Journal of Equine Veterinary Science 89 (2020) 102987

Contents lists available at ScienceDirect

Journal of Equine Veterinary Science

journal homepage: www.j-evs.com

Original Research

Historical Aspects of Equine Embryo Transfer

W.R. (Twink) Allen*, Sandra Wilsher
The Paul Mellon Laboratory of Equine Reproduction, Newmarket, Suffolk, United Kingdom
Sharjah Equine Hospital Reproduction Laboratory, Sharjah, United Arab Emirates

a r t i c l e i n f o a b s t r a c t

Article history: Early embryo transfer in equids was undertaken simultaneously in the early 1970s in Cambridge, En-
Received 1 November 2019 gland, and Kyoto, Japan. Both groups achieved limited success when flushing the uterine horn ipsilateral
Received in revised form to the side of ovulation but the rates improved markedly when the whole uterus was flushed on real-
1 March 2020
ization of the continued movement of the embryo throughout the uterine lumen after day 6. Initial
Accepted 2 March 2020
Available online 10 March 2020
transfers of embryos to recipient mares were carried out surgically, but nonsurgical transfer via the
cervix has been used subsequently with increasing success, culminating in pregnancy rates of 75%e90%
today. Experimental use of embryo transfer in horses and donkeys demonstrated the unique ability of
Keywords:
Embryo transfer
equids to carry to term a full range of interspecies hybrid conceptuses and extraspecies pregnancies
Interspecies created by embryo transfer. Furthermore, splitting of day 4e8 cell embryos and day 6 compact morulae
Extraspecies allowed the creation of genetically identical twin foals. But despite these and other significant advances
Placenta over the past 45 years, a persisting limitation is the relatively low embryo recovery rates from donor
Endometrial cups mares treated with exogenous gonadotropins in attempts to induce them to superovulate. This is due to
Equine chorionic gonadotropin (eCG) the toughness of the ovarian tunica albuginea which forces ovulation through the ventrally situated
ovulation fossa where multiple follicles compete with each other and luteinize before they can ovulate
properly.
© 2020 Elsevier Inc. All rights reserved.

1. Introduction pregnancies from nonsurgical transfer in their first attempts [3],

they subsequently achieved a very encouraging 40% pregnancy rate
Equine embryo transfer can be said to have begun in the early after transferring 15 day 6 morulae/early blastocysts to synchro-
1970s. The author (WRA) and his “boss”, Tim Rowson, at the nized recipient mares via a long needle passed through the anterior
renowned Animal Research Station in Cambridge, adapted surgical fornix of the vagina and thence through the wall of the uterus,
methods of embryo recovery and transfer developed previously thereby avoiding passage through the cervix which they assumed
there in cattle and sheep to the horse and obtained a number of had been the cause of their failure to achieve any pregnancies from
pregnancies, including three from the transfer of interspecies their first attempts.
hybrid mule (\horse
_donkey) and hinny (\ donkey
_ horse) Two major barriers existed in the 1970s which greatly limited
embryos, into the opposite species of the dam of the embryo [1,2]. the early development of embryo transfer in horses compared with
Simultaneously, a group of workers in Japan were using nonsurgical the other domestic farm species such as cattle, sheep, goats, and
methods to both recover and transfer horse embryos between pigs. First, and perhaps most important at that time, the registra-
donor and recipient mares and although they failed to achieve any tion authorities for most pure “breeds” (rather than “types”) of
horses throughout the world, and very much led by the national
Jockey Clubs representing the Thoroughbred racehorse, would not
Animal welfare/ethical statement: There are no ethical considerations as this is a
permit entry into their respective stud books of progeny conceived
review and all data cited is from historic studies undertaken under UK Home Office by either artificial insemination or embryo transfer. Second, there
approval. did not then, and nor does there still today, exist a practical and
Conflict of interest statement: There are no conflicts of interest. efficient method for inducing superovulation in donor mares, as
* Corresponding author at: W.R. (Twink) Allen, The Paul Mellon Laboratory of
existed for the female donors of the other farm species. But despite
Equine Reproduction, 18 Woodditton Road, Newmarket, Suffolk, CB8 9BJ, United
Kingdom. these very real drawbacks, a number of significant advances in both
E-mail address: [email protected] (W.R. Allen). the methodologies and results of equine embryo transfer occurred

https://doi.org/10.1016/j.jevs.2020.102987
0737-0806/© 2020 Elsevier Inc. All rights reserved.
2 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987

during that early period and this paper attempts to review that of LH during estrus in the mare contrasts sharply with the rapid,
progress. But before doing so, it is pertinent to highlight some of the more pulsatile release of LH that occurs some 24e30 hours before
unusual aspects of the estrous cycle and early pregnancy in the ovulation in other farm species [9] and it is very likely responsible
mare which have impinged, both positively and negatively, on the for some of the twin and early diestrous second ovulations which
development and widescale use today of embryo transfer and occur in some mares [10].
related embryo technologies in equids. The marked variability which exists in both the duration of
estrus and in the occurrence of ovulation in relation to the onset of
2. The Estrous Cycle of the Mare estrous signs precludes any reference to estrus as a pivotal point on
which to calculate the age of embryos for recovery purposes. The
The mare is a seasonally polyestrous animal with a distinct time of ovulation is the only meaningful parameter in this regard
breeding season during spring, summer, and early autumn. The and day 0 is usually taken as the day on which ovulation is first
estrous cycle is normally 21e23 days in length, consisting of detected transrectally as having occurred. However, because donor
16e17 days of luteal activity or diestrus and 4e6 days of estrus, the mares were often only examined once daily by rectal palpation in
duration of which differs markedly between individual mares and the 1970s, any estimate of the time of ovulation was subject to an
is also influenced by season, being prolonged at the start of a inherent error which could be as great as 24 hours. This was re-
breeding season in early spring, shortest at the summer solstice, flected in the considerable variation found in the stage of devel-
and lengthening again in autumn [4,5]. opment of horse morulae or blastocysts recovered on days 6e8 for
Some 8e12 follicles >1 cm in diameter are usually present in the embryo transfer [11].
ovaries on the first day of estrus, but only one of these, or occa-
sionally 2 or 3 in some breeds of horses, progress to ovulation. The 3. Synchronization of Ovulation
follicle reaches 3e5 cm in diameter before ovulating and is there-
fore readily palpable and ultrasonographically visible through the Tim Rowson, Bob Moor, and other colleagues at the Animal
wall of the rectum. Ovulation can only take place through the small Research Station in Cambridge demonstrated that the pregnancy
semicircular region on the ventromedial surface of the ovary rate after surgical embryo transfer in sheep and cattle was highest
known as the ovulation fossa (Fig. 1) which is not covered by the when ovulation was exactly synchronized in donor and recipient
tough, fibrous tunica albuginea that envelops the rest of the ovary animals, and began to drop off sharply when the degree of donor-
[6] and it occurs usually 12e24 hours before cessation of the recipient synchrony exceeded 24 hours either way in the cow or
behavioral signs of estrus [7]. Luteinization of the resulting corpus 48 hours either way in the ewe [12]. Only the one published study,
hemorrhagicum occurs quickly so that peripheral plasma proges- that of Oguri and Tsutsumi [13], had seriously examined the
terone concentrations have generally risen to 1e2 ng/mL by question of donor-recipient synchrony in relation to equine embryo
24 hours, and frequently reach as high as 3e5 ng/mL by 48 hours, transfer. From the transfer of 79 horse blastocysts using two
after ovulation. This rapid rise permits the use of daily plasma different nonsurgical techniques, initial conception rates of 63%,
progesterone measurements during estrus as an accurate method 54%, 41%, 38%, and 0%, and live foal rates of 50%, 35%, 30%, 19%, and
for determining the day of ovulation for the purpose of embryo 0%, respectively, were obtained when the degree of synchrony
transfer. between recipient and donor mares was, respectively, 2, 1, 0, þ1,
Pituitary gonadotropin release patterns in the cycling mare and þ2 days. Although the situation was slightly confounded by the
differ markedly from those shown by the other large domestic use of two different transfer methods, and only half the number of
species. Serum FSH concentrations show a biphasic rise with a animals in the 2- and þ2-day groups compared with the other
pronounced mid-diestrous surge and a second, generally smaller, three groups, the figures nonetheless gave a clear indication that a
peak which coincides roughly with ovulation [8]. Luteinizing hor- degree of asynchrony as great as 48 hours or more was not disad-
mone concentrations on the other hand, remain baseline vantageous in terms of pregnancy rate, provided the recipient had
throughout diestrus but rise steadily during estrus to reach a peak ovulated after, not before, the donor.
1e2 days after, rather than before, ovulation. This prolonged release In a more recent study [14] involving the nonsurgical transfer of
a total of 78 day 10 blastocysts to recipient mares which ranged
from 9 to þ 2 days with respect to the donor, pregnancy rates
remained high in recipients that had ovulated up to 4 days after, but
not before, the donor and 3 of 8 mares that were as much as 7 days
negatively asynchronous with the donor became pregnant
(Table 1). Interestingly, these 3 minus 7 day embryos slowed their
rate of development, as monitored ultrasonographically and from
measurement of the onset of equine chorionic gonadotropin (eCG)
secretion in the recipient mares, to the “ovarian stage” of the
recipient rather than that of the original donor (Fig. 2).
The critical period for recognition of pregnancy in the mare
occurs around day 12 after ovulation [15]. This important message
from the embryo to prevent luteolysis clearly involves suppression
of the normal, cyclical release of prostaglandin F2a by the endo-
metrium at the end of diestrus [16] and, surprisingly, it seems not to
be related to the large quantities of estrogens secreted by the
blastocyst from as early as day 10 [17e20]. It is conceivable that
handling, changes in temperature, and any other adverse condi-
tions which the blastocyst experiences during a nonsurgical re-
covery and transfer process may temporarily halt its development,
Fig. 1. A sectioned horse ovary showing the preovulatory follicle adjacent to the thereby enabling it to survive so well when transferred to a very
ventral ovulation fossa through which it will ovulate. negatively asynchronous recipient uterus. Conversely, such delayed
 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987 3

Table 1
Influence of donor-recipient asynchrony on pregnancy rates and early embryonic loss rates after transfer of day 10 embryos.

Degree of Donor: Recipient Days After Ovulation of the No. of Recipient Mares Pregnant (%)
Asynchrony (Days) Recipient on Day of ET
2 Days Post ET 10 Days Post ET With Embryonic Heartbeat
a
þ4 14 3/6 (50%) 0/6 (0%) 0/6 (0%)a
þ2 12 6/6 (100%) 6/6 (100%) bc 6/6 (100%) bd
0 10 6/6 (100%) 6/6 (100%) bc 6/6 (100%) bd
1 9 6/8 (75%) 5/8 (63%) abc 5/8 (63%) abd
2 8 6/8 (75%) 6/8 (75%) bc 5/8 (63%) abd
3 7 6/8 (75%) 6/8 (75%) bc 6/8 (75%) bcd
4 6 8/8 (100%) 8/8 (100%) bc 7/8 (83%) bcd
5 5 6/8 (75%) 6/8 (75%) bc 6/8 (75%) bcd
6 4 5/6 (83%) 5/6 (83%) bc 5/6 (83%) bcd
7 3 8/8 (100%) 6/8 (75%) bc 3/8 (38%) abc
9 1 4/6 (67%) 2/6 (33%) ab 0/6 (0%)a
Control AI derived, non-ET pregnancies 9/9 (100%) 9/9 (100%)c 9/9 (100%)d

Abbreviations: AI, artificial insemination; ET, embryo transfer.

a,b,c,d
Values with different superscripts within a column are significantly different at P < .05.
Adapted from Wilsher et al [14].

development may prevent the blastocyst from liberating its after the second injection (day 20) and approximately 80% of them
maternal recognition of pregnancy signal in time to prevent ovulating during a 4 day period between days 20 and 24 [23]. The
luteolysis in an “advanced recipient”. success of this double prostaglandin synchronization schedule,
The development of efficient methods to control and synchro- with or without injection of human chorionic gonadotropin (hCG)
nize estrus and ovulation in the large domestic species provided a given on days 7 and 20 after the first prostaglandin treatment to
keystone for the successful commercial exploitation of embryo hasten ovulation in each prostaglandin-induced estrus, was
transfer in cattle [12]. The horse was no exception and the need to confirmed [24,25], and Eric Palmer showed that daily administra-
be able to control ovulation in groups of donor and recipient mares tion of the orally active synthetic progestagen, allyl trenbolone
was perhaps even greater than in the cow because of the much (Regumate, Roussel Uclaf, Paris), to cycling mares for 10 days
longer period of estrus and the marked individual variation be- accompanied by prostaglandin analog on the last day of progesta-
tween mares in the time from the onset of estrous behavior to the gen administration and hCG a further 10 days later, likewise
occurrence of ovulation. The ability of a single injection of a rela- resulted in synchronized ovulation in over 75% of the treated ani-
tively low dose of prostaglandin F or various prostaglandin analogs mals [26].
to induce rapid luteolysis in the diestrous mare [21,22] provided the
basis of the treatment regimens developed over the early years to 4. Superovulation in the Mare
control ovulation in mares. Two injections of a prostaglandin
analog such as cloprostenol (Estrumate, ICI Pharmaceuticals Divi- Equine chorionic gonadotropin, formerly called pregnant mare
sion, Cheshire, UK), given 14 days apart to a group of randomly serum gonadotropin, the high-molecular-weight glycoprotein
cycling mares, resulted in >90% of them displaying estrus by 6 days hormone which uniquely expresses both FSH-like and LH-like

Fig. 2. Peripheral serum progesterone (closed symbols) and eCG (open symbols) profiles in two recipient mares to which a day-10 embryo was transferred when they were only
3 days after ovulation to create a donor-recipient asynchrony of as much as 7 days. Note, how the eCG concentrations and the progesterone concentrations from the first secondary
ovulation do not begin to rise until conceptus age day 45 after ovulation, some 7 or 8 days later than normal.
4 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987

activities [27,28] remained the most widely used and effective Subsequently, the same group [36] used a commercially avail-
gonadotropin for stimulating ovulation in cattle, sheep, goats, and able equine pituitary extract (Pitropin, Biological Specialties, Wis-
pigs in the 1970s and 1980s. It was therefore somewhat ironic that consin) to similarly treat cycling mares daily for 6 days during mid-
eCG, which is present in high concentration in the blood of mares estrus or late diestrus/early estrus. As in the previous experiments,
between 40 and 120 days of pregnancy [29,30], proved completely ovulation rates in the treated and control mares were assessed by
ineffective with regard to stimulating follicle development when rectal palpation only, and on this basis, 6 of the 8 treated mares
given to cycling and pregnant mares, even in repeated doses as high were considered to have produced between 2 and 4 ovulations each
as 150,000 iu [7,31]. Studies by the late Francesca Stewart working during the estrus after treatment. However, when they were sub-
with the author [32,33] demonstrated that the gonadotropic re- jected to nonsurgical uterine flushing between days 8 and 10 after
ceptors in equids, as distinct from those in other species, have the induced ovulation, the embryo recovery rate was only 35% of
somehow evolved the ability to differentiate between the binding that expected based on ovulation rate in the 8 treated animals,
properties of their own chorionic gonadotropin (eCG) and those of compared with 67% of that expected in 20 untreated control mares.
their pituitary gonadotropins, FSH and LH. That such an unusual A subsequent study by the same group [37] found that the
mechanism allowing differential recognition of, and response to, conception rate was markedly reduced, compared with that in
different types of gonadotropin molecules should have evolved in untreated controls, in mares treated daily for 14 days with equine
the mare is perhaps not surprising when one considers that at the pituitary extract to induce single and/or multiple ovulations early
time of peak eCG production around day 60 of gestation, a total in the breeding season (31% vs. 60%). In addition, only one
quantity in excess of 1,000,000 i.u eCG is likely to be circulating in conceptus was still present at day 50 after ovulation in all 8 of the
her blood. Hence, were such a protective system not to exist, her 27 treated mares that conceived during the estrus in which two or
ovaries would be hyperstimulated and “blown apart” each time she more ovulations had been diagnosed per rectum.
became pregnantdas indeed can occur in the ovaries of a donkey The incidence of spontaneous twin ovulations in the mare is
carrying either an interspecific hinny conceptus or an extraspecific strongly breed-dependent, ranging from as high as 20% in the
horse conceptus created by embryo transfer (Fig. 3; [31]). The larger, more modern breeds such as the Thoroughbred, to zero in
greatly increased ratio of FSH to LH found to exist in horse (1.4) and smaller, more primitive breeds such as the Welsh Mountain Pony.
hinny (0.8) CG compared with donkey CG (0.1) [28] was very likely Although the Wisconsin studies demonstrated that administration
the cause of the marked differences in ovarian responses to these of equine pituitary gonadotropins in sufficient quantity could
variations in fetal genotype. induce some degree of multiple ovulations in a proportion of
One early advance with regard to superovulation in the mare treated mares, it became clear that this type of therapy would not
came from experiments undertaken in Wisconsin [34] which be advantageous for routine use in a commercial equine embryo
induced estrus and ovulation in 25 of 29 seasonally anestrous pony transfer program. The cost and impracticality of the treatment
mares by giving daily injections for 14 consecutive days of a crude regimen, the variability in response rates and the very significant
extract of equine pituitary glands. Based on rectal palpation of the reduction in pregnancy rate in treated mares combined to
ovaries, the authors noted that 14 of the 25 treated mares which did outweigh any possible advantages. However, the more recent and
ovulate had two or more ovulations. Subsequently, the same group encouraging results from attempted induction of multiple follicle
[35] obtained a similar proportion of double and triple ovulations growth in donor mares by the administration of recombinant-
when giving the same type of course of daily injections of equine derived equine pituitary gonadotropins [38] may provide a useful
pituitary extract, with or without a subsequent injection of hCG, for and practical regime for improving oocyte recovery rates when
14 days to seasonally anestrous mares and for 6 consecutive days to undertaking ovum pick-up.
cycling mares, between either days 11e16 of diestrus or days 1e6 of Back in the 1970s, the most practical solution to the failure of
estrus. Once again, however, the ovulation figures were based on superovulation was to increase the frequency of ovulation in
rectal palpation alone and were not supported by any attempts at selected donor mares by repeatedly shortening their estrous cycle
embryo recovery. Furthermore, the repeated injections of a crude with prostaglandins. In the author's laboratory, donor mares were
pituitary extract caused the death of one mare and gave rise to routinely injected with 250 mg of prostaglandin analog (Clopros-
injection site abscesses of many of the others. tenol; ICI Pharmaceuticals Division) immediately after undergoing
nonsurgical embryo recovery on day 7 or 8 after ovulation. This
induced a return to estrus in 2e3 days when each mare was again
inseminated with the required semen and usually ovulated within
4e5 days. After a further 7 days, the mares were ready to be flushed
again and in this way the interval between successive embryo re-
covery attempts was shortened from the expected 21 or 22 days of
the normal estrous cycle to about 13e15 days. Repeated shortening
of the estrous cycle of donor mares in this manner may or may not
lower their fertility and from one pony mare flushed a total of 6
times during a period of 83 days in 1979, produced 5 normal
blastocysts. On the other hand, subsequent studies have indicated
that repeated embryo recovery attempts may induce deleterious
changes in the uteri of donor mares [39,40].

5. Surgical Embryo Recovery

Because the horse embryo does not enter the uterus until days
6e6.5 after ovulation [13,41], embryo recovery at any stage before
Fig. 3. Pregnant uterus and hyperstimulated ovaries in a Jenny donkey carrying a
transferred horse conceptus. One ovary has ruptured as a result of excessive growth of
day 6 (late morula) necessitated entry into the abdomen via a
secondary follicles and accessory corpora lutea stimulated by the high concentrations ventral midline laparotomy under general anesthesia followed by
of horse eCG secreted by the larger horse endometrial cups. flushing of the ipsilateral oviduct with 30e40 mL of medium. The
 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987 5

flushing technique was similar to that described for the cow [42]
although two factors complicated the process in the mare. First, the
ovarian ligament on the right side of the animal is considerably
shorter than that on the left so that, in nulliparous pony mares,
excessive traction was often necessary to exteriorize the right ovary
and oviduct sufficiently through a midventral incision to enable the
oviduct to be flushed. This interfered with the flushing technique
and carried the risk of rupture of the ovarian ligament with
resulting hemorrhage. Second, the uterotubal junction in the mare
takes the form of a firmly occluded papilla protruding into the
uterine lumen (Fig. 4) [43]. Although this permits fluids to flow
from the oviduct into the uterus, its very-effective valve-like action
completely prevented the reverse passage of fluid from the uterus
into the oviduct. At that time, it was not possible to achieve
retrograde oviduct flush by injecting medium into the uterine
lumen and then forcing it into the oviduct, as was carried out in the
cow.
Fig. 5. Illustration of the technique of surgical embryo recovery in equids. The rigid
The method used successfully for many years in the author's plastic cannula was held between thumb and forefinger in the ampulla of the oviduct
laboratory [1,2,31] involved the insertion of a 5e7 cm length of and the fluted end of the glass pipette was held also by digital pressure in the tip of the
firm, polystyrene cannula (5 mm O.D) into the dilated fimbria of the ipsilateral uterine horn adjacent to the uterotubal papilla.
oviduct. This was connected by a short length of flexible tubing to a
20 mL syringe containing the flushing medium. The cannula was
held in place in the oviduct and back-flow of fluid prevented by firm mares between 1 and 3 days after ovulation yielded 28 embryos
manual pressure. A 1-cm longitudinal excision was made in the between the 1- and 8-cell stages of development and 4 freshly
outer serosa and myometrium of the ipsilateral uterine horn about ovulated unfertilized oocytes. Many old, unfertilized oocytes from
3 cm posterior to the uterotubal junction. The endometrium was previous estrous cycles were also recovered in most flushes [46].
pierced with the blades of a pair of rounded scissors and the fluted
end of a short length of glass Pasteur pipette, previously shaped in a 6. Nonsurgical Embryo Recovery
Bunsen burner flame, was inserted into the incision and held firmly
against the uterotubal junction by manual pressure. The other end Certain features of the mare's reproductive tract make the
of the glass tube was tapered and connected to a length of soft whole process of nonsurgical uterine flushing much simpler than
rubber tubing which passed to the embryo collection dish (Fig. 5). its equivalent in the cow. The mare's cervix is relatively short and is
Some 30e40 mL of flushing medium [44] was forced through the more easily distended in diestrus than that in the cow, therefore
oviduct and collected into 2 or 3 embryo dishes. allowing the insertion of a much larger diameter catheter into the
Despite the impracticality of the method, a high recovery rate of uterus. Furthermore, the horns of the mare's uterus are not coiled
early-stage embryos and/or unfertilized oocytes could be achieved like those of a cow and they diverge from each other at a much
when flushing young, fertile mares [45]. For example, Allen and greater angle, thereby more easily permitting the passage of a
Rowson [2] obtained 23 embryos from 30 Welsh Pony mares (77% flushing catheter from the body of the uterus into each uterine
recovery rate) flushed between days 1 and 6 after ovulation. In a horn. Two nonsurgical flushing methods were used originally in the
subsequent experiment, surgical flushing of the left oviduct of 39 mare.

6.1. Single-Horn Flushing

A rigid instrument [3]; Fig. 6 [2]; or a flexible rubber or poly-

thene two-way catheter [11,47] was passed through the cervix into
the body of the uterus. With the operator's arm in the rectum the
catheter was then manipulated into the horn of the uterus ipsi-
lateral to the recently ovulated follicle so that the inflatable rubber
cuff was situated at the base of the horn. The cuff was inflated with
40e50 mL of air or water to isolate the lumen of the horn from the
remainder of the uterus (Fig. 7) and 50e60 mL of medium was
flushed rapidly into the horn using a catheter-tipped 60 mL syringe
attached directly to the free end of the catheter. While massaging
the fluid-filled uterine horn per rectum, the medium was then
either positively withdrawn into the same syringe or was allowed
to flow back out of the catheter into a suitable collecting vessel. The
whole process was repeated twice.

6.2. Whole Uterus Flushing

This method soon replaced the ipsilateral horn only flush

[13,36,48,49] and it simply involved passing the flushing instru-
ment, either the rigid instrument [46] or the flexible two-way
Fig. 4. The tightly closed and protruding uterotubal papilla or junction (UTJ) at the catheter with inflatable cuff through the cervix into the body of
anterior tip of a uterine horn in a normal diestrous mare. the uterus. The cuff was inflated and a slight backward tension
6 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987

Fig. 6. The older method of single uterine horn embryo nonsurgical embryo recovery
using a rigid, metal, 3-way flushing apparatus manipulated into the base of the ipsi-
lateral uterine horn with the operator's arm in the mare's rectum.

maintained on the catheter so that the cuff was kept adjacent to, and
hence occluded, the internal os of the cervix (Fig. 8). The flushing
medium (500e1000 mL of phosphate buffered saline containing 5%
Fig. 8. The whole uterus, nonsurgical, embryo flushing technique using the flexible
v:v bovine serum albumin) was run into the uterus under gravity
two-way 18-FG catheter shown in Fig. 7 inserted into the body of the uterus and
until the lumen was distended (Fig. 8) and it was then allowed to run withdrawn so that the inflated cuff occluded the internal os of the cervix. Around
back out into a suitable collecting vessel while the operator 500 mL of flushing medium was run into the uterus and recovered by gravity while the
massaged the uterus per rectum. The whole process was repeated operator continued to massage the uterine horns per rectum to “flush” the embryo out
two or three times so that, eventually, a total in excess of 2 L of from between folds in the endometrium.

medium had been infused into and recovered from the uterus in
three stages. Because embryo filters did not exist in these early days,
horn flushes of pony mares and Jenny donkeys which had been
the medium was collected back into 500 mL measuring cylinders
covered naturally by either a fertile horse stallion or a fertile Jack
which were allowed to stand on the laboratory bench for
donkey, yielded 26 intraspecies and interspecies hybrid embryos to
15e20 minutes to enable the embryo to sink to the bottom of the
give an overall recovery rate of only 48%.
cylinder. The excess flushing medium was carefully aspirated or
In their later article, Oguri and Tsutsumi [13] reported greatly
poured off and the remaining 20e30 mL placed in a concave embryo
improved embryo recovery rates when using a similar, rigid three-
dish and searched for the embryo using a dissecting microscope.
way flushing instrument passed only through the cervix so that the
The whole uterus flushing technique, although it used consid-
whole uterus was flushed with a total of 1500 mL of medium in
erably more medium, immediately doubled the embryo recovery
three “lots”, each of 500 mL. In their first experiment they recov-
rate compared with the ipsilateral single horn method. In earlier
ered 18 blastocysts from 20 attempted flushes of pony and half-
studies [2,3] using the original rigid flushing instrument (Fig. 6) to
breed mares between 6 and 7 days after ovulation and in the
flush only the ipsilateral uterine horn in fertile pony mares, embryo
subsequent much larger study [48] involving the recovery of a total
recovery rates of only 49% and 40%, respectively, were achieved.
Likewise, Tischner [11] recovered only 11 embryos from 21 mares
(52%) by the ipsilateral horn method and, in the author's laboratory
during the 4-year period from 1976 to 1980, a total of 54 single-

Fig. 7. The infusion of 60e100 mL of the flushing medium into the ipsilateral uterine Fig. 9. Surgical transfer of a horse embryo via midline laparotomy using a Pasteur
horn via a catheter-tipped syringe connected to the flexible 18-FG catheter with its pipette passed through a hole punched in the uterine horn by a blunted 18 gauge
inflatable cuff occluding the exit from the ipsilateral uterine horn into the uterine body. needle.
 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987 7

(±50%) occurs spontaneously in all breeds of horses [7,50,51] with a

strong likelihood in lactating mares for the conceptus to implant in
the uterine horn contralateral to the implantation side of the pre-
vious conceptus [52]. Furthermore, if the blastocyst was destined to
migrate for whatever reason, it appeared to do so very soon after
entering the uterus on day 6 after ovulation, so it has already vacated
the ipsilateral horn by day 7 or 8. Transrectal ultrasonography and
intrauterine videoendoscopy have since shown convincingly that
the equine conceptus moves constantly throughout the entire
uterine lumen between days 6 and 16 after ovulation [53].
In due course, a more precise timing of 144e168 hours for the
entry of the embryo into the uterus was determined [54], and it was
also reported at the other extreme that larger embryos flushed at
day 9 or later were more likely to be damaged during collection and
transfer [55,56]. Hence, nowadays, in a clinical setting, most em-
bryos destined for transfer to recipient mares are recovered for
Fig. 10. Genetic drift in the nursery. Donkey (2n ¼ 62), Przewalski horse (2n ¼ 66), and embryo transfer on day 7 or 8.
a common zebra (2n ¼ 44) foals with their recipient horse (2n ¼ 64) mothers.

7. Embryo Transfer
of 148 embryos from 200 attempted flushes between days 6 and 9
after ovulation, they made a direct comparison between the ipsi- 7.1. Surgical Transfer
lateral uterine horn and whole uterus flushing methods. While the
recovery rate from the former was only 44%, it rose to 87% when the The method used for surgical transfer of embryos in the mare
whole uterus was flushed. was very similar to that described in cattle [42]. The ipsilateral
The whole uterus flushing method was also adopted in the uterine horn was exteriorized through a mid-ventral laparotomy
author's laboratory during the 1981 breeding season when incision and a small hole made in the wall of the uterus and
attempting to recover donkey blastocysts for surgical transfer to endometrium by forcing through a blunted 18 gauge needle, the tip
mares. The same flexible plastic, 18 French gauge human degusta- of which was filled with solder and rounded off on a grindstone.
tion tube catheter used for single-horn flushes (Fig. 8) was used and The embryo was taken up in a minimal volume of medium
three “lots”, each of 300e500 mL, of medium were infused into the (0.05e0.01 mL) in the tip of a Pasteur pipette attached by a short
uterus under gravity and allowed to flow out again while manip- length of rubber tubing to a 1 mL tuberculin syringe. The tip of the
ulating the uterus per rectum to aid the fluid expulsion. In this pipette was eased through the hole in the uterine wall (Fig. 9) and
manner, a total of 27 donkey and 8 horse blastocysts were recov- passed posteriorly in the uterine lumen for 2e3 cm before gently
ered from the attempted flushing of 37 Jenny donkeys and 11 pony expelling the embryo.
mares between days 6 and 8 after ovulation, to give an overall re- Surgical transfer of equine embryos gave high pregnancy rates,
covery rate of 72% which compared favorably with the average provided the transfers were carried out >3 days after ovulation. In
recovery rate of 40%e50% obtained during the previous 4 years earlier experiments, the author and Tim Rowson [2] obtained
when using the ipsilateral horn flushing method. A similar recovery pregnancies in only 7 of 16 mares (44%) to which a single embryo
rate was reported by Douglas [36] who obtained 14 embryos from was transferred surgically between days 1 and 6 after ovulation;
the attempted whole-uterus flushing of 20 pony mares between embryos recovered on days 1e3 were transferred to the ipsilateral
days 8 and 16 after ovulation. oviduct, whereas those recovered on Days 4e6 were transferred to
From the available figures, and especially those provided by the the ipsilateral uterine horn. However, analysis of the results
Japanese workers, it was apparent that whole uterus flushing gave a showed that whereas only one of 8 embryos transferred on days
marked improvement in embryo recovery rate to that achieved by 1e3 continued to develop, 6 of 8 (75%) transferred to the uterus on
flushing only the ipsilateral uterine horn [49]. And when Butterfield days 4e6 gave rise to pregnancies that were still ongoing at 42 days
and Matthews published their excellent article in 1978 describing after ovulation.
intrauterine migration of the equine conceptus, it became generally Once the basic methods of nonsurgical recovery and surgical
realized that a high rate of transuterine migration of the conceptus transfer of embryos was established in the author's laboratory in

Fig. 11. Two-day old (A) donkey and (B) horse foals with their respective extraspecies recipient horse and donkey mothers. Note the large size of the donkey foal having benefitted
from gestation in the larger horse uterus.
8 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987

Fig. 12. The elongated, toothed, Wilsher forceps used to grasp and straighten the recipient mare's cervix to facilitate nonsurgical embryo transfer.

the 1970s, relatively few intraspecific transfers between horses complete expulsion of the medium and embryo when the plunger
were carried out, the great majority involving the transfer of of the syringe was depressed. With a gloved hand protecting the
interspecific hybrid mule or hinny embryos to a synchronized end of the pipette as it passed into the vagina, the operator's index
recipient of the opposite species (donkey and horse, respectively). finger was used to open the external os of the cervix and direct the
Subsequently this was extended to include reciprocal extraspecies catheter into the uterus. The embryo was then deposited either in
transfer of horse and donkey embryos to unmated synchronized the body of the uterus or, by manipulation with the operator's free
recipients of the other species and it also included transfer of hand in the rectum, in the ipsilateral horn.
Przewalski's Wild Horse (2n ¼ 66) and Common Zebra (2n ¼ 44) Just as nonsurgical recovery of equine embryos gave rise to
embryos of domestic horse recipient mares (2n ¼ 64; Fig. 10). Yet marked differences in recovery rates between research groups, so
despite the xenogeneic nature of these transfers, the conception the conception rates resulting from nonsurgical transfers varied
rate remained relatively high. For example, during 1979e1981, in- considerably. In their original study, Oguri and Tsutsumi [3] failed
clusive surgical transfer of 8 horse embryos to unmated Jenny to obtain any pregnancies when transferring 10 blastocysts and one
donkeys resulted in 5 pregnancies (62%) and transfer of a total of 34 morula via the cervix to the ipsilateral horn of precisely synchro-
donkey embryos to unmated pony mares during the same 3 year nized recipient mares. In their later study, however, they achieved a
period resulted in 24 pregnancies (71%; Fig. 11; [31]). 40% pregnancy rate after the nonsurgical transfer of 15 day 6
In these original studies ventral midline laparotomy under blastocysts to synchronized recipient mares via a long needle
general anesthesia was used and this remained the method of passed through the anterior fornix of the vagina and thence
choice for the author (WRA) until nonsurgical embryo transfer through the wall of the uterus, therefore avoiding passage through
methods gave equivalent pregnancy rates. However, midline lapa- the cervix, all as described previously. In pony mares, Allen and
rotomy was not the only surgical method and other research fa- Rowson [2] obtained five pregnancies from the nonsurgical transfer
cilities were transferring embryos through the flank using local of 7 blastocysts via the cervix. Also, as described previously, transfer
infiltration anesthesia in the standing, sedated recipient mare [57]. was carried out using a disposable plastic cattle insemination
pipette and the embryo was deposited in the body of the uterus
7.2. Nonsurgical Transfer without any attempt to direct the pipette into either of the two
uterine horns.
This was a very straightforward technique in the mare whereby In the later Japanese study [48], 82 blastocysts were transferred
the embryo, in 0.2e0.5 mL of medium, was drawn into the lumen of nonsurgically to unmated recipient mares. Of these, 45 were
a rigid plastic cattle insemination pipette or similar device using a transferred transcervically as described previously and the
syringe attached to the other end of the pipette. A 1e2 mL cushion remaining 37 were transferred via the cervix and deposited in the
of air was drawn into the pipette before the medium to ensure middle of the uterine horn ipsilateral to the previous ovulation after

Fig. 13. (A) Passing the Wilsher forceps into the vagina of the recipient mare facilitated by a duck-billed vaginoscope and visualization using a pencil torch. (B) The cervix has been
grasped by the Wilsher forceps and pulled backward to straighten the cervical canal and so allow easier passage of the insemination catheter carrying the embryo in 2e3 mL of
medium through into the body of the mare's uterus.
 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987 9

countries even banning the use of surgical transfer. Nonsurgical

transfer methods typically involve transfer of the embryo held in a
minimum volume (0.1e0.3 mL) of medium in a 0.25 or 0.5 mL
semen straw inserted into a rigid embryo transfer gun passed
through the cervix via the operator's gloved hand has become the
worldwide norm and, depending on the experience and resulting
expertise of the operator, gives pregnancy rates of 75%e80% in
well-synchronized recipient mares. However, a more recent
modification of the method involving insertion of an autoclaved,
duck-billed speculum into the mare's vagina provides clear visu-
alization of the cervix, which in turn, enables insertion of a pair of
elongated, toothed Wilsher forceps (Fig. 12) to grasp the external
cervical os and pull the cervix backward to “straighten” its lumen.
This then allows the embryo handler to pass an insemination
pipette containing the embryo in a much larger volume of medium
(2e3 mL) easily through the cervix to deposit the embryo in the
body or horn of the uterus (Fig. 13). The great advantage of the
Wilsher forceps method is the insertion of only sterilized in-
Fig. 14. A recipient mare and its Welsh pony transferred foal in Krakow, Poland, after
struments into the vagina and through the cervix to maintain ste-
transport to Poland as a day 6 embryo in the oviduct of a rabbit.
rility and ease of passage through the cervix. As a result, the
technique routinely achieves pregnancy rates of >90%, even when
performed by even relatively inexperienced operators [59,60].
positioning the pipette by manipulation per rectum. No significant
difference in conception rate resulted from the two different 7.3. Storage, Transport, and Manipulation of Equine Embryos
transfer methods, the figures being 42% for transcervical transfer
and 44% by intracervical transfer. However, some pregnancy losses Little was published about the storage and transport of equine
occurred in both groups of recipient mares such that the initial embryos in the early days. In 1975, the author and colleagues [61]
figure of 42% for overall conception rate was reduced to 33% in recovered 6 morulae surgically from Welsh pony mares at the
terms of the birth of live foals. Animal Research Station in Cambridge and transported them by car
As the desire to undertake nonsurgical embryo transfer to Krakow, Poland, in the ligated oviducts of two rabbits, the
increased in the 1980s, much work was undertaken at other labo- journey lasting 33 hours. Five of the embryos had continued to
ratories, in particular, Colorado State University in Fort Collins, develop normally in the rabbits and four of these were transferred
University of California in Davis and Unite  Physiologie de la (3 surgically and 1 non-surgically) to Polish recipient mares in
Reproduction et des Comportements (INRA), Nouzilly, France, to which ovulation had been synchronized to within þ-48 hours by
hone the nonsurgical technique. This included improving knowl- the prior use of prostaglandin and hCG; the fifth embryo was
edge regarding the synchrony of recipients, the use of and phar- accidently lost. Three pregnancies developed from which one foal
macological manipulation of noncyclic recipients, and elucidating was aborted at day 292 of gestation and 2 colt foals were born live
factors affecting pregnancy rates and embryonic death after at term (Fig. 14).
nonsurgical embryo transfer [58]. Much of this work was also Early attempts in Cambridge to freeze surplus equine morulae
presented at the now established 4-yearly International Sympo- and early blastocysts proved unsuccessful. A freezing method used
sium on Equine Embryo Transfer, the first of which was held in New routinely for sheep embryos [62] was used and although the em-
York in 1984 under the sponsorship of the Dorothy Russell Have- bryos appeared viable and relatively undamaged when thawed,
meyer Foundation. transfer of 9 of them to suitable recipients failed to result in any live
Over the subsequent years as pregnancy rates improved, intra- foals; 8 mares were suspected of having become pregnant but no
cervical embryo transfer became the norm with some European conceptus developed beyond 40 days of gestation. The resistance of

Fig. 15. Two pairs of genetically identical twin foals produced by embryo bisection. (A) “Romulus” and “Remus” being held by the author and Mr Peter Burrell CBE, architect and the
Director of the British National Stud in Newmarket. (B) “Ophelia” and “Desdemona” with their recipient mothers at the Equine Fertility Unit in Newmarket.
10 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987

pellucida. These were then embedded in a small chip of agar gel

and the whole placed in the ligated oviducts of anestrous ewes for
3e5 days as described by Willadsen [63]. Of the total of 20 divided
embryos made in this manner and placed in sheep, 18 (90%) had
continued to develop normally to the morula or blastocyst stage
when subsequently recovered from the ewe; these were trans-
ferred surgically to the uteri of synchronized recipient mares, some
of which had been the donor of the original embryo 4e5 days
previously. A total of 10 pregnancies resulted and in five instances
only one half of the original embryo continued to develop, the other
half failing to progress in its recipient mare; in all three cases where
the donor also served as the recipient for one half of its manipu-
lated embryo, pregnancy did not result. Two sets of monozygotic
twins were born alive, one pair of colts in 1980 and a pair of fillies in
1981 (Fig. 15). Both these sets of identical twins originated from
separation of the two 4-cell embryos into 4 one-cell embryos. With
Fig. 16. A 43-day donkey-in-horse conceptus showing the poorly developed donkey the first of these, only 3 of the 4 reconstituted quarter embryos
chorionic girdle which had failed to detach and invade the horse maternal endome- continued to develop in the sheep and were, therefore, retrans-
trium at days 36e38 to form eCG-secreting endometrial cups.
ferred to mares. With the second manipulation, all 4 reconstituted
embryos progressed to the morula/blastocyst stage in the ewe and
the mare to superovulation and the resulting difficulty in obtaining were transferred to recipient mares. Although 3 pregnancies
sufficient numbers of embryos for investigation severely limited commenced, one of these failed between 43 and 51 days of gesta-
the amount of useful experimental study that could be applied to tion [64].
determining the optimum cooling and freezing requirements of These preliminary experiments demonstrated quite clearly that
equine embryos. the equine embryo like that of the ewe [63] and cow [65] remains
During 1979 and 1980, eight 2- to 8-cell horse embryos, recov- pluripotent until at least the 8-cell stage and may not be destroyed
ered surgically from the oviducts of Welsh pony mares between by manipulation and a reduction in overall cell number. Because
days 1 and 3 after ovulation had their zonae pellucida removed and the chances of live births of spontaneously occurring identical twin
the individual blastomeres were reconstituted in varying pro- horses is extremely remote, embryo micromanipulation could
portions of the original cell number in empty porcine zonae provide a useful tool to produce multiple monozygotic equine

Fig. 17. Depicting the recovery of a day 6 mule embryo from its horse mother followed by its bisection using a micromanipulator and transfer of the resulting demi-embryos, one to
a horse and the other to a donkey recipient. (A and B) The small and typically mule-like endometrial cups in the endometrium of the horse recipient (A) compared with the much
larger and hinny-like endometrial cups that developed from the other demi-mule embryo transferred to the donkey recipient (B) and which produced much higher eCG con-
centrations in maternal blood (C) thereby demonstrating clearly the profound influence of maternal uterine environment on endometrial cup development.
 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987 11

offspring that would be genetically identical and therefore immu- useful in equids than it is in other large domestic species. The
nologically compatible. These would have obvious potential value equine blastocyst is robust and reasonably undemanding in its
as experimental animals. environmental requirements. It is easily recovered by nonsurgical
means and it survives well if transferred correctly, even in an
7.4. Interspecies Embryo Transfer imperfectly synchronized recipient uterus. These features, coupled
with the desire to reproduce a whole variety of characters of
Equidae is one of the few genera of mammals in which mating excellence in the various breeds of horses has driven the rapid
between different member species with widely differing chromo- development and wide-scale application of equine embryo transfer
some complements can result in the birth of live hybrid offspring in recent years after the rather slow and uncertain start in the 1970s
[66]. For example, a Mongolian Wild Horse with a karyotype of 66 and early 1980s. But a major stumbling block remains. Namely, the
could be inseminated with semen from a Cape Mountain zebra resistance of the mare's ovaries to stimulation by exogenous go-
with a karyotype of only 32 and produce a live hybrid foal [67]. nadotrophins, combined with the unusually tough tunica albuginea
Allen and Rowson [1] originally reported the successful transfer which surrounds her ovaries and limits ovulation and release of the
of an interspecies mule embryo to a Jenny donkey and, vice versa, a oocyte to the ovulation fossa, thereby preventing any synchronous
hinny embryo to a horse mare and subsequent attempts to transfer ovulation of multiple follicles such as can be achieved successfully
intraspecific embryos between horses and donkeys were surpris- and repeatedly in other domestic animal species ranging from the
ingly successful [31]. From a total of 8 horse embryos recovered mouse to the cow.
nonsurgically at the blastocyst stage and transferred surgically to
the uteri of recipient Jenny donkeys, 5 established pregnancies
Acknowledgments
resulted. Two of these pregnancies were subsequently terminated
surgically on days 63 and 71 of gestation for gross and histological
The senior author (WRA) expresses his profound gratitude to
examinations of the endometrial cups and other fetal tissues, one
the late Thaddeus Mann FRS, Tim Rowson FRS and Cyril Adams, and
conceptus was aborted spontaneously on day 82 and the other two
to Bob Moor FRS, all of the now defunct Animal Research Station in
were carried normally to term and born live at 346 and 364 days of
Cambridge, for their kind practical support and wonderful aca-
gestation, respectively ([31]; Fig. 11). And with extraspecies trans-
demic stimuli nearly half a century ago.
fers, for example, a donkey embryo transferred to a horse recipient,
The historic research detailed in this paper was supported by
an even higher success rate was achieved. For example, transfer of a
funding from the Thoroughbred Breeders’ Association, the Horse-
total of 34 donkey embryos into synchronized recipient horse
race Betting Levy Board and the Dorothy Russell Havemeyer
mares resulted in 24 established xenogeneic pregnancies, giving a
Foundation of America.
transfer success rate of 71%.
The horse conceptuses in Jenny donkeys gave rise to large and
active endometrial cups with resulting high concentrations of eCG References
in the maternal circulation and they seemed readily able to implant
and continue normally to term. This was not the case with the [1] Allen WR, Rowson LEA. Transfer of ova between horses and donkeys. Proc. 7th
Int Congr Anim Reprod AI, Munich. 1972. p. 484e6.
donkey conceptuses transferred to mares, however, and of the 24 [2] Allen WR, Rowson LEA. Surgical and non-surgical egg transfer in horses.
such pregnancies established, only one resulted in the birth of a live J Reprod Fertil Suppl 1975;(23):525e32.
foal on day 346 (Fig. 11). In 12 other cases, abortion or resorption of [3] Oguri N, Tsutsumi Y. Non-surgical recovery of equine eggs and an attempt at
non-surgical egg transfer in horses. J Reprod Fertil 1972;31:187e98.
the already dead conceptus occurred between 47 and 105 days of [4] Andrews FN, McKenzie FF. Estrus ovulation and related phenomena in the
gestation and in the remaining 11 recipient mares, surgical removal mare. Res Bull Univ Mo Agric Stat 1941:329e40.
of the conceptus between days 60 and 87 showed definite evidence [5] Osborne VE. An analysis of the pattern of ovulation as it occurs in the annual
reproductive cycle of the mare in Australia. Aust Vet J 1966;42:149e64.
of placental and fetal damage resulting from a generalized maternal
[6] Heinze H. Pelviscopy in the mare. J Reprod Fertil Suppl 1975;(23):319e27.
cell-mediated immune attack against the xenogeneic conceptus. In [7] Day FT. Clinical and experimental observations on reproduction in the mare.
marked contrast to the “horse-in-donkey” pregnancies, none of the J Agric Sci 1940;30:244e56.
[8] Evans MJ, Irvine CHG. Serum concentrations of FSH, LH and progesterone
“donkey-in-horse” pregnancies showed any endometrial cup
during the oestrous cycle and early pregnancy in the mare. J Reprod Fertil
development as a result of failure of the donkey chorionic girdle to Suppl 1975;(23):193e200.
invade the horse endometrium between days 35 and 40 of gesta- [9] Geschwind II. Dynamics of pituitary gonadotropin secretion. J Anim Sci
tion (Fig. 16). 1972;34:19e28.
[10] Stabenfeldt GH, Hughes JT, Evans JW, Geschwind II. Unique aspects of the
The marked influence of uterine environment on development reproductive cycle of the mare. J Reprod Fertil Suppl 1975;(23):155e64.
of the equine conceptus was perhaps best shown by bisecting a day [11] Tischner M. Evaluation of deep frozen semen in stallions. J Reprod Fertil Suppl
6 mule morula and transferring one demi-mule embryo back to a 1979;(27):53e61.
[12] Rowson LEA. The role of reproductive research in animal reproduction.
recipient horse mare and the other demi-mule embryo to a recip- J Reprod Fertil 1971;26:11e28.
ient Jenny donkey (Fig. 17; [68]). Both conceptuses developed [13] Oguri N, Tsutsumi Y. Non-surgical egg transfer in mares. J Reprod Fertil
normally until they were removed surgically at 60 days of gestation. 1974;41:313e9.
[14] Wilsher S, Clutton-Brock A, Allen WR. Successful transfer of Day 10 horse
While the mule-in-horse pregnancy showed the typical small and embryos: influence of donor-recipient asynchrony on embryo development.
already necrotic mule endometrial cups (Fig. 17A) that produced Reproduction 2010;139:575e85.
low concentrations of eCG in the maternal circulation (Fig. 17C), the [15] Hershman L, Douglas RH. The critical period for maternal recognition of
pregnancy in Pony mares. J Reprod Fertil Suppl 1979;(27):395e406.
mule-in-donkey pregnancy showed much larger, still-active [16] Neely DP, Kindahl H, Stabenfeldt GH, Edqvist IE, Hughes JP. Prostaglandin
endometrial cups (Fig. 17B) that had secreted very high concen- release patterns in the mare: physiological, pathophysiological and thera-
trations of eCG into maternal blood (Fig. 17C). Thus, in the “wrong” peutic responses. J Reprod Fertil Suppl 1979;(27):181e9.
[17] Flood PF, Betteridge KJ, Irvine DS. Estrogens and androgens in blastocoelic
uterus, the mule conceptus developed like a hinny pregnancy.
fluid and cultures of cells from equine conceptuses of 10 to 22 days gestation.
J Reprod Fertil Suppl 1979;(27):413e9.
8. Conclusions [18] Zavy MT, Mayer R, Vernon MW, Bazer FW, Sharp DC. An investigation of the
uterine luminal environment of non-pregnant and pregnant Pony mares.
J Reprod Fertil Suppl 1979;(27):403e10.
Many features of the reproductive anatomy and physiology of [19] Heap RB, Hamon M, Allen WR. Studies on estrogen synthesis in the equine
the mare make the technique of embryo transfer simpler and more conceptus. J Reprod Fertil Suppl 1982;(32):343e52.
12 W.R. Allen, S. Wilsher / Journal of Equine Veterinary Science 89 (2020) 102987

[20] Wilsher S, Allen WR. Intrauterine administration of plant oil inhibits luteolysis [44] Whittingham DG. Fertilization of mouse eggs in vitro. Nature 1968;220:
in the mare. Equine Vet J 2011;43:99e105. 592e5.
[21] Douglas RH, Ginther OJ. Effect of prostaglandin F2alpha on length of diestrus [45] Onuma H, Ohnami Y. Retention of tubal eggs in mares. J Reprod Fertil Suppl
in mares. Prostaglandins 1972;2:265e74. 1975;(23):507e16.
[22] Allen WR, Rowson LEA. Control of the mare’s estrous cycle by prostaglandins. [46] Van Niekerk CH, Gerneke WH. Persistence and parthenogenetic cleavage of
J Reprod Fertil 1973;33:539e49. tubal ova in the mare. Onderstepoort J Vet Res 1966;31:195e204.
[23] Palmer E, Jousset B. Synchronization of estrus in mares with a prostaglandin [47] Vogelsang SG, Sorensen JD, Potter GD, Burns SJ, Kraemer DC. Fertility of donor
analogue and hCG. J Reprod Fertil Suppl 1975;(23):269e77. mares following non-surgical collection of embryos. J Reprod Fertil Suppl
[24] Hyland JH, Bristol F. Synchronization of estrus and timed insemination in 1979;(27):383e95.
mares. J Reprod Fertil Suppl 1979;(27):251e60. [48] Oguri N, Tsutsumi Y. Non-surgical transfer of equine embryos. Arch Androl
[25] Voss JL, Wallace RA, Squires EL, Pickett BW, Shideler BK. Effects of synchro- 1980;102:108. Abstract.
nization and frequency of insemination on fertility. J Reprod Fertil Suppl [49] Squires EL. Equine embryo transfer. Equine Vet Data 1980;1:201e12.
1979;(27):257e69. [50] Bain AM, Howey WP. Ovulation and transuterine migration of the conceptus
[26] Palmer E. Reproductive management of mares without detection of estrus. in Thoroughbred mares. J Reprod Fertil Suppl 1975;(23):541e50.
J Reprod Fertil Suppl 1979;(27):263e70. [51] Butterfield RM, Mathews RG. Ovulation and the movement of the conceptus
[27] Papkoff H. Chemical and biological properties of the subunits of pregnant in the first 35 days of pregnancy in thoroughbred mares. J Reprod Fertil Suppl
mare serum gonadotropin. Biochem Biophys Res Commun 1974;58:397e411. 1979;(27):447e56.
[28] Stewart F, Allen WR, Moor RM. Pregnant mare serum gonadotropin: ratio of [52] Allen WE, Newcombe JR. Relationship between early pregnancy site in
follicle stimulating hormone and luteinizing hormone activities measured by consecutive gestations in mares. Equine Vet J 1981;15:51e8.
radioreceptor assay. J Endocr 1977;71:371e88. [53] Ginther OJ. Mobility of the early equine conceptus. Theriogenology 1983;19:
[29] Cole HH, Hart GH. The potency of blood serum of mares in progressive stages 603e11.
of pregnancy in effecting the sexual maturity of the immature rat. Am J [54] Battut I, Colchen S, Fieni F, Tainturier D, Bruyas JF. Success rates when
Physiol 1930;93:57e64. attempting to nonsurgically collect equine embryos at 144,156 or 168 hours
[30] Allen WR. The immunological measurement of pregnant mare serum after ovulation. Equine Vet J Suppl 1997;(25):60e2.
gonadotrophin. J Endocr 1969;43:593e600. [55] McKinnon AO, Squires EL. Morphological assessment of the equine embryo.
[31] Allen WR. Immunological and endocrinological studies of endometrial cups J Am Vet Med Ass 1988;192:401e6.
during equine pregnancy; symposium in honour of EC amoroso. J Reprod [56] Carnevale EM, Ramirez RJ, Squires EL, Alvarenga MA, Vanderwall DK,
Fertil Suppl 1982;(31):57e94. McCue PM. Factors affecting pregnancy rates and early embryonic death after
[32] Stewart F, Allen WR. The binding of FSH, LH and PMSG to equine gonadal equine embryo transfer. Theriogenology 2000;54:965e79.
tissues. J Reprod Fertil Suppl 1979;(27):431e41. [57] Squires EL, Garcia RH, Ginther OJ. Factors affecting the success of equine
[33] Stewart F, Allen WR. Biological functions and receptor binding activities of embryo transfer. Equine Vet J Suppl 1985;(3):920e5.
equine chorionic gonadotropins. J Reprod Fertil 1981;62:527e39. [58] McKinnon AO, Squires EL. Equine embryo transfer. Vet Clin North Am Equine
[34] Douglas RH, Nuti L, Ginther OJ. Induction of ovulation and multiple ovulation Pract 1988;4:305e33.
in seasonally anovulatory mares with equine pituitary fractions. Ther- [59] Wilsher S, Allen WR. An improved method for non-surgical embryo transfer in
iogenology 1974;133:459e64. the mare. Equine Vet Edu 2004;16:39e44.
[35] Lapin DR, Ginther OJ. Induction of ovulation and multiple ovulations in [60] Cuervo-Arango J, Claes AN, Stout TAE. Effect of embryo transfer technique on
seasonally anovulatory and ovulatory mares with an equine pituitary extract. the likelihood of pregnancy in the mare: a comparison of conventional and
J Anim Sci 1977;44:834e41. Wilsher's forceps-assisted transfer. Vet Rec 2018;183:323.
[36] Douglas RH. Review of induction of superovulation and embryo transfer in the [61] Allen WR, Stewart F, Trouson AO, Tischner M, Bielanski W. Viability of horse
equine. Theriogenology 1979;11:33e44. embryos after storage and long-distance transport in the rabbit. J Reprod
[37] Woods GL, Ginther OJ. Pregnancy rate and incidence of multiple foetuses in mare Fertil 1976;47:387e94.
with induced multiple ovulations. J Reprod Fertil Suppl 1982;(32):415e21. [62] Willadsen SM, Polge EJC, Rowson LEA, Moor RM. Deep freezing of sheep
[38] Roser JF, Meyers-Brown G. Enhancing fertility in mares: recombinant equine embryos. J Reprod Fertil 1976;46:151e9.
gonadotropins. J Equine Vet Sci 2019;76:6e13. [63] Willadsen SM. A method for culture of micromanipulated sheep embryos and
[39] Carnevale EM, Beisner AE, McCue PM, Bass LD, Squires EL. Uterine changes its use to produce monozygotic twins. Nature (Lond) 1979;277:298e301.
associated with repeated inseminations and embryo collections in mares. Proc [64] Allen WR, Pashen RL. Production of monozygotic (identical) horse twins by
Am Ass Equine Practnrs 2005;51:202e3. embryo micromanipulation. J Reprod Fertil 1984;71:607e13.
[40] Campbell MLH. Embryo transfer in competition horses: managing mares and [65] Willadsen SM, Polge EJC. Attempts to produce monozygotic quadruplets in
expectations. Equine Vet Educ 2014;26:322e7. cattle by blastomere separation. Vet Rec 1981;108:211e4.
[41] Hamilton WJ, Day FT. Cleavage stages of the ova of the horse with notes on [66] Gray AP. Mammalian hybrids. Slough, UK: Commonwealth Agricultural
ovulation. J Anat 1945;79:127e36. Bureaux; 1971.
[42] Rowson LEA, Moor RM, Lawson RAS. Fertility following egg transfer in the [67] Allen WR, Short RV. Interspecific and extraspecific pregnancies in equid:
cow; effect of method, medium and synchronization of oestrus. J Reprod Fertil anything goes. J Hered 1997;88:384e92.
1969;18:511e23. [68] Allen WR, Skidmore JA, Stewart F, Antczak DF. Effects of fetal genotype and
[43] Sisson S, Grossman JD. The anatomy of domestic animals. Philadelphia, PA: uterine environment on placental development in equids. J Reprod Fertil
WB Saunders; 1953. 1993;97:55e60.