Effects of additional prostaglandin F2α and estradiol-17β during Ovsynch in lactating dairy cows
Article Outline
- Abstract
- Introduction
- Materials and Methods
- Results
- Discussion
- Acknowledgments
- Supplementary data
- References
- Copyright
Abstract
This study was designed to evaluate whether decreasing circulating progesterone (P4) or increasing circulating estradiol-17β (E2) near the time of artificial insemination (AI) in an Ovsynch protocol would increase pregnancies per AI (P/AI) in lactating dairy cows. Six hundred nineteen lactating Holstein cows (n
=772 inseminations) received Ovsynch (GnRH–7 d-PGF2α–56 h-GnRH–16 h-timed AI). Cows were randomized in a 2×2 factorial experiment of 4 treatments to receive or not receive 25mg of PGF2α 24h after the standard PGF2α of Ovsynch, or 0.5mg of E2 at the time of the final GnRH of Ovsynch, or both. Blood samples were collected 24h after normal PGF2α and at final GnRH to evaluate circulating P4. Ovarian ultrasound was done at final GnRH to determine preovulatory follicle size. Ovulation was confirmed by ultrasound 5 d after AI. Treatment with additional PGF2α increased the percentage of cows that had complete luteal regression (95.6%) compared with control cows (84.6%). In contrast, additional PGF2α had no detectable effect on P/AI (control = 41.5% vs. + PGF2α=44.7%). Supplementation with E2 increased expression of estrus (84.4 vs. 37.2%), but had no effect on overall fertility and even tended to have a negative effect on fertility in cows that ovulated to the second GnRH (control = 51.5% vs. +E2=44.0%). Thus, additional treatments with PGF2α or E2 during Ovsynch can be used to increase synchronization and expression of estrus during Ovsynch, although the lack of improvement in fertility makes these treatments unwarranted.
Key words: additional prostaglandin F2α, estradiol, Ovsynch
Introduction
Reproductive efficiency has declined over the past several years as milk yield has increased in lactating dairy cows (Washburn et al., 2002). As milk production increases, duration of estrus decreases (Lopez et al., 2004; Wiltbank et al., 2006), and current rates of detection of estrus on dairy farms are generally not optimal for efficient reproductive management programs (Nebel and McGilliard, 1993; Senger, 1994; Peralta et al., 2005). Timed AI protocols such as Ovsynch (GnRH–7 d–PGF2α–2 d-GnRH–16 h–AI) were developed to reduce or eliminate dependence of reproductive management on detection of estrus (Pursley et al., 1995, 1997a,b). Although Ovsynch improves submission rates, pregnancies per AI (P/AI) remain similar to that in cows inseminated based on detection of estrus (1997a,b).
In spite of the management advantages of the Ovsynch protocol, circulating hormone concentrations are likely not optimized during this protocol. For instance, previous studies have indicated that incomplete luteal regression occurs in some cows that are treated with Ovsynch (Pursley et al., 1997b; Moreira et al., 2000, 2001; Souza et al., 2007). Increased circulating progesterone (P4) concentrations near the time of AI could produce a suboptimal environment in the reproductive tract for sperm and ovum transport and result in lowered fertility. Rates of PGF2α-induced corpus luteum (CL) regression vary among studies (109/116=94%, Pursley et al., 1997b; 265/327=81%, Moreira et al., 2000; 143/150=95%, Gümen et al., 2003). Additionally, other studies have reported varying percentages of cows with low P4 by 48h following PGF2α, regardless of high or low P4 at the time of PGF2α (195/226=86.3%, Cartmill et al., 2001; 513/611=84%, El-Zarkouny et al., 2004; 867/919=94%, Souza et al., 2007). Elevation of circulating P4 >0.5 ng/mL near timed AI reduced fertility by more than 50% (Souza et al., 2007). Therefore, our study was designed to determine whether a second PGF2α treatment 1 d after the first PGF2α could lower P4 and increase P/AI during Ovsynch.
In addition to low circulating P4 concentrations near AI, an elevated estrogen concentration is critical for optimal fertility (Hawk and Cooper, 1975; Sartori et al., 2002b; Souza et al., 2007). During Ovsynch, the second GnRH is given to precisely induce the time of ovulation; however, circulating estradiol-17β (E2) concentrations may not have reached optimal concentrations, as indicated by a lack of estrus in most cows treated with Ovsynch (Souza et al., 2007). In addition, lactating dairy cows have an elevated rate of E2 metabolism (Sangsritavong et al., 2002; Wiltbank et al., 2006) that decreases circulating E2 concentrations (Sartori et al., 2002a; Lopez et al., 2004). Many reproductive problems may be caused by reduced E2 during the periovulatory period such as inefficient sperm transport, suboptimal oviductal/uterine environment, impaired oocyte fertilization, and poorer embryonic quality (Hawk and Cooper, 1975; Ryan et al., 1993; Sartori et al., 2002b). In a recent study, supplementation with 1mg of E2 at 8h before the final GnRH of Ovsynch did not cause an overall improvement in fertility but increased fertility in cows that ovulated a medium-sized follicle, cows with low BCS, cows at first service, and in primiparous cows (Souza et al., 2007). Nevertheless, another study reported no improvement in fertility when 0.25mg of estradiol cypionate (ECP) was given at the time of the final GnRH of Ovsynch (Sellars et al., 2006). In the present study, we chose to use 0.5mg of native E2, a dose that has previously been shown to elevate E2 to normal preovulatory concentrations (Souza et al., 2005). For ease of conducting the experiment, this dose of E2 was given at the time of second GnRH during the Ovsynch protocol. This was similar to the timing of ECP treatment in a previous study (Sellars et al., 2006), but 8h later than in a previous study with E2 supplementation (Souza et al., 2007).
Based on previous research in which E2 supplementation was only effective in cows ovulating medium-sized follicles (Souza et al., 2007), we designed a procedure to increase synchrony of follicular development during our reproductive management protocol. For first AI, we did this by presynchronizing cows with Ovsynch and then initiating a second Ovsynch procedure at 7 d after the final GnRH of the presynchronization Ovsynch, a procedure that we have termed double-Ovsynch (Souza et al., 2008). In addition, we resynchronized cows with a GnRH injection at 7 d before the start of the normal GnRH of Resynch. These procedures were done not only to increase fertility during Ovsynch and Resynch, but also to improve follicular/luteal synchrony during Ovsynch to increase the likelihood of detecting positive effects of supplemental treatments.
Thus, the overall objective of this research was to test the effects of decreasing circulating P4 and increasing circulating E2 near timed AI during Ovsynch. We hypothesized that treatment with additional PGF2α would reduce circulating P4 near AI and increase fertility during Ovsynch. We expected that the cows that would be most likely to benefit from the extra PGF2α treatment would be those with elevated P4 at the time of the second PGF2α treatment. Our second hypothesis was that supplementation with 0.5mg of E2 at the time of the second GnRH would increase expression of estrus and improve fertility in cows ovulating medium-sized follicles, during first service, and in primiparous cows. We anticipated that presynchronization with double-Ovsynch would produce a high percentage of cows with medium-sized follicles (optimizing response to E2) and increase the percentage of cows with a d 7 CL (potentially less responsive to PGF2α; increasing the importance of a second PGF2α) during Ovsynch.
Materials and Methods
Cows
Six hundred nineteen lactating Holstein cows, receiving a total of 782 inseminations, were housed in free-stalls on 2 commercial dairy farms near Middleton and Sauk City, Wisconsin. Although 782 inseminations were performed during this study, 2 cows were sold and 1 died before the pregnancy diagnosis at farm A, and 4 cows were sold before the pregnancy diagnosis on farm B. Additionally, 2 cows at farm B were rebred again on the day after timed AI, and 1 cow accidentally received an extra injection. Thus, 10 cows were removed from the study and our final results utilized information from 772 inseminations.
Normal Reproductive Management of Herds
All cows receiving first service were presynchronized beginning at 48 to 54 DIM with an injection of GnRH (100μg; OvaCyst, IVX Animal Health, St. Joseph, MO). Seven days later, all cows received an injection of PGF2α (25mg; ProstaMate, IVX Animal Health), followed 3 d later by another injection of GnRH. One week later, all cows began the breeding Ovsynch (GnRH–7 d–PGF2α–56 h–GnRH–16 h-timed AI). No detection of estrus was performed at either farm before first AI.
After first AI, both farms performed detection of estrus in returning cows. At farm A, a professional AI technician used tail chalk on cows daily, and cows were inseminated by 1 technician based on removed tail chalk. Estrus was detected in 12.9% of cows that had received a previous AI and would have gone on to receive a pregnancy evaluation and Resynch. At farm B, the farm owner visually detected estrus daily and performed all of the inseminations. Estrus was detected in 21.6% of cows that had received a previous AI and would have gone on to have a pregnancy evaluation and Resynch. Pregnancy diagnosis was performed in all cows at d 39 after AI. Cows not detected in estrus after their previous AI began Resynch with an injection of GnRH on d 25 after AI. Seven days later, these cows began Ovsynch with an injection of GnRH (d 32). Seven days after that, cows were diagnosed as pregnant or open (d 39 after AI). At farm A, the veterinarian used ultrasound in the pregnancy diagnoses, whereas the veterinarian at farm B used rectal palpation as a means to diagnose pregnancy. Only cows diagnosed as open continued the Resynch protocol by receiving an injection of PGF2α, followed 56h later by GnRH, with AI 16h after GnRH.
All cows, regardless of service number, received an injection of human chorionic gonadotropin (hCG, 2,000 IU; Chorulon, Intervet, Millsboro, DE) on d 6 after the final GnRH injection of Ovsynch. This treatment was done in the routine reproductive management of both farms based on previous results showing an improvement in fertility with this treatment (Santos et al., 2001). All hormone injections were administered intramuscularly.
Experimental Design
The experimental period was from November 2006 to March 2007. This study had a 2×2 (±PGF2α and ±E2) factorial design consisting of 4 treatments with all cows randomly assigned to 1 of 4 treatment groups (see Figure 1). Thus, one treatment was Ovsynch with no additional treatments (control; n=197), the second treatment group consisted of an additional injection of PGF2α (25mg) at 1 d after the normal PGF2α during Ovsynch (+PGF2α; n=182), the third treatment consisted of an injection (0.5mg i.m.) of E2 concurrently with the final GnRH injection (+E2; n=196), and the fourth treatment consisted of administering both the PGF2α and the E2 (+PGF2α+E2; n=197). The University of Wisconsin-Madison, College of Agriculture and Life Sciences Animal Care Committee approved all animal procedures.
Figure 1.
All cows received Ovsynch and were randomized to receive or not receive an additional 25-mg PGF2α injection on 1 d after the normal PGF2α of Ovsynch and to receive (or not) 0.5
mg of estradiol-17β (E2) concurrently with the final GnRH injection. BS1 = blood sample 1, collected at 1 d after the normal PGF2α of Ovsynch; BS2 = blood sample 2, collected at the time of the final GnRH. Ultrasound was performed at the final GnRH and also 6 d later (not shown); TAI = timed AI.
The E2 solution was prepared as follows: E2 (Sigma Chemical Co., St. Louis, MO) was weighed and benzyl alcohol (EM Science, Cherry Hill, NJ) added to bring the solution to 5 mg/mL. Sesame oil (EM Science) was then added to the preparation to obtain a final solution of 0.5 mg/mL.
Blood samples were collected via coccygeal venipuncture from most cows (90.4%) on the day after normal PGF2α during Ovsynch (BS1) and again at the time of the final GnRH during Ovsynch (BS2). Expression of estrus was evaluated in cows near AI using the same methods of visual estrus detection used on both farms at other times and data were recorded for our analysis. No estrus data were recorded in the first week on farm A, and the last 2 wk of estrus data were accidentally discarded at farm B leaving estrus data available for 646 AI (83.9% of total). In addition, ovarian ultrasound examinations (Easi-Scan, BCF Technology Ltd., Livingston, United Kingdom) were performed on most cows (82.4%) at the time of the final GnRH injection to determine the size of the dominant follicle. If a cow did not have any follicles >8mm in diameter, she did not continue in the study (n=7). A second scan was performed 6 d after the final GnRH injection to determine the percentage of cows that ovulated to the second GnRH treatment. Ovulation was designated if the dominant follicle disappeared and a CL was detected on the same ovary. If it was unclear which follicle ovulated (n=79), the cow was included in ovulation percentage analyses, but not in analyses of ovulatory follicle size.
Hormone Assays
Blood samples were centrifuged at 1,600×g for 20min and then stored at −20°C until assayed for P4 concentrations. Circulating P4 was evaluated from unextracted sera using an antibody-coated tube RIA kit (Diagnostic Products Corporation, Los Angeles, CA). The sensitivity of the P4 assay was 0.01 ng/mL. The intra- and interassay coefficients of variation were 6.5 and 6.0%, respectively.
Statistical Analysis
The analyses were performed using the GLIMMIX procedure in SAS (SAS Institute, Cary, NC). For analysis of P/AI, there was no farm (P=0.9) or treatment×farm interaction (P=0.8), so the data from both farms were combined. Farm (A or B), treatment (control, +PGF2α, +E2, and +PGF2α+E2), service number (first or later service), parity (primiparous or multiparous), and interactions among these variables were all included in the original model for analysis of P/AI. Only treatment and service number remained in the final model. All other variables were found not to have a significant effect (P>0.10) on P/AI and were removed from the model. Similar analyses were done for other, less critical, variables that were analyzed in the study (percentage ovulating to second GnRH, percentage with low P4, percentage expressing estrus).
The Guided Data Analysis application of SAS was used to test the variables for normality of residuals and homogeneity of variances. Normality of the data for circulating P4 could not be achieved through data transformation, so the NPAR1WAY procedure for nonparametric analysis was used.
Results
General Results
The overall percentage of cows that ovulated to the second GnRH treatment was 89.9% (572/636), with double ovulations occurring in 4.83% (25/518) of the cows that ovulated. There was a tendency for an effect of service number and parity on percentage of cows that ovulated to the second GnRH treatment, with primiparous cows having a lesser (P=0.070) percentage ovulating (86.4%; 140/162) compared with multiparous cows (91.1%; 432/474), and first-service cows with a greater (P=0.106) percentage ovulating (91.5%; 344/376) than Resynch cows (87.7%; 228/260).
To determine a cutoff point to designate cows in which the CL had fully regressed in this experiment, an analysis was done between P4 concentration at the time of the final GnRH treatment and P/AI (shown in Figure 2). A value of 0.4 ng/mL was chosen from this analysis based on a reduction of more than 50% in P/AI compared with cows with lesser P4. This value is similar to that determined in a previous experiment (0.5 ng/mL) in which we used an ELISA analysis of P4 concentration at 48h after the PGF2α treatment (8h sooner) of Ovsynch (Souza et al., 2007). In all subsequent data analyses of low P4 at the time of the second GnRH (BS2), a value of 0.4 ng/mL was used as the cutoff. For analysis of P4 at BS1 (1 d after first PGF2α treatment), results were reclassified in such a way as to provide approximately even numbers of cows per analyzed group and therefore did not use this cutoff.
Figure 2.
Effect of different circulating progesterone (P4) concentrations at second GnRH treatment on pregnancies per AI (P/AI). Levels of P4 concentrations are 1: <0.05 ng/mL (n
=74); 2: 0.05–0.099 ng/mL (n=115); 3: 0.1–0.149 ng/mL (n=131); 4: 0.15-0.199 ng/mL (n=112); 5: 0.2–0.249 ng/mL (n=80); 6: 0.25–0.299 ng/mL (n=59); 7: 0.3–0.399 ng/mL (n=56); 8: 0.4-0.699 ng/mL (n=37); and 9: ≥0.7 ng/mL (n=33). A,BBars with different letters tend to differ (P<0.1). In addition, group 9 was different from groups 1 (P=0.050), 2 (P=0.012), 3 (P=0.007), and 4 (P=0.010); group 8 was different from groups 3 (P=0.034) and 4 (P=0.049).
As shown in Table 1, there was no treatment effect (P=0.349) on P/AI by analysis of the 4 different treatments. There was no interaction between the +PGF2α and the +E2 treatments (P>0.10); therefore, subsequent analyses focused on the main effects of these 2 treatments. Regardless of treatment, an effect of service number was detected (P<0.001) on P/AI, with first-service cows having greater P/AI (49.8%, n=453) than cows at later services (33.5%, n=319). Treatment also had no effect on percentage of cows that ovulated to the second GnRH treatment of Ovsynch (P=0.427). In contrast, cows receiving +PGF2α and +PGF2α+E2 had a greater (P<0.001) percentage of cows with luteal regression (95.7 and 95.5%, respectively) as indicated by low P4 (<0.4 ng/mL) at the time of the final GnRH compared with cows in the control (84.8%) and +E2 (84.3%) treatments. In addition, more (P<0.001) cows expressed estrus in the +E2 (83.3%) and +PGF2α+E2 (85.4%) treatments compared with control (38.4%) and +PGF2α (35.9%) cows.
Table 1. Summary of the effect of all treatments for overall pregnancies per AI (P/AI), percentage ovulating to the second GnRH of Ovsynch, percentage of cows with low circulating progesterone (P4) at the time of the final GnRH, and percentage of cows expressing estrus
| Item | Treatment1 | Treatment effect | |||
|---|---|---|---|---|---|
| Control | +PGF2α | +E2 | +PGF2α+E2 | ||
| P/AI, % (n/total) | 39.6 (78/197) | 48.6 (88/182) | 43.4 (85/196) | 41.1 (81/197) | 0.349 |
| Ovulation to second GnRH, % (n/total) | 86.3 (139/161) | 89.4 (135/151) | 92.5 (148/160) | 91.5 (150/164) | 0.427 |
| Low P4 at second GnRH, % (n/total) | 84.8a (151/178) | 95.7b (155/162) | 84.3a (150/178) | 95.5b (171/179) | <0.001 |
| Expression of estrus, % (n/total) | 38.4a (63/164) | 35.9a (52/145) | 83.3b (135/162) | 85.4b (135/158) | <0.001 |
a,bMeans with different superscript letters within a row differ (P |
1Treatments: control = Ovsynch with no additional treatments; +PGF2α = additional injection of PGF2α (25 |
Effect of Extra PGF2α Treatment
In Table 2 are shown the main effects of an extra PGF2α treatment. There was no difference (P=0.340) in overall P/AI for cows receiving or not receiving an additional PGF2α treatment 1 d after the normal PGF2α of Ovsynch (at the time of BS1; 44.7 vs. 41.5%). Percentage of cows that ovulated to the second GnRH of Ovsynch was also not different (P=0.657) between the treatments (90.4 vs. 89.4%). After removal of cows that did not ovulate to the second GnRH treatment of Ovsynch, there was a somewhat higher P/AI than without correcting for ovulation, but there was still no difference (P=0.514) in P/AI for cows receiving the additional PGF2α (control=46.0%; +PGF2α=49.1%). For the third analysis, cows that had high P4 at the time of the final GnRH injection were removed. It was anticipated that this correction would remove any effect of PGF2α treatment because cows without CL regression would be removed from the analysis. Again, there was a somewhat greater P/AI than without any corrections, particularly for the cows that did not receive the additional PGF2α, and there was no difference in P/AI for PGF2α main effects (50.0 vs. 50.8%; P=0.769) after this correction. As expected, more (P<0.001) cows had low P4 (<0.4 ng/mL) at the time of the final GnRH injection in the cows receiving a second injection of PGF2α during Ovsynch (95.6%) compared with cows that did not receive a second PGF2α (84.6%; Table 2).
Table 2. Effects of an additional treatment with PGF2α on pregnancies per AI (P/AI), percentage of cows ovulating to the second GnRH, and percentage of cows with regressed corpus luteum as indicated by low circulating progesterone (P4; <0.4 ng/mL) at second GnRH treatment
| Treatment | Overall P/AI, % (n/total) | Ovulation to second GnRH, % (n/total) | P/AI in ovul. cows,1% (n/total) | Low P4 at second GnRH, % (n/total) | P/AI-Low P4 and ovul. cows, % (n/total) |
|---|---|---|---|---|---|
| −PGF2α | 41.5 (163/393) | 89.4 (287/321) | 46.0 (132/287) | 84.6 (301/356) | 50.8 (125/246) |
| +PGF2α | 44.7 (169/379) | 90.4 (285/315) | 49.1 (140/285) | 95.6 (326/341) | 50.0 (134/268) |
| P-value | 0.340 | 0.657 | 0.514 | <0.001 | 0.769 |
1Ovul. cows = cows that ovulated in response to the final GnRH. |
Table 3 shows the effect of service number on the response to the second PGF2α treatment. There was a tendency (P=0.073) for an effect of service number on the percentage of cows with low P4 at the second GnRH treatment (first service = 374/408, 91.7%; later service = 253/289, 87.5%). More first-service cows receiving an additional PGF2α had low P4 (97.5%) compared with −PGF2α cows at first service (86.1%; P=0.001) or −PGF2α cows at later services (82.4%; P<0.001). The cows at later services that received an additional PGF2α had an intermediate percentage of cows with low P4 at second GnRH (92.9%), and this value was different (P=0.045) compared with cows at later services that did not receive an additional PGF2α (82.4%). There was no significant effect of the second PGF2α treatment on P/AI in cows either at first service or later services. In contrast, there was an effect of service number on P/AI with first service being higher than later services whether the second PGF2α treatment was given (52.7 vs. 33.5%; P=0.002) or not given (47.0 vs. 33.5%; P=0.041).
Table 3. Effects of an additional PGF2α treatment at 24 h after first PGF2α treatment on pregnancies per AI (P/AI) and percentage of cows with low progesterone (P4) at final GnRH for cows at first or later services
| Treatment | P/AI, % (n/total) | P-value | Low P4, % (n/total) | P-value | ||
|---|---|---|---|---|---|---|
| First service | Later services | First service | Later services | |||
| −PGF2α | 47.0 (109/232) | 33.5 (54/161) | 0.041 | 86.1 (179/208) | 82.4 (122/148) | 0.788 |
| +PGF2α | 52.7 (116/221) | 33.5 (53/158) | 0.002 | 97.5 (195/200) | 92.9 (131/141) | 0.208 |
| P-value | 0.241 | 1.000 | 0.001 | 0.045 | ||
As discussed previously and shown in Figure 2, P/AI were lower (P<0.05) or tended to be lower (P<0.1) for cows with greater P4 (≥0.4 ng/mL) at BS2. To more closely evaluate the relationship between P4 concentrations at BS1 and BS2 and the effect of an additional PGF2α treatment, cows were reclassified into 7 classes (of approximately 50 cows without additional PGF2α and 50 cows receiving an additional PGF2α in each class) based on P4 concentration at BS1. Table 4 shows the P4 concentrations at BS2 in these 7 classes for cows that received or did not receive the second treatment with PGF2α. The additional PGF2α injection reduced or tended to reduce P4 concentrations in cows with P4 concentrations of 0.450 ng/mL or greater at BS1. In contrast, when P4 was below this concentration at BS1, the additional PGF2α treatment had no effect (P>0.1) in further reducing P4 concentrations compared with cows that did not receive the additional PGF2α.
Table 4. Concentrations of progesterone (P4) at the time of second GnRH treatment (BS2) based on P4 concentrations at 24 h after PGF2α treatment (BS1) when the additional PGF2α treatment was administered1
| P4 at BS1 (ng/mL) | P4 at BS2 (ng/mL) by treatment | P-value | |
|---|---|---|---|
| −PGF2α | +PGF2α | ||
| ≥1.275 | 1.178 | 0.338 | 0.014 |
| 0.9–1.274 | 0.328 | 0.203 | 0.004 |
| 0.725–0.899 | 0.374 | 0.209 | 0.056 |
| 0.575–0.724 | 0.215 | 0.142 | 0.010 |
| 0.450–0.574 | 0.168 | 0.155 | 0.100 |
| 0.270–0.449 | 0.175 | 0.117 | 0.268 |
| ≤0.269 | 0.101 | 0.065 | 0.241 |
1BS1 = blood sample collected 1 d after the normal PGF2α injection; BS2 = blood sample collected at the time of the final GnRH injection (56 h after normal PGF2α). |
To perform more valid analyses of binomial data by P4 concentration, the data were divided into thirds, based on the P4 concentration at BS1 (Table 5). This division resulted in cows being classified as high (P4 >0.878 ng/mL; n=228); medium (0.878 ng/mL≥P4≥0.515 ng/mL; n=228), or low (P4 <0.515 ng/mL; n=229) P4 at BS1. Treatment with an additional PGF2α increased the percentage of cows with low P4 (<0.4 ng/mL) at BS2 (time of the second GnRH) in cows classified as high or medium P4 at BS1 but did not change this percentage in cows classified as low P4 at BS1. This was obviously because most cows (∼96%) with lower P4 at BS1 also had low P4 at second GnRH (BS2). Surprisingly, the additional PGF2α treatment had no effect (P=0.451) on P/AI in cows that had high P4 (control = 33.0% vs. +PGF2α=37.7%) at the time of BS1, despite the dramatic increase in percentage of cows with low P4 concentrations after the second PGF2α treatment in these cows. Cows classified with medium P4 concentrations also had no increase (P=0.234) in P/AI in cows treated with the second PGF2α (52.3%) compared with nontreated cows (44.6%). As expected, the second PGF2α treatment had no effect on P/AI for the cows with low P4 at BS1 (47.1 vs. 46.4%; P=0.894).
Table 5. Effects of an additional PGF2α on the percentage of cows with low circulating progesterone (P4; ≤0.4 ng/mL) at the time of the final GnRH and pregnancies per AI (P/AI) after Ovsynch in cows classified into 3 groups based on P4 concentration at BS1 (just before additional PGF2α)1
| Item | Progesterone class2 at BS1 | P-value for contrast | ||||
|---|---|---|---|---|---|---|
| High (H) | Medium (M) | Low (L) | H vs. L | H vs. M | M vs. L | |
| Low P4 at BS2, % (n/total) | ||||||
| −PGF2α | 63.2 (67/106) | 91.7 (111/121) | 95.8 (114/119) | <0.001 | <0.001 | 0.567 |
| +PGF2α | 91.0 (111/122) | 97.2 (104/107) | 99.1 (109/110) | 0.071 | 0.196 | 0.958 |
| P-value | <0.001 | 0.089 | 0.158 | |||
| P/AI, % (n/total) | ||||||
| −PGF2α | 33.0 (35/106) | 44.6 (54/121) | 47.1 (56/119) | 0.064 | 0.148 | 0.994 |
| +PGF2α | 37.7 (46/122) | 52.3 (56/107) | 46.4 (51/110) | 0.448 | 0.052 | 0.628 |
| P-value | 0.451 | 0.234 | 0.894 | |||
1BS1 = blood sample collected 1 d after the normal PGF2α injection; BS2 = blood sample collected at the time of the final GnRH injection (56 h after normal PGF2α). |
2Classes: low: <0.515 ng/mL; n |
A comparison was made of P/AI in cows with differing concentrations of P4 at BS1 (using division into thirds as described in previous paragraph). When all cows were analyzed (with or without PGF2α; data not shown), the cows with high P4 BS1 had lower P/AI than cows with medium (high = 35.5% vs. medium = 48.2%; P=0.018) or low (high = 35.5% vs. low = 46.7%; P=0.045) P4. When only cows without supplemental PGF2α were analyzed (Table 5), there was a tendency (P=0.064) for cows with high P4 to have lower P/AI than cows with low P4 at BS1. When only cows receiving supplemental PGF2α were analyzed, there was a tendency (P=0.052) for cows with high P4 to have lower P/AI than cows with medium P4 at BS1.
Estradiol-17β
Table 6 shows the main effects of E2 treatment on various experimental endpoints. When all cows were included in the analysis, the percentage of cows that ovulated to the second GnRH treatment tended to differ (P=0.057) among treatments. Therefore, analyses were made including only cows that ovulated to the second GnRH of Ovsynch (middle column). In addition, a final analysis was done including only cows that ovulated to the second GnRH and had low P4 at the time of second GnRH of Ovsynch (final columns).
Table 6. Effects of supplementation with 0.5mg of estradiol-17β (+E2) for all cows, or including only those cows that ovulated to the final GnRH, or only cows that ovulated and had low circulating progesterone (P4; <0.4 ng/mL) at final GnRH during Ovsynch
| Item | All cows | Ovulating cows | Low P41 and ovulating cows | ||||||
|---|---|---|---|---|---|---|---|---|---|
| −E2 | +E2 | P-value | −E2 | +E2 | P-value | −E2 | +E2 | P-value | |
| Ovulation to second GnRH, % (n/total) | 87.8 (274/312) | 92.0 (297/323) | 0.057 | 100 | 100 | — | 100 | 100 | — |
| P/AI, % (n/total) | 43.9 (166/379) | 42.2 (166/393) | 0.731 | 51.5 (141/274) | 44.0 (131/298) | 0.108 | 54.4 (136/250) | 46.6 (123/264) | 0.105 |
| Expression of estrus, % (n/total) | 37.2 (115/309) | 84.4 (270/320) | <0.001 | 36.6 (86/235) | 84.7 (216/255) | <0.001 | 37.2 (80/215) | 87.7 (200/228) | <0.001 |
| P/AI, % (n/total) | |||||||||
| Cows with estrus | 53.9 (62/115) | 45.6 (123/270) | 0.511 | 61.6 (53/86) | 46.3 (100/216) | 0.114 | 65.0 (52/80) | 47.5 (95/200) | 0.065 |
| Cows without estrus | 39.7 (77/194) | 30.0 (15/50) | 0.723 | 47.7 (71/149) | 38.5 (15/39) | 0.842 | 50.4 (68/135) | 46.4 (13/28) | 0.983 |
| P-value | 0.081 | 0.248 | 0.180 | 0.862 | 0.167 | 0.167 | |||
1Low P4 is defined as <0.4 ng/mL at time of second GnRH treatment of Ovsynch. |
There was no difference (P=0.731) in overall P/AI for cows receiving or not receiving E2 (42.2 vs. 43.9%, respectively; Table 6). However, after removing cows that did not ovulate, the cows not receiving E2 tended (P=0.108) to have a greater P/AI than cows receiving E2 (51.5 vs. 44.0%). This trend in lower P/AI for E2-treated cows continued once all cows were removed that did not have full CL regression by the time of the final GnRH (54.4 vs. 46.6%, control vs. E2; P=0.105).
As expected, a greater (P<0.001) percentage of cows expressed estrus in the group receiving E2 than in those not receiving E2 (84.4 vs. 37.2%, respectively; Table 6). There was no difference in P/AI between cows receiving or not receiving E2 when only cows that expressed estrus were used in the analysis (45.6 vs. 53.9%, respectively; P=0.511) and similarly when only cows that did not express estrus were used in the analysis (30.0 vs. 39.7%, respectively; P=0.723). In contrast, when all cows were removed that did not ovulate or have complete CL regression, there was a tendency (P=0.065) for cows not receiving E2 and showing estrus to have a greater P/AI (65.0%) compared with cows that received E2 and expressed estrus (47.5%). This difference was not observed for cows not expressing estrus.
Expression of estrus was affected by ovulatory follicle size (data not shown) only in the cows not receiving E2 (P=0.030) but not in the E2 treatment group (P>0.1). For cows not receiving E2, those that ovulated medium-sized follicles (14–18mm) had a greater estrous response (40.5%; n=170) compared with cows ovulating small follicles (<14mm; 18.6%; n=58), with cows ovulating large follicles (>18mm) having intermediate estrous expression (26.7%; n=25).
We had hypothesized that first-service cows and cows ovulating medium-sized follicles would benefit from E2 supplementation at the time of the final GnRH based on previous results (Souza et al., 2007). However, there was no difference (P=0.998) in P/AI between the 2 groups (Table 7) at first service (−E2=50.2% vs. +E2=49.3%) or later services (−E2=34.4% vs. +E2=32.7%). However, there was a difference between first versus later services in cows treated or not treated with supplemental E2 (Table 7).
Table 7. Effects of supplementation with 0.5 mg of estradiol-17β (+E2) at the final GnRH of Ovsynch in cows during their first or later services on overall pregnancies per AI (P/AI)
| Treatment | Overall P/AI (%) | P-value | |
|---|---|---|---|
| First service | Later services | ||
| −E2 | 50.2 (114/228) | 34.4 (52/151) | 0.014 |
| +E2 | 49.3 (111/225) | 32.7 (55/168) | 0.006 |
| P-value | 0.998 | 0.989 | |
The overall distribution of ovulatory follicle size for cows at first or later services is shown in Figure 3. For comparisons of P/AI, ovulatory follicles were divided into 3 classes (Table 8). The middle class (medium) was defined as the mean±1 standard deviation. Therefore, the follicles included in the medium follicle class ranged from 14 to 18mm in diameter (n=342; 69.4% of all single ovulating cows). Small follicles were ≤13mm (n=106; 21.5%), and large follicles were ≥19mm (n=45; 9.1%) in diameter. We found no effect of follicle class on P/AI (P=0.92) and no interaction between follicle class and treatment with E2 (interaction P=0.14). However, there was a strong effect of service number (P<0.001) on P/AI and an interaction between service number and follicle class (P=0.035). All of these subgroups (first or later services; small, medium, large ovulatory follicle) were analyzed for effect of E2 treatment on P/AI and there were no significant effects of E2 treatment in any of the subgroups (Table 8).
Figure 3.
Distribution of ovulatory follicle size for cows at first or later services. Only single ovulating cows are included in the figure.
Table 8. Effects of supplementation with 0.5mg of estradiol-17β (+E2) at the final GnRH of Ovsynch on pregnancies per AI (P/AI) in cows during their first or later services within 3 different follicle classes based on ovulatory follicle size in single ovulating cows
| Treatment | Follicle class | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| ≤13mm | 14 to 18mm | ≥19mm | |||||||
| First service | Later services | P-value | First service | Later services | P-value | First service | Later services | P-value | |
| −E2 | 67.7 (21/31) | 29.4 (5/17) | 0.072 | 52.3 (57/109) | 39.7 (25/63) | 0.385 | 66.7 (8/12) | 50.0 (4/8) | 0.881 |
| +E2 | 55.3 (21/38) | 30.0 (6/20) | 0.278 | 49.5 (46/93) | 40.3 (31/77) | 0.633 | 66.7 (8/12) | 7.7 (1/13) | 0.056 |
| P-value | 0.719 | 1.000 | 0.978 | 0.999 | 1.000 | 0.215 | |||
Discussion
The overall fertility exceeded 40% (43%; 332/772) in this study. This was primarily because of the exceptionally high P/AI at the first service, approaching 50%, throughout the study (49.7%; 225/453). The overall P/AI at later services was only 33.5% (107/319) but these represented only 41.3% (319/772) of the services in this study because of the high first-service P/AI and detection of estrus in the herds after first service. The high first-service fertility may have been because of the use of the double-Ovsynch protocol for presynchronization; however, that was not tested in this study. Double-Ovsynch was used to ensure that anovular cows were cycling before Ovsynch and to tightly synchronize the age of CL at second PGF2α and the size of the ovulatory follicle. About 70% of single-ovulating cows ovulated medium-sized follicles (14–18mm) with only about 20% of cows ovulating small follicles and less than 10% ovulating very large follicles. This seems to be a more consistent follicular size than in a previous study from our laboratory (Souza et al., 2007) in which less than 50% of cows ovulated medium-sized follicles (48.7%; 303/622) and a greater percentage of cows ovulated large follicles (22.5%; 140/622; ≥20mm). This previous study utilized a 2 PGF2α presynchronization protocol (14 d between the 2 PGF2α injections with 11 d from last PGF2α to start of Ovsynch). Thus, the reproductive management strategy used in this study seemed particularly appropriate for testing our hypotheses on the effects of reducing circulating P4 and increasing E2 concentrations during the Ovsynch protocol.
Effect of Additional PGF2α
Our first hypothesis was that treatment with an additional PGF2α 1 d after the normal PGF2α treatment of Ovsynch would reduce circulating P4 near AI, resulting in an increase in fertility, particularly in cows that did not have complete regression of the CL in response to the first PGF2α treatment (elevated P4 at BS1). This hypothesis was based on the knowledge that the CL does not completely regress in a percentage of cows receiving Ovsynch and P/AI is reduced for those cows (Moreira et al., 2000, 2001; Souza et al., 2007). Similar to our previous results (Souza et al., 2007), an elevated P4 concentration near the time of the second GnRH treatment dramatically reduced P/AI to less than 50% of values found in cows with low P4 (Figure 2). Treatment with PGF2α 1 d after the normal PGF2α of Ovsynch increased the percentage of cows with low P4 at the time of final GnRH of Ovsynch from 84.6% without the second PGF2α to 95.6% in cows receiving the second PGF2α. Our finding of 15.4% of cows with incomplete regression of the CL is higher than in some previous studies (Pursley et al., 1997b; Gümen et al., 2003; Souza et al., 2007), similar to others (Cartmill et al., 2001; El-Zarkouny et al., 2004), and lower than in one study (Moreira et al., 2000). The largest previous study reporting luteal regression during Ovsynch found 94% of cows (867/919) with low P4 at 48h after the first PGF2α of Ovsynch (Souza et al., 2007). The lower incidence of complete luteal regression in our study may have been because of an elevated percentage of cows with a d 7 CL at the time of the PGF2α treatment of Ovsynch after use of double-Ovsynch for presynchronization or a set-up GnRH before the Resynch program. Previous studies have indicated that a d 7 CL may not respond as well to PGF2α as a more mature CL (Momont and Seguin, 1984). Treatment with the second PGF2α was clearly effective in regressing the CL, based on the large decrease in percentage of cows with high P4 at the second GnRH of Ovsynch (15.4 vs. 4.4% in cows without or with the second PGF2α treatment) or based on the decrease in P4 concentrations between BS1 and BS2 (Table 4) in cows with medium to high P4 at BS1. Thus, a second PGF2α treatment during Ovsynch is effective in increasing luteal regression during the protocol.
Despite the improved luteal regression, there was no overall improvement in P/AI in response to the second PGF2α. In retrospect, this is understandable based on a simple calculation of expected results. The second PGF2α increased the percentage of cows with low P4 from 84.6 to 95.6%, a difference of 11 percentage units. Because overall P/AI was 43% in this study, it seems likely that the greatest potential increase in P/AI because of the second PGF2α would be 4.7% (43%×0.11). Our observed improvement of 3.2% is not far from this value. Clearly, a statistically detectable improvement in P/AI from the second PGF2α will require a study with 1,386 (4.7% difference) to 2,983 (3.2% difference) cows per treatment (Win Episcope 2.0 calculations with one-tailed test) because of the relatively small magnitude of the treatment effect.
We anticipated that we would be able to select the cows that would show improvement after the second PGF2α treatment by evaluating the P4 concentrations in cows at 1 d after the first PGF2α, just before we gave the second PGF2α treatment. Surprisingly, the second PGF2α treatment did not increase P/AI in cows with high P4 at BS1, despite a major increase in the percentage of cows with low P4 at second GnRH (63.2 vs. 91.0%; Table 5). The same was true for cows with medium and low concentrations of P4 at BS1. However, the largest numerical increase in P/AI in response to the second PGF2α was for cows with a medium concentration of P4 at the time of the second PGF2α treatment. This unexpected observation may indicate that timing in the decrease of P4 during Ovsynch is critical. For example, cows with medium but not high P4 may have an earlier decrease in P4 that might be important for optimal fertility. Cows with high P4 on 1 d after the normal PGF2α treatment continued to have low fertility, regardless of whether they received the second PGF2α treatment and subsequently had complete CL regression. Perhaps administration of the extra PGF2α earlier (∼12h following the normal PGF2α) would have reduced circulating P4 sooner and had a more positive impact on P/AI.
Effect of E2 Supplementation
Our second hypothesis was that supplementation with 0.5mg of E2 at the time of the second GnRH injection would increase expression of estrus in all cows and improve P/AI in cows ovulating medium-sized follicles and cows at first service. This hypothesis was based on previous research (Souza et al., 2007) indicating that supplementation with 1mg of E2 8h before the final GnRH dramatically increased expression of estrus and increased fertility in cows at first service (but not at later services), primiparous cows (but not multiparous), and cows ovulating medium-sized follicles (but not those ovulating smaller and larger follicles).
Clearly, supplementation with E2 increased expression of estrus from 37.2% in cows not receiving E2 to 84.4% in E2-treated cows. This increase occurred for all subgroups of cows analyzed in the study (primiparous and multiparous; first or later services; smaller, medium, or larger ovulatory follicles). Expression of estrus in our control group was somewhat greater than the ∼20% reported in previous studies that gave the second GnRH injection at 48h after the PGF2α treatment (Jobst et al., 2000; Pancarci et al., 2002; Perry et al., 2005) but similar to the 44.4% reported when the final GnRH treatment was delayed until 56h (Souza et al., 2007). Our high rates of expression of estrus following E2 treatment were similar to the rates reported previously for cows supplemented with 1mg of E2 (Souza et al., 2007) or when ECP at 24h post-PGF2α was substituted for the final GnRH of Ovsynch (Pancarci et al., 2002; Cerri et al., 2004; Galvão et al., 2004).
In contrast to the results of Souza et al. (2007), we found no effect of E2 supplementation on fertility. There was no effect of E2 on overall P/AI. When we considered only cows that were synchronized to the protocol (ovulated and low P4), similar to what was done by Souza et al. (2007), there was a tendency for a negative effect of E2 supplementation on P/AI. In addition, there was no indication of positive E2 effects in first-service cows, as was observed by Souza et al. (2007). Finally, our experimental protocol attempted to more tightly synchronize the size of the ovulatory follicle at the time of the second GnRH (using double-Ovsynch and a preliminary GnRH during Resynch) and measurements were made of the size of the ovulatory follicle to increase our chances of observing a potential effect of E2 on fertility. Nevertheless, there was no indication of any effect of E2 supplementation on P/AI in cows ovulating small, medium, or large follicles. Thus, it is clear that E2 supplementation in the reproductive management protocol used in this experiment was without a positive effect on fertility.
One of the major differences between our present study and our previous study (Souza et al., 2007) was the timing of the E2 treatment in relation to the final GnRH. In the previous study, E2 was administered 8h before the final GnRH of Ovsynch (48h after PGF2α). This would allow a dramatic increase in circulating E2, peaking at 4h before the second GnRH, and beginning to decline by GnRH treatment, with no difference in circulating E2 at 4h after the second GnRH (Souza et al., 2005). In contrast, the protocol used in our present experiment would be expected to attain peak E2 concentrations at 4h after the second GnRH treatment and would not return to baseline levels until 12h after GnRH. This shift in timing of the E2 peak could produce very different effects of E2 on the reproductive tract and fertility. Although E2 supplementation at the time of the final GnRH caused estrus in most cows, there was no improvement in P/AI in cows that showed estrus versus cows that did not show estrus during this protocol. This result sharply contrasts with the results of Souza et al. (2007). It also contrasts with the results obtained when the Heatsynch protocol was used. Cows that expressed estrus after receiving ECP had much greater P/AI than cows not expressing estrus (Lopes et al., 2000; Cerri et al., 2004; Kasimanickam et al., 2005). The only other result in which estrogen was given at the time of the second GnRH (0.25mg of ECP) also had numerically lower P/AI in cows treated with estrogen than in control cows (Sellars et al., 2006). Other physiological explanations for these results are also possible such as insufficient dose of E2, presynchronization/resynchronization protocol reducing E2 effect, human chorionic gonadotropin post-AI reducing E2 effect, E2 suppressing the ability of GnRH to induce a normal LH surge, or even no role for supplementary E2 during the Ovsynch protocol. However, suboptimal timing of the E2 may be the most plausible explanation to reconcile previous and present results.
In conclusion, this study has provided novel information on the effects of adding a second PGF2α or supplementing with E2 during the Ovsynch protocol. It is encouraging that adding a second treatment with PGF2α can produce increased efficiency of luteolysis. However, no detectable improvements in fertility accompanied this improvement in percentage of cows with complete luteolysis. Our results were not encouraging for E2 supplementation, with no improvement in fertility when E2 was given at the same time as the second GnRH treatment of Ovsynch. In any future experiments with these treatments it seems warranted that the timing of PGF2α and E2 treatment be carefully considered.
Acknowledgments
The authors thank Mystic Valley Dairy (Sauk City, WI), Wagner Dairy (Middleton, WI), and their staff for all their help and the use of their herds to conduct these on-farm trials. IVX Animal Health Inc. (St. Joseph, MO) donated hormones used for this research.
Supplementary data
Interpretive summary.
References
- . Stage of cycle, incidence, and timing of ovulation, and pregnancy rates in dairy cattle after three timed breeding protocols. J. Dairy Sci. 2001;84:1051–1059
- . Timed artificial insemination with estradiol cypionate or insemination at estrus in high-producing dairy cows. J. Dairy Sci. 2004;87:3704–3715
- . Pregnancy in dairy cows after synchronized ovulation regimens with or without presynchronization and progesterone. J. Dairy Sci. 2004;87:1024–1037
- . Effect of addition of a progesterone intravaginal insert to a timed insemination protocol using estradiol cypionate on ovulation rate, pregnancy rate, and late embryonic loss in lactating dairy cows. J. Anim. Sci. 2004;82:3508–3517
- . Follicular size and response to Ovsynch versus detection of estrus in anovular and ovular lactating dairy cows. J. Dairy Sci. 2003;86:3184–3194
- . Improvement of sperm transport by the administration of estradiol to estrous in ewes. J. Anim. Sci. 1975;41:1400–1406
- . Evaluation of reproductive performance in lactating dairy cows with prostaglandin F2alpha, gonadotropin-releasing hormone, and timed artificial insemination. J. Dairy Sci. 2000;83:2366–2372
- . Fertility following fixed-time AI or insemination at observed estrus in Ovsynch and Heatsynch programs in lactating dairy cows. Theriogenology. 2005;63:2550–2559
- . Use of estradiol cypionate for timed insemination. J. Dairy Sci. 2000;83(Suppl. 1):910;(Abstr.)
- . Relationship between level of milk production and estrous behavior of lactating dairy cows. Anim. Reprod. Sci. 2004;81:209–223
- . Influence of day of estrous cycle on response to PGF2α products: Implications for AI programs for dairy cattle. In: Proc. 10th Int. Congr. Anim. Reprod. Artificial Insemination. University of Illinois at Urbana-Champaign, IL. Univ. Illinois Urbana-Champaign, IL. 1984;p. 336.1–336.3
- . Effects of presynchronization and bovine somatotropin on pregnancy rates to a timed artificial insemination protocol in lactating dairy cows. J. Dairy Sci. 2001;84:1646–1659
- . Use of bovine somatotropin in lactating diary cows receiving timed artificial insemination. J. Dairy Sci. 2000;83:1237–1247
- . Interactions of high milk yield and reproductive performance in dairy cows. J. Dairy Sci. 1993;76:3257–3268
- . Use of estradiol cypionate in a presynchronized timed artificial insemination program for lactating dairy cattle. J. Dairy Sci. 2002;85:122–131
- . Comparison of three estrus detection systems during summer in a large commercial dairy herd. Anim. Reprod. Sci. 2005;87:59–72
- . Relationship between follicle size at insemination and pregnancy success. Proc. Natl. Acad. Sci. USA. 2005;102:5268–5273
- . Reproductive management of lactating dairy cows using synchronization of ovulation. J. Dairy Sci. 1997;80:301–306
- . Synchronization of ovulation in dairy cows using PGF2α and GnRH. Theriogenology. 1995;44:915–923
- . Pregnancy rates per artificial insemination for cows and heifers inseminated at a synchronized ovulation or synchronized estrus. J. Dairy Sci. 1997;80:295–300
- . Comparing early embryo mortality in dairy cows during hot and cool seasons of the year. Theriogenology. 1993;39:719–737
- . High feed intake increases liver blood flow and metabolism of progesterone and estradiol-17β in dairy cattle. J. Dairy Sci. 2002;85:2831–2842
- . Effect of human chorionic gonadotropin on luteal function and reproductive performance of high-producing lactating Holstein dairy cows. J. Anim. Sci. 2001;79:2881–2894
- . Fertilization and early embryonic development in heifers and lactating cows in summer and lactating and dry cows in winter. J. Dairy Sci. 2002;85:2803–2812
- . Ovarian structures and circulating steroids in heifers and lactating cows in summer and lactating cows and dry cows in winter. J. Dairy Sci. 2002;85:2813–2822
- . Time and incidence of ovulation and conception rates after incorporating estradiol cypionate into a timed artificial insemination protocol. J. Dairy Sci. 2006;89:620–626
- . The estrus detection problem: New concepts, technologies, and possibilities. J. Dairy Sci. 1994;77:2745–2753
- . A new presynchronization system (Double-Ovsynch) increases fertility at first postpartum timed AI in lactating dairy cows. Theriogenology. 2008;70:208–215
- . Profiles of circulating estradiol-17β after different estrogen treatments in lactating dairy cows. Anim. Reprod. 2005;2:224–232
- . Supplementation with estradiol-17β before the last GnRH of the Ovsynch protocol in lactating dairy cows. J. Dairy Sci. 2007;90:4623–4634
- . Trends in reproductive performance in southeastern Holstein and Jersey DHI herds. J. Dairy Sci. 2002;85:244–251
- . Changes in reproductive physiology of lactating dairy cows due to elevated steroid metabolism. Theriogenology. 2006;65:17–29
PII: S0022-0302(09)70452-7
doi:10.3168/jds.2008-1289
© 2009 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.
