when only cows that ovulated to the first GnRH of the 5-day timed AI protocol were considered, luteolysis and P/AI on Day 35 were, respectively, 54.6 and 27.4% for G6G-S, 94.6 and 48.7% for G6G-T, 85.7 and 30.8% for CIDR-S, and 96.3 and 34.5% for CIDR-T. Therefore, the larger dose of PGF 2 administered as a single injection did not overcome the refractory response of the early CL to luteolysis observed in previous studies that used the conventional luteolytic dose [7]. Nevertheless, there were cows with a newly formed CL that under-

10 10 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx went complete luteolysis (45.7%) after single injection of 1 mg of cloprostenol on Day 5 after induction of ovulation. Age of the CL can vary, because not all cows have an LH surge and ovulate at the same time after treatment with GnRH. This subtle variation in age of the CL associated with individual changes in the process of luteinization might result in altered kinetics and molecular control of luteal development and consequent differences in the timing to undergo complete luteolysis in response to PGF 2 [19]. When cows were presynchronized, as in G6G in Experiment 1 and all cows in Experiment 2, dividing the PGF 2 into two injections was critical to fertility, regardless of type of PGF 2 used. However, the single PGF 2 injection did not affect fertility in non-presynchronized cows in Experiment 1. Similar responses have also been reported for dairy heifers [15] and for some [20], but not all studies with beef cattle [18], in which the use of the standard luteolytic dose of PGF 2 on Day 5 only was effective as twice the amount divided into two injections. In the current study, response to twice the luteolytic dose of PGF 2 given as a single injection in the 5-day timed AI protocol was dependent on the proportion of cows with a newly formed CL. In dairy heifers subjected to the 5-day timed AI protocol, ovulation to the initial GnRH was low, and 40% of them have a new CL on the day of the PGF 2 injection of the protocol [21]. Conversely, in presynchronized dairy cows, ovulation to the first GnRH is high, usually 60% [3,5,7]. Therefore, the rate of ovulation at the beginning of the protocol will determine the effectiveness of single administration of PGF 2 on Day 5, or the need for an additional PGF 2 treatment 7 to 24 h later. The competence of the ovulatory follicle is critical for fertility of lactating dairy cows; extending dominance typically reduces P/AI [22]. Dynamic biochemical and molecular changes occur in the follicle as it acquires and maintains dominance [23,24]. Reducing the interval from follicle emergence to ovulation improves early embryo quality [5] and P/AI of lactating dairy cows following synchronized ovulation [7] or spontaneous estrus [22]. As dominance extends, the oocyte is overexposed to LH pulses, which alters hormonal concentrations in the follicular fluid and results in compromised oocyte maturation [25]. These changes culminate with ovulation of an aged oocyte of reduced fertility [22,26]. By using the 5-day timed AI program, which limits follicle dominance, we initially hypothesized that supplementation with progesterone during the protocol would result in fertility similar to that of presynchronized cows. Contrary to that hypothesis, presynchronization of the estrous cycle with G6G improved fertility of dairy cows compared using a CIDR, as shown in Experiment 1, particularly in cows subjected to the two PGF 2 injections that optimized regression of newly formed CL. Presynchronization of grazing dairy cows resulted in similar ovarian responses and benefits to fertility as those previously described to high-producing dairy cows [1,3], including increased ovulation to the first GnRH and presence of CL at initiation of a timed AI protocol. The former benefited fertility when associated with adequate luteolysis and the latter had a major effect on fertility, regardless of the synchronization treatment. Likely, the progesterone supplemented by a single CIDR insert was not effective to reverse the negative effect of low endogenous progesterone concentrations at protocol initiation as previously suggested [27]. Nevertheless, in grazing cows in which the estrous cycles were synchronized with two doses of PGF 2 given 13 days apart, the use of a CIDR insert for 5 days before the second PGF 2 improved P/AI [28]. Interestingly, the benefit to P/AI was observed primarily in cows in early diestrus, which is characterized by low concentrations of endogenous progesterone. In Experiment 1, responses to synchronization treatments differed according to breed of cows. Jersey and crossbred cows did not benefit from presynchronization with G6G, whereas for Holstein cows, presynchronization and dividing the PGF 2 into two injections were critical to improve P/AI. In fact, compared with Jersey and crossbred cows, Holsteins were less fertile when subjected to the CIDR treatment. Although the presence of CL on the day of the first GnRH of the timed AI and ovulation to the first GnRH did not differ among breeds, Holstein cows had lower progesterone concentrations on study Day 3. This difference likely reflected catabolism of progesterone in larger and more productive cows with greater feed intake [29]. In that regard, progesterone concentration during development of the ovulatory follicle is critical to fertility [6]. Perhaps for Jersey and crossbred cows, presynchronization was less critical due to greater concentrations of progesterone resulting from the CIDR in these cows when compared with Holsteins. They are smaller and produce less milk than Holsteins, therefore, likely had less feed intake that has been shown to influence ovarian steroid catabolism [29]. Furthermore, the G6G, which increased ovulation to the first GnRH, only improved P/AI of Holstein cows, and not those of Jersey and crossbreds. In fact, for Holsteins, ovulation to the first GnRH of the 5-d timed AI protocol was critical to

11 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx 11 fertility when luteolysis was optimized with the two- PGF 2 injection treatment, but the same was not observed for Jerseys and crossbreds. 5. Conclusions Presynchronizing the estrous cycle and dividing the dose of PGF 2 into two injections increased P/AI in cows subjected to the 5-d timed AI protocol. The benefit of dividing the dose of PGF 2 was observed in grazing dairy cows receiving cloprostenol and highproducing dairy cows receiving dinoprost when presynchronization of the estrous cycle was used. Presynchronization of grazing dairy cows increased the proportion with a CL at the beginning of the timed AI protocol and ovulation to the first GnRH of the protocol, both of which improved fertility when adequate luteolysis was obtained. Nevertheless, the benefits of presynchronization were observed primarily in Holstein cows. For Jersey and crossbred cows, progesterone supplementation during the 5-d timed AI protocol constituted an alternative strategy that resulted in similar fertility compared with presynchronized cows. A single PGF 2 injection containing twice the standard luteolytic dose of cloprostenol or dinoprost on Day 5 of the timed AI protocol resulted in inadequate CL regression, especially in presynchronized cows, which compromised fertility of dairy cows. Acknowledgments The authors thank the owner and staff of Alliance Dairy (Trenton, Florida, USA), Piedmont Dairy (Trenton, Florida, USA) and South Point, Dairy (Chiefland, Florida, USA) for use of their cows and facilities. Our appreciation is extended to Nilo Francisco and Pat Cilo Cilo for assistance with conduct of the experiments, and Peter Hansen for comments and suggestions during the preparation of this manuscript. The first two authors contributed equally to this research. References [1] Moreira F, Orlandi C, Risco CA, Mattos R, Lopes F, Thatcher WW. Effects of presynchronization and bovine somatotropin on pregnancy rates to a timed artificial insemination protocol in lactating dairy cows. J Dairy Sci 2001;84: [2] El-Zarkouny SZ, Cartmill JA, Hensley BA, Stevenson JS. Pregnancy in dairy cows after synchronized ovulation regimens with or without presynchronization and progesterone. J Dairy Sci 2004;87: [3] Bello NM, Steibel JP, Pursley JR. Optimizing ovulation to first GnRH improved outcomes to each hormonal injection of Ovsynch in lactating dairy cows. J Dairy Sci 2006;89: [4] Vasconcelos JL, Silcox RW, Rosa GJ, Pursley JR, Wiltbank MC. Synchronization rate, size of the ovulatory follicle, and pregnancy rate after synchronization of ovulation beginning on different days of the estrous cycle in lactating dairy cows. Theriogenology 1999;52: [5] Cerri RL, Rutigliano HM, Chebel RC, Santos JE. Period of dominance of the ovulatory follicle influences embryo quality in lactating dairy cows. Reproduction 2009;137: [6] Bisinotto RS, Chebel RC, Santos JE. Follicular wave of the ovulatory follicle and not cyclic status influences fertility of dairy cows. J Dairy Sci 2010;93: [7] Santos JE, Narciso CD, Rivera F, Thatcher WW, Chebel RC. Effect of reducing the period of follicle dominance in a timed artificial insemination protocol on reproduction of dairy cows. J Dairy Sci 2010;93: [8] Ribeiro ES, Cerri RL, Bisinotto RS, Lima FS, Silvestre FT, Favoreto MG, et al. Reproductive performance of grazing dairy cows following presynchronization and resynchronization protocols. J Dairy Sci 2011;94: [9] McDougall S. Effects of treatment of anestrous dairy cows with gonadotropin-releasing hormone, prostaglandin, and progesterone. J Dairy Sci 2010;93: [10] Melendez P, Gonzalez G, Aguilar E, Loera O, Risco C, Archbald LF. Comparison of two estrus-synchronization protocols and timed artificial insemination in dairy cattle. J Dairy Sci 2006;89: [11] Chebel RC, Al-Hassan MJ, Fricke PM, Santos JE, Lima JR, Martel CA, et al. Supplementation of progesterone via controlled internal drug release inserts during ovulation synchronization protocols in lactating dairy cows. J Dairy Sci 2010;93: [12] Cerri RL, Rutigliano HM, Bruno RG, Santos JE. Progesterone concentration, follicular development and induction of cyclicity in dairy cows receiving intravaginal progesterone inserts. Anim Reprod Sci 2009;110: [13] Ferguson JD, Galligan DT, Thomsen N. Principal descriptors of body condition score in Holstein cows. J Dairy Sci 1994;77: [14] NRC. Nutrient requirements of dairy cattle. 7 th Rev Ed, Natl Acad Sci Washington, DC, [15] Rabaglino MB, Risco CA, Thatcher MJ, Kim IH, Santos JE, Thatcher WW. Application of one injection of prostaglandin F(2alpha) in the five-day Co-Synch CIDR protocol for estrous synchronization and resynchronization of dairy heifers. J Dairy Sci 2010;93: [16] Bridges GA, Helser LA, Grum DE, Mussard ML, Gasser CL, Day ML. Decreasing the interval between GnRH and PGF 2 from 7 to 5 days and lengthening proestrus increases timed-ai pregnancy rates in beef cows. Theriogenology 2008;15: [17] Miyamoto A, Shirasuna K, Sasahara K. Local regulation of corpus luteum development and regression in the cow: impact of angiogenic and vasoactive factors. Domest Anim Endocrinol 2009;37: [18] Kasimanickam R, Day ML, Rudolph JS, Hall JB, Whittier WD. Two doses of prostaglandin improve pregnancy rates to timed-ai in a 5-day progesterone-based synchronization protocol in beef cows. Theriogenology 2009;71: [19] Wiltbank MC, Ottobre JS. Regulation of intraluteal production of prostaglandins. Reprod Biol Endocrinol 2003;1:91.

12 12 E.S. Ribeiro et al. / Theriogenology xx (2012) xxx [20] Cruppe LH, Souto LA, Maquivar M, Gunn P, Mussard ML, Wolfenson D, et al. Use of two coincident doses of PGF 2 with the 5-d CO-synch plus CIDR estrous synchronization program. J Anim Sci 2010;88 (E-Suppl. 2):768 (Abstr.). [21] Lima FS, Ayres H, Favoreto MG, Bisinotto RS, Greco LF, Ribeiro ES, et al. Effects of gonadotropin-releasing hormone at initiation of the 5-d timed artificial insemination (AI) program and timing of induction of ovulation relative to AI on ovarian dynamics and fertility of dairy heifers.. J Dairy Sci 2011;94: [22] Bleach EC, Glencross RG, Knight PG. Association between ovarian follicle development and pregnancy rates in dairy cows undergoing spontaneous oestrous cycles. Reproduction 2004; 127: [23] Aerts JM, Bols PE. Ovarian follicular dynamics. A review with emphasis on the bovine species. part II: Antral development, exogenous influence and future prospects. Reprod Domest Anim 2010;45: [24] Tripathi A, Kumar KV, Chaube SK. Meiotic cell cycle arrest in mammalian oocytes. J Cell Physiol 2010;223: [25] Revah I, Butler WR. Prolonged dominance of follicles and reduced viability of bovine oocytes. J Reprod Fert 1996;106: [26] Mihm M, Baguisi A, Boland MP, Roche JF. Association between the duration of dominance of the ovulatory follicle and pregnancy rate in beef heifers. J Reprod Fertil 1994;102: [27] Lima JR, Rivera FA, Narciso CD, Oliveira R, Chebel RC, Santos JE. Effect of increasing amounts of supplemental progesterone in a timed artificial insemination protocol on fertility of lactating dairy cows. J Dairy Sci 2009;92: [28] Xu ZZ, Burton LJ, Macmillan KL. Reproductive performance of lactating dairy cows following estrus synchronization regimens with PGF 2 and progesterone. Theriogenology 1997;47: [29] Wiltbank M, Lopez H, Sartori R, Sangsritavong S, Gümen A. Changes in reproductive physiology of lactating dairy cows due to elevated steroid metabolism. Theriogenology 2006; 65:17 29.