Fifty thoroughbred mares from 6 to 9 years of age, weighing 430 to 470 kg, were used between March and August 2009. Before the experiment, mares underwent clinical evaluation by transrectal palpation and ultrasonographic examination of the genital tract.

Unless otherwise stated, all chemicals were purchased from Sigma-Aldrich (USA). The base medium for IVM was tissue culture medium (TCM)-199 (Invitrogen, USA) supplemented with 0.6 mM cysteine, 1.0 mM pyruvate, 10 mU/mL FSH (Folltropin-V; Bioniche Animal Health, Canada), 50 ng/mL EGF, 1 µg/mL insulin-transferrin-sodium, 25 µg/mL gentamycin and 10% (v/v) fetal bovine serum (FBS) from the same batch. In vitro culture (IVC) medium was a 1:1 mixture of Dulbecco's modified Eagle's medium (D-MEM) (Invitrogen) and nutrient mixture F-12 (D-MEM/F-12) (Invitrogen) supplemented with 10% FBS.

A real-time ultrasound scanner (Mylab30Vet; Esaote, Italy) equipped with a 7.5 MHz convex array transducer (model EC123) housed in a hard plastic vaginal device with stainless steel needle guidance was used. Two types of needles were compared, a 12-G double lumen needle (V-EOAD-1260L; Cook Medical, Australia) and a 14-G double lumen needle system using a short disposable needle (Bovi-vet; Kruuse, Denmark). A vacuum pump was attached to the needle, and the aspiration pressure was adjusted to -150 mmHg for the 12-G needle and -200 mmHg for the 14-G needle. Before follicle aspiration, mares were restrained in stocks and sedated with 0.6 mg/kg xylazine intravenously (iv), 0.03 mg/kg acepromazine iv and 0.01 mg/kg butorphanol tartrate iv, as well as 0.1 mg/kg propantheline bromide iv for rectal relaxation. The ultrasound transducer was inserted into the fornix of the vagina, and the ovary was attracted transrectally to lie against the vaginal wall near the transducer. The follicles were aspirated by inserting the needle into the follicular cavity while viewing through the monitor of the ultrasound scanner, then massaged per rectum and flushed continuously with 150 to 200 mL of commercial flushing solution (Vigro; Bioniche Animal Health) containing 10 units/mL heparin. After the aspiration procedure, the fluid was immediately taken to a laboratory and examined under a stereomicroscope for recovery of cumulus-oocyte complexes (COCs).

Recovered COCs were classified as follows according to the morphology of the cumulus (Fig. 1) [14]: (1) compact (Co, COCs with cumulus or corona cells tightly surrounding the oocyte); (2) expanded (Ex, COCs with well-expanded cumulus or surrounded by a mucous matrix); (3) denuded (De, COCs with only corona radiata or a partial layer of cumulus). After washing in IVM medium, each COC was placed into one well of a four-well multi-dish (Nunc, Denmark) containing 500 µL IVM medium and cultured at 38.5℃ in a humidified atmosphere of 5% CO₂ in air for 13 to 16 h (Ex) or 24 to 27 h (Co).

After transport from the slaughterhouse to the laboratory at 30℃ (within 0.5-1 h), ovaries were cleaned by trimming the adnexa with scissors and washing with sterile saline. All visible follicles were opened with a scalpel and the follicular wall was scraped using a bone curette. Scraped contents were washed into petri dishes with D-MEM/F-12 supplemented with 0.05% (w/v) polyvinyl alcohol and examined under a stereomicroscope. Cumulus-oocyte complexes were classified and held in a mixture of 40% (v/v) TCM-199 with Earle's salts (Invitrogen) and 40% (v/v) TCM-199 with Hanks' salts (Invitrogen) with 20% (v/v) FBS [20] at room temperature for 23 h (holding treatment). After the holding treatment, COCs were washed with fresh IVM medium and cultured at 38.5℃ in a humidified atmosphere of 5% CO₂ in air for 19 to 20 h.

Skin fibroblasts from Warmblood stallions were seeded into a four-well plate and grown in D-MEM/F-12 (Invitrogen) supplemented with 15% (v/v) FBS until a complete monolayer formed. Donor cells were induced to synchronize at the G0/G1 stage of the cell cycle by contact inhibition for 24 to 72 h. Cells of the same passage (3-6 passages) were used in each replicate for the various treatments. A single cell suspension was prepared by trypsinization of cultured cells and resuspension in D-MEM containing 0.4% (w/v) BSA (MEM-BSA) prior to nuclear transfer.

After IVM, cumulus-cell-free oocytes from which cumulus cells had been removed by exposure to 15% (w/v) hyaluronidase and repeated pipetting through a small glass pipette were incubated for 10 min in manipulation medium (MEM-BSA) containing 5 µg/mL cytochalasin B. Following incubation, the oocytes were transferred into a drop of manipulation medium on a glass dish (WillCo-dish; Willco Wells, The Netherlands) and the zonae pellucidae were penetrated with a 12-µm blunt glass pipette (Humagen, USA) using a Piezo drill (PMM-150FU; PRIME TECH, Japan). Oocytes were enucleated by aspirating the first polar body and the adjacent cytoplasm containing the MII chromosomes, which were confirmed by the absence of metaphase spindles under a microscope (Ti-U; Nikon, Japan) equipped with a polarization-based optical instrument (Oosight; Cambridge Research and Instrumentation, USA). Following enucleation, a single donor cell was inserted into the perivitelline space of each oocyte. The oocyte-cell couplets were then placed on a 0.5 mm fusion chamber overlaid with 280 mM mannitol containing 0.05 mM CaCl₂ and 0.1 mM MgCl₂. Membrane fusion was induced by applying two pulses of 2.2 to 2.4 kV/cm direct current for 15 µsec using a cell fusion generator (LF101; Nepa Gene, Japan). After incubation in MEM-BSA for 15 min, non-fused couplets were subjected to subsequent electric pulses. Reconstructed oocytes were activated at 1 to 2 h after fusion by exposure to medium containing 5 µM ionomycin for 4 min, followed by washing and incubation in IVC medium containing 2 mM 6-dimethylaminopurine at 38.5℃ in a humidified atmosphere of 5% CO₂ in air for 5 h.

Following the activation procedure, SCNT embryos were transferred to recipient mares (1-3 embryos/mare) by standing flank laparotomy within 24 h of ovulation [29]. Briefly, after treatment with a combination of 0.6 mg/kg xylazine iv, 0.03 mg/kg acepromazine iv and 0.02 mg/kg butorphanol tartrate iv, approximately 150 mL of lidocaine was instilled into the paralumbar fossa of the mare for local anesthesia (inverted L block). The ovary and oviduct ipsilateral to ovulation were exteriorized, and SCNT embryos were transferred into the mid-ampullary region of the oviduct using a 14 cm catheter (Tom Cat Catheter; Kendall, USA). After initial detection of pregnancy by transrectal ultrasonography (5 MHz transducer) between Days 14 and 16 (Day 0 = surgical transfer), mares were subsequently examined at Days 21, 30, 50 and 60 of gestation. After Day 100, pregnancy examinations were performed by transabdominal ultrasonography (2.5 MHz transducer) at monthly intervals.

Parentage analysis was performed on the foal produced by SCNT to confirm genetic identity with the somatic donor cells. The DNA from blood and cells was extracted by the standard proteinase K/phenol-chloroform procedure [21]. The extracted genomic DNA samples were subjected to microsatellite assay using a 17-plex Applied Biosystems (USA) horse genotyping kit (VHL20, HTG4, AHT4, HMS7, HTG6, HMS6, HTG7, HMS3, AHT5, ASB2, HTG10, HMS2, ASB17, LEX3, HMS1, CA425, and ASB23) labeled with one of the fluorescent dyes, FAM, NED, PET or VIC. Polymerase chain reaction (PCR) products were generated in 15 µL reactions containing 10 ng DNA, 2.5 µL StockMarks PCR buffer, 4 µL dNTP mix, 4 µL of 17 primer sets (one for each of the microsatellite markers) and 2.5 U Taq DNA polymerase (Applied Biosystems). Amplification was carried out in a GeneAmp PCR System 9700 thermal cycler (Applied Biosystems). The reaction profiles included the following: one cycle of denaturation at 95℃ for 10 min, followed by 30 cycles of denaturation at 95℃ for 30 sec, annealing at 60℃ for 30 sec and extension at 72℃ for 60 sec and then final extension at 72℃ for 60 min. PCR products were resolved with an internal size standard (GeneScan 500 LIZ; Applied Biosystems). Genotyping was performed by electrophoresis on an automated DNA sequencer (ABI 3100; Applied Biosystems), and genotype data were analyzed using the GeneScan and Genotyper software (Applied Biosystems).

Equine mitochondrial D-loop analysis was performed to confirm mitochondrial genome identity on the foal produced by SCNT and on the oocyte donor cell used for SCNT. We sequenced the equine mitochondrial D-loop, which is ~1 kb including the D-loop adjacent region, in 11 mtDNA samples. Two primer sets for amplification and sequencing analysis were designed based on GenBank sequences (NCBI accession no. X79547). The primer sequences of the two PCR primer sets were: D-loop1-Forward: TCCCCCTCATATTAAACCAGAA, Reverse: CATGTCAGGTGGGTATGGTG; D-loop2-Forward: AACGTTTCCTCCCAAGGACT, Reverse: GTGTGAGCATGGGCTGATTA. PCR was performed in a mixture of 1.25 pM of each primer, 30 ng genomic DNA, 250 µM dNTPs and 1.5 U Taq DNA polymerase (Applied Biosystems) in the buffer provided by the manufacturer. Amplification was carried out in a GeneAmp PCR System 9700 thermal cycler (Applied Biosystems) under touch-down conditions [9]. Nucleotide sequencing was performed using the Big Dye Terminator version 3.1 Cycle Sequencing kit (Applied Biosystems) and an ABI 3730 Genetic Analyzer (Applied Biosystems). Sequence variants were verified by means of chromatograms using the DNA star software.

In experiment 1, the type of needle used for TVUFA in mares was examined. Two types of needle were compared, a 12-G double lumen needle and a 14-G double lumen needle system using a short disposable needle. To determine the effects of follicular size on recovery rates, the diameters of ovarian follicles were first measured. Follicles were classified into three types: (1) Group A (from 25 to 30 mm); (2) Group B (from 30 to 35 mm); (3) Group C (> 35 mm). In experiment 2 and 3, the developmental competence of oocytes recovered in vivo using TVUFA with a short needle was investigated by IVM and SCNT. To determine the maturation rate according to the status of COCs, oocytes were denuded and examined for condition of the cytoplasm and extrusion of the first polar body (PB). In vitro matured oocytes recovered by scraping follicles from ovaries of slaughtered mares were used at the same time to compare their in vivo development after SCNT and ET. In experiment 4, nuclear DNA extracted from donor cells and blood samples of the cloned foal, the oocyte donor and the recipient mare and analyzed to determine the identity of the cloned foal. Mitochondrial DNA was also analyzed to confirm that the foal was not genetically related to the recipient mare.

All analyses were performed using the Statistical Analysis System (ver. 9.1; SAS Institute, USA). Data were analyzed using a general linear model followed by the least significant difference mean separation procedure when treatments differed at p < 0.05. Percentage data were arcsine transformed prior to analysis to maintain homogeneity of variances. Percentage data are expressed as the means ± SEM.

The recovery rate (47.5%) with the 14-G short needle system did not differ significantly from that (35.0%) using the 12-G long needle (Table 1). Comparison of the recovery rate according to the size of follicles revealed recovery rates of Group A, B and C of 55.7% (n = 56), 41.4% (n = 68) and 36.5% (n = 24), respectively.

Table 2 shows the IVM status of equine oocytes according to the type of COCs. The maturation rate of Ex oocytes (97.1%) was not different from that of Co oocytes matured for 24 to 26 h or for 28 to 30 h (83.3% and 91.0%, respectively). The percentage of oocytes without PB (13.9%) in Co oocytes matured for 24 to 26 h was not significantly different from that (0.0%) in Ex oocytes.

In vivo efficiency of oocytes recovered by TVUFA (in vivo oocytes) was compared with that of oocytes derived from slaughtered ovaries (in vitro oocytes) after SCNT and ET (Table 3). The fusion rate of in vivo oocytes was not different from that of in vitro oocytes. After the transfer of SCNT embryos derived from in vitro oocytes, one embryo formed embryonic vesicles (1/42 = 2.4%) at early pregnancy examination, but this could not be maintained over 30 days. Among 26 SCNT embryos derived from in vivo oocytes, three formed embryonic vesicles (3/26 = 11.5%) after transfer to 13 recipient mares and one went to term (1/26 = 3.8%, gestation periods: 342 days) (Fig. 2). The cloned foal has grown to date without any observed abnormalities.

Parentage analysis data of the donor cell, the cloned foal, the oocyte donor and the recipient mare are shown in Tables 4 and 5. Analysis of seventeen equine specific microsatellite loci confirmed that the cloned foal was genetically identical to the donor cell, and there was no genetic relationship between the cloned foal and its surrogate mother (Table 4). Analysis of mitochondrial DNA revealed perfect homology between the cloned foal and the oocyte donor, while the mtDNA of the cloned foal was different from that of the donor cell and the surrogate mother (Table 5).