Effects of delaying the removal of the intravaginal progesterone device by 24 h in the 5d Co-synch protocol in heifers

In beef herds, increasing animal welfare, improving reproductive performance and easing animal management are key goals in farm economics. This study explored whether delaying the removal of the intravaginal progesterone device by 24 h in heifers synchronized with a 5d Co-synch 72 h protocol could improve reproductive eciency of xed-time articial insemination (FTAI). In Experiment 1, we examined whether such a modied protocol would provide an acceptable, total synchronization rate (TSR) in cycling Holstein heifers. Heifers (13.4 ± 0.69 mo.) were randomly assigned to the standard 5d Co-synch 56 h protocol (5dCo56; n = 10), the standard 5d Co-synch 72 h protocol (5dCo72; n = 17), or the modied 5d Co-synch 72 h protocol, in which removal of the progesterone device was delayed by 24 h (6dCo48; n = 19). Heifers were considered synchronized (TSR) if serum progesterone value > 1.0 ng/mL and the corpus luteus (CL) was detectable by ultrasonography on Day 5, if progesterone < 1.0 ng/mL and at least one follicle had a diameter > 8 mm on Day 7.5, and if progesterone > 1.0 ng/mL and CL was detectable on Day 15. In Experiment 2, 309 cycling beef heifers on 18 commercial farms were subjected to the 5d Co-synch 72 h or 6d Co-synch 48 h protocol and conception rate (CR) studied. These results suggest that delaying removal of the intravaginal progesterone device by 24 h during the standard 5d Co-synch 72 h protocol provides satisfactory results without reducing reproductive eciency of heifers.


Results
In experiment 1, the three protocols in dairy heifers led no differences on TSRs of 80.0% (5dCo56), 88.2% (5dCo72) and 89.5% (6dCo48). In experiment 2, the CR from the beef heifers, observed during two consecutive reproductive seasons did not differ: 59.7% for 5dCo72 and 62.0% for 6dCo48 (P = 0.907). Heifer age, body condition score or stress score did not signi cantly affect conception rate.

Conclusions
These results suggest that delaying removal of the intravaginal progesterone device by 24 h during the standard 5d Co-synch 72 h protocol provides satisfactory results without reducing reproductive e ciency of heifers.
In GnRH-based protocols, the initial GnRH is intended to induce LH release and ovulation, with emergence of a new follicular wave approximately 2 d later. The 5d Co-synch protocol, in which FTAI is performed 72 h after the rst prostaglandin administration, shortens the follicular dominance by two days and lengthens the proestrus phase by 16-24 hours, ultimately leading to higher mean conception rate (CR) [8,[11][12][13][14].
In contrast, early work [15] suggested that the 5d Co-synch 56 h (5dCo56) protocol was more appropriate for beef heifers, probably because following P4 device removal, heifers tend to display estrus earlier than cows. However, another study [16] reported higher conception rate in dairy heifers if they received the nal GnRH administration concurrent with AI at 72 h after PGF than if they received it at 16 h before AI. Still another study found no fertility difference between beef heifers who received FTAI at 66 or 72 h after GnRH [17].
A shortened proestrus phase should not affect heifer synchronization or CR, as long as prostaglandin administration is maintained on Days 5 and 6 of the protocol. In fact, the serum concentration of progesterone induced by an intravaginal device is signi cantly lower than that induced by a natural CL [18]. Concentration of progesterone (P4) will remain low for 24 h until the second prostaglandin administration in animals that do not undergo a luteolytic response to prostaglandin on Day 5. A preliminary study [19] found no difference in CR in dairy heifers with a conventional 5 day Co-synch 72 h protocol or a modi ed protocol (Day 0: GnRH + IPD; Day 5: PGF; Day 6: PGF + IPD removal; Day 8: GnRH + FTAI). They described that delaying IPD removal from dairy heifers by 24 h during the 5d Co-synch protocol avoided the demonstration of estrus earlier than 16 h prior to FTAI. However, that study did not examine whether the modi ed protocol altered the ovarian dynamics compared to the 72 h protocol, nor did it compare total synchronization rates (TSR). Thus, the optimal window when ovulation should be induced with GnRH relative to FTAI in heifers subjected to the 5d Co-synch protocol warrants further research. Indeed, modi cation of existing Co-synch protocols may increase the versatility of FTAI programs on beef farms. This modi ed protocol, previously proposed [19], henceforth referred to as "6dCo48", comprises a 5d GnRH-based Co-synch protocol without presynchronization but with GnRH on Day 0 and with prostaglandin administration on Days 5 and 6. In the 6dCo48 protocol, removal of the IPD is delayed by 24 h relative to the conventional 5dCo72 protocol. We compared these two protocols in terms of ovarian synchronization and conception rates, with the aim at testing its value in the eld, in beef heifers. Therefore, the objective of Experiment 1 was to examine whether the 6dCo48 protocol would provide an adequate TSR compared to 5dCo72 and 5dCo56 protocols in dairy heifers, based on analysis of follicular growth, luteolysis and ovulation. Dairy heifers were chosen (easier to obtain an adequate sample of animals homogeneous in age, size, weight and ovarian stage, under a controlled environment) to check that this modi cation did not signi cantly worsen the ovarian response. The objective of Experiment 2 was to compare CR of beef heifers in commercial beef herds after FTAI using the 6dCo48 or 5dCo72 protocols (the two protocols with the best synchronization rate in Experiment 1). Our hypothesis was that delaying by 24 h the removal of the intravaginal progesterone device within a 5dCo72 protocol in heifers would not negatively affect conception rates after FTAI and might improve the synchronization rate of heifers, making it an additional synchronization protocol available to beef farmers.

Methods
Experiment 1: synchronization study Experiment 1 was conducted in a small sample of Holstein heifers to ensure that the new protocol (6dCo48) would not induce a worse synchronization rate than conventional protocols. This phase of the study was performed in Holstein heifers because it was easier to obtain an adequate sample of animals homogeneous in age, size, weight and ovarian stage that could be maintained in the same controlled environment within a proper handling facility. Moreover, dairy heifers show a comparable ovarian physiology to beef heifers [20,21].
Holstein heifers on a single dairy farm in Toledo, in central Spain, were housed in free-stall barns and fed a total mixed ration adapted to their daily requirements [22], with ad libitum access to feed and water. A total of 46 dairy heifers were included, 29 during a rst replicate (March 2017) and 17 during a second replicate (April 2017). To be included in the study, animals had to be in adequate health; show a body condition score (de ned below) ≥ 2 and < 3.5; lack a history of arti cial insemination; and lack anoestrus. Animals needed to present P4 concentrations higher than 1 ng/mL in one or both samples taken 7 days apart previous day 0 of the study. In addition, initial ovarian cyclicity had to be observed by transrectal ultrasonography on Days − 7 and 0 (see Sect. 2.1.1). Of the 51 heifers screened for inclusion, ve were excluded because they showed no CL on at least one day, and subsequent P4 assay showed concentrations below 1 ng/mL. On Day 0 of the protocols, body condition score (BCS) on a scale from 1 (cachectic) to 5 (extremely obese) [23] and stress score (SS) were determined. SS was measured on a scale from 1 to 5 [24] when animals were in the chute, where a score of 1 meant that the animal was calm, without unexpected, sudden movements; 2, the animal was slightly restless; 3, the animal was squirming and occasionally shaking; 4, the animal moved continuously with very vigorous movements that shook the chute; and 5, the animal was rearing, twisting the body and struggling violently [24].
In the rst replicate (March 2017), a total of 29 dairy heifers were assessed for age, BCS and SS (Table 1) and randomly assigned to undergo the 5dCo72 protocol (n = 9), 5dCo56 protocol (n = 10) or 6dCo48 protocol (n = 10). All protocols involved the insertion of an intravaginal progesterone device. In the second replicate (April 2017), dairy heifers (Table 1) were randomly assigned to undergo our 6dCo48 protocol (n = 9) or 5dCo72 protocol (n = 8), since the 5dCo72 protocol gave as good TSR as the 5dCo56 protocol in the rst replicate (see Results).  The protocol 5dCo72 was performed as described [8]. Brie y, it consisted of intramuscular administration of 100 µg GnRH . This dosage has been demonstrated to be safe and not associated with increased rate of multiple ovulation [25]. On Day six, 25 mg prostaglandin was administered intramuscularly. On Day eight, 72 h after removal of the intravaginal progesterone device, FTAI was performed and 100 µg GnRH was administered intramuscularly. The 5dCo56 protocol was identical to 5dCo72, except that FTAI and GnRH treatment were performed at 56 h after removal of the progesterone device. The 6dCo48 protocol was identical to 5dCo72, except that the progesterone device was removed on Day six, simultaneously with the second prostaglandin administration (Fig. 1).
Unfortunately, CR after FTAI could not be analyzed because heifers received conventional or sex-sorted semen from different bulls based on farmer assessments. This may have confounded analyses of CR. Nevertheless, we considered the TSR results to be su ciently reliable to proceed to a eld test of the modi ed protocol in beef heifers in Experiment 2 of the study.

Ultrasonography and blood analyses
Transrectal ultrasonography (US) was performed by the same experienced veterinarian on every heifer using a SIUI CTS 800® immediately centrifuged at 4,500 g for 15 min, and the plasma was transferred to a fresh tube and stored at -80 ˚C until progesterone determination. Plasma progesterone concentrations were measured in a single analysis using an enzyme immunoassay kit (Demeditec Diagnostics, Kiel-Wellsee, Germany) as described [28]. Assay sensitivity was 0.045 ng/mL and the manufacturer-speci ed intra-assay variation coe cient was 5%.  Abbreviations: SD = Standard Deviation; Age: age at study inclusion (Day 0); BCS = Body Condition Score at Day 0, scale 1-5; SS = Stress Score (1-5) at Day 0; P4 = plasma progesterone concentration on Day 0; 5dCo72 = 5 day Co-synch 72 h protocol with intravaginal progesterone device (IPD); 6dCo48 = 6 day Co-synch protocol 48 h with IPD. *P-value after Kruskal-Wallis analysis.

Statistical analyses
All data were analyzed using SPSS® 25 (IBM, Armonk, NY, USA). Probability values less than or equal to 0.05 were considered signi cant and those between 0.05 and 0.10 were considered trends. All data were reported as mean (percentage) or as mean ± SD.
Inter-group differences were assessed for signi cance using the chi-squared and Student's t test when data were normally distributed or using non-parametric analyses (Kruskal-Wallis) when data were skewed. Results from both phases of the study were analyzed using logistic regression that included farm as a xed factor and several possible confounding factors in a stepwise forward method based on the Wald statistic criterion P > 0. 10. In Experiment 2 of the study, data were separately analyzed in two regression models, one with data on all 309 animals, and the other with data on 122 animals with the P4 information.

Experiment 1
Replicates were clustered by introducing the factor "replicate" in the regression model. Percentages of heifers synchronized at the different time points in the protocol and totally synchronized are summarized in the table included in the Additional le 1. TSR did not vary signi cantly with any of the categorical factors included in the study, including synchronization protocol.
The age on D0 was not different between totally synchronized heifers (13.5 ± 0.89 months) and other heifers (13.6 ± 0.65 mo.; P = 0.699). Similarly, the size of the largest follicle on Day 7.5 was not different between totally synchronized heifers (12.8 ± 2.29 mm) and others (12.8 ± 1.16 mm; P = 0.968). The P4 values at different time points in the protocol did not in uence the probability of synchronization on the nearest subsequent measurement date, nor the probability of being totally synchronized at the end of the protocol. The exception was P4 on Day 7.5: each 1-ng/mL P4 increase on Day 7.5 was associated with signi cantly lower probability of total synchronization at the end of the protocol (OR 0.02, 95% CI 0.001-0.416; P = 0.012).  Table 3. Farm and semen did not signi cantly in uence conception rate (P = 0.907 and 0.329, respectively) in the regression model. Similarly, season (P = 0.924), interaction between age and protocol (P = 0.746) and interaction between AI technician and protocol (P = 0.706) did not signi cantly affect conception rate.

Discussion
In this study, the 6dCo48 protocol, a modi ed version of the conventional 5dCo72 protocol, achieved an ovarian synchronization rate of 89.5% and CR of 62.0%, which were not different from those obtained with the conventional protocol. Thus, the 6dCo48 would be an alternative protocol to optimize the reproductive management of beef heifers subjected to FTAI, enhancing exibility when considering implementing FTAI strategies in beef cattle farms.
If farmers decide to implement different protocols for cows (5dCo72h) and heifers (5dCo56h), arti cial insemination of a single herd must occur in two steps, once in heifers and 16 h later in cows. With the protocol described here, in contrast, beef heifers and multiparous cows can be synchronized and inseminated at the same time. Moreover, as hypothesized, our modi cation of the conventional 5d Co-synch protocols did not worsen reproductive outcomes in heifers, probably because we maintained prostaglandin administration on Day 5. This mitigated the effect of proestrus shortening on reproductive e ciency. Prostaglandin administration on Days 5 and 6 promotes corpus luteum (CL) regression [30] and increases CR in dairy and beef heifers [31,32].
While some studies question the usefulness of the second prostaglandin dose in heifers [17,33,34], we maintained it in our protocol because it is indispensable for FTAI programs in multiparous cows [35,36], and we aimed to develop an additional protocol that would work with a similar handling for beef cows and heifers.
Nevertheless, we cannot exclude that the modi ed protocol may be associated with higher risk that a heifer has an old oocyte when inseminated [8]. Some protocols administer GnRH on Day 0 to enhance subsequent response to GnRH on Day 8 in beef heifers [37] and dairy heifers [30], ultimately increasing CR [38]. Although some studies suggest that GnRH on Day 0 provides no reproductive bene ts in dairy or beef heifers [5,33] and may even induce pregnancy loss in acyclic heifers [17], we decided to keep the GnRH on Day 0 in our protocol, based on the inclusion of cyclic heifers. Indeed, an FTAI protocol in which cows received prostaglandin 3 days before the protocol began and then were arti cially inseminated at 66 h after IPD removal led to a higher percentage of cows showing a new follicular wave after the initial GnRH as well as higher conception rate than the classical 5d Co-synch 72 h protocol [39]. Such pre-synchronization has been associated with higher CR in dairy heifers [40], higher fertility rate after FTAI in multiparous beef cattle [39], and greater ovarian synchronization in beef heifers [41]. In beef heifers, however, another study did not observe different CRs between the 5d Co-synch 72 h protocol and a prostaglandin 6d-IPD protocol, consisting of prostaglandin (25 mg) on Day − 9, GnRH (100 µg) and IPD insertion on Day − 6, PG and IPD removal on Day 0, and FTAI at 66 h after IPD removal [42]. Unlike the prostaglandin 6d-IPD protocol, the modi ed protocol in the present study did not include pre-synchronization with prostaglandin.
In Experiment 1, mean age of the 46 dairy heifers in the rst part of the study was 13.5 ± 0.86 months, and did not differ between synchronized and other animals. Similarly, the age of beef heifers in the second part of this study did not in uence the probability of pregnancy after FTAI. For adequately developed beef heifers, reaching puberty before FTAI may be even more important for fertility than age [42]. Age may not have in uenced conception rate in our study because the heifers (21.7 ± 3.54 months old) had reached puberty and had adequate body condition. Since these two factors are the main determinants of fertility in heifers [43,44], they may also explain the lack of differences on TSR between our experimental groups. We also cannot exclude that the lack of a signi cant difference in TSR is an artifact of the relatively small samples in our study.
Body condition score did not affect the TSR in dairy heifers, which is not surprising given that scores were in the physiological range between 2.5 and 3.5. In contrast, cachectic or extremely obese body condition reduces reproductive performance in cattle [45][46][47].
As in dairy heifers, BCS in our beef heifers also did not affect CR, independently of synchronization protocol. The best reproductive results can be achieved in beef herds [48] with an appropriate nutrition plan [49][50][51] that allows them to achieve a balanced BCS of 2.5-3 at the beginning of the reproductive season. This balanced BCS in heifers is associated with appropriate hormonal regulation [48,51].
This may help explain why SS did not in uence TSR. SS was similarly low (around 2) in pregnant and non-pregnant beef heifers, which may re ect that beef farms in our study had already implemented stress-reducing measures at the chute. Studies have suggested that SS < 3 indicates good animal temperament [55][56][57][58]. The uniformly low SS in our animals may help explain why conception rate did not differ signi cantly with beef breed or with the interaction between breed and FTAI protocol. Our results suggest that although CR can depend on temperament in certain breeds [55,56,58], animal temperament in our study was modulated mainly by management.
We found no signi cant difference in synchronization rates between animals strati ed by P4 concentrations on Day 0, using a strati cation approach like previous work [26]. That previous work suggested that initial P4 in dairy cows undergoing FTAI should range between 0.5 and 6 ng/mL at the rst GnRH administration [26]. When we categorized our animals by initial P4 concentration, we broadened the de nition of "intermediate" concentrations to consider that heifers show higher P4 concentrations than dairy cows [59]. We did nd a tendency toward higher TSR in dairy heifers with intermediate P4 concentrations on Day 0, but this may not be real, since only one heifer in Experiment 1 was in the estrus stage of its cycle on Day 0, and more than 50% of animals showed initial P4 values > 7 ng/mL. Our failure to observe a robust dependence of TSR or CR on initial P4 concentration may re ect differences in ovarian physiology between dairy cattle and heifers [60], suggesting that beef heifers can show a strong ovarian response to FTAI protocols independently of initial P4. At the same time, we cannot exclude that the lack of signi cant variation in TSR or CR re ects the small samples in our study.
Nevertheless, we did nd that P4 concentration on Day 7.5, when luteolysis should happen, inversely and signi cantly affected the rate of synchronization on that day (OR 0.02, 95% CI 0.001-0.416). This may re ect that progesterone inhibits luteinizing hormone release through a negative feedback loop, disrupting ovulation at the end of the synchronization protocol [61].
We found no difference on follicle size around 12.8 ± 2.29 mm before AI across all protocols, also, without differences to the 13.4 ± 0.3 mm reported for beef heifers under the 7d and 5d Co-synch 72 h protocols [62] and the 11.0 ± 0.5 mm reported for dairy heifers under the 5d Co-synch 72 h protocol with PGF 2α presynchronization at 2 days prior to IPD insertion [40]. Therefore, delaying removal of the intravaginal progesterone device did not reduce the TSR.

Page 9/15
In conclusion, our modi cation of the 5d Co-synch protocol for heifers, in which removal of the intravaginal progesterone device is delayed by 24 h, achieves reproductive results as good as those obtained with 5d Co-synch protocols that stagger insemination of heifers and cows. Therefore, the 6dCo48 protocol seems to be a suitable protocol for beef heifers on commercial farms, which can provide another tool to expand FTAI possibilities. Availability of data and materials

Abbreviations
The datasets during and/or analysed during the current study available from the corresponding author on reasonable request.
The authors declare that they have no competing interests Funding CEVA Santé Animale partially funded hormonal treatments.
Authors' contributions AFN designed the work, acquired eld data, analyzed and interpreted all the data and major contribution in drafting the manuscript; SSL: acquired and analyzed eld data, revised manuscript; JLPP: acquired and analyzed laboratory data and revised manuscript; NPV: analyzed data, drafted the manuscript; AHM: acquired and analyzed laboratory data and revised manuscript; JVGM: designed the work, interpreted all the data and revised manuscript; SA: designed the work, acquired, analyzed and interpreted all data, major contribution in reviewing the manuscript. All authors read and approved the nal manuscript.