BIOLOGY OF REPRODUCTION 57, 165-171 (1997)
Effects of Equine Chorionic Gonadotropin, Human Chorionic Gonadotropin, and
Laparoscopic Artificial Insemination on Embryo, Endocrine, and
Luteal Characteristics in the Domestic Cat'
Terri L. Roth,2 Barbara A. Wolfe, Julie A. Long, JoGayle Howard, and David E. Wildt
Conservation & Research Center, National Zoological Park, Smithsonian Institution, Front Royal, Virginia 22630
ABSTRACT
The effects of gonadotropin treatment and laparoscopic ar-
tificial insemination (Al) on embryo quality, serum progesterone
and estradiol concentrations, and luteal progesterone content
were examined in the domestic cat. These data were compared
to similar historical data reported for naturally estrual, mated
queens. All queens in this study (n = 32) were treated with eCG
followed by 1) natural breeding (eCG-NB), 2) NB and hCG(eCG-NB-hCG), 3) NB and a sham Al procedure (eCG-NB-sham
Al), or 4) hCG and actual Al (eCG-hCG-AI). Queens ovulating
in response to treatment were ovariohysterectomized, and ovi-
ducts and uteri were flushed to collect embryos. Ovarian struc-
tures were recorded, corpora lutea (CL) were excised and eval-
uated for progesterone content, and serum was analyzed for
estradiol-1 7p and progesterone. Follicle and CL numbers ranged
from 0 to 28 and 2 to 42 per cat, respectively, and treatment
means did not differ (p > 0.05) among groups. Embryos were
recovered from oviducts and uterine horns in all treatment
groups, and recovery ranged from 60-96%. Mean embryo num-
ber per queen ranged from 8.2 2.6 to 23.2 ? 3.8 and did not
differ (p 0.05) among groups. However, the proportions of
unfertilized oocytes were greater (p < 0.05) for groups treated
with hCG and/or artificially inseminated, and the proportion of
blastocysts produced (31 of 107, 29.0%) was lower (p < 0.05)
in the eCG-hCG-AI group than for any other treatment (range,
59 of 116 [50.9%] to 67 of 116 [57.8%]). Not all queens in
each group produced good-quality embryos (eCG-NB, 5 of 5;
eCG-NB-hCG, 5 of 8; eCG-NB-sham Al, 2 of 5; and eCG-hCG-
Al, 3 of 6). Serum progesterone and estradiol-170, and total
luteal progesterone per ovary did not differ (p 0.05) among
treatments. Compared to historical controls (naturally estrual,
mated queens), eCG-NB queens produced > 4 times as many
good-quality embryos and blastocysts. Similarly, eCG-hCG-AI-
treated queens produced > 4 times the number of oocytes and
embryos, although a high proportion of these were poor quality
and did not develop to blastocysts. Together, these results indi-
cate that queens treated with eCG are capable of consistently
producing many good-quality embryos, at least half of which
develop to blastocysts in culture. These data support the use of
eCG in felids and suggest that other factors are responsible for
reduced pregnancy success and small litter sizes following as-
sisted reproduction.
INTRODUCTION
In 1992, domestic cats treated with eCG and hCG gave
birth to kittens following laparoscopic artificial insemina-
tion (AI) [1]. Since then, similar protocols have been ap-
Accepted March 10, 1997.
Received December 9, 1996.
'This research was supported, in part, by grant HD 23853 from the
National Institute of Child Health and Human Development, the Ralston
Purina Big Cat Survival Fund, the Philip Reed Foundation, Friends of the
National Zoo, and the Smithsonian Scholarly Studies Program.2Correspondence: Terri L. Roth, Center for Research of Endangered
Wildlife, Cincinnati Zoo and Botanical Garden, 3400 Vine Street, Cincin-
nati, OH 45220. FAX: (513) 569-8213.
plied to several nondomestic felid species with the goal of
eventually using assisted reproduction to facilitate conserv-
ing and managing endangered populations. On numerous
occasions, these efforts have been rewarded by the birth of
offspring [2-7]; however, pregnancies are not consistently
produced by AI in most felid species. The cheetah and do-
mestic cat are exceptions, with approximately 50% becom-
ing pregnant after AI [1, 8]. However, resulting litters are
smaller than those typically produced by naturally estrual,
mated females, despite numerous fresh ovulation sites on
the ovaries at AI.
In vitro fertilization (IVF) has proven useful for studying
gamete function and the influence of exogenous gonadotro-
pins among felids. IVF embryos have been produced using
oocytes collected from gonadotropin-treated females of
several felid species [9-14], and offspring occasionally
have resulted after the transfer of these embryos [9, 11, 12,
14]. However, again, pregnancy rates are low following IVF
and embryo transfer, and when pregnancies do result, litter
sizes are small despite transferral of many embryos into
recipients [9, 11, 12, 14]. These reports suggest that IVF-
generated embryos are not all developmentally competent,
an assertion partially supported by the failure of most IVF
embryos to develop to the blastocyst stage in vitro [15-18].
This in vitro developmental block has not been overcome
by modifying the culture system [15, 16, 18, 19] and is not
demonstrated by embryos collected from naturally estrual,
mated queens [17]. Together, AI and IVF/embryo transfer
results have led us to question the quality of oocytes pro-
duced by felids treated with exogenous gonadotropins.
Follicular development can be stimulated in cats with
exogenous FSH or eCG [9, 12, 20, 21], and ovulation can
be induced with GnRH or hCG [22]. For nondomestic fe-
lids, eCG followed by hCG has become the regimen of
choice to avoid animal stress associated with multiple FSH
injections. Of concern, however, are reports that eCG treat-
ment is associated with abnormal oocyte production [23-
25], poor-quality embryos [26-28], and/or low pregnancy
success [29, 30] in several domestic and laboratory species.
Recent data in cattle suggest that single FSH injections,
administered subcutaneously, are effective for inducing
multiple ovulations [31, 32]. Therefore, FSH could perhaps
substitute for eCG in felids, if in fact eCG is adversely
affecting oocyte quality and/or the maternal milieu during
early embryo development. However, eCG is not the only
factor potentially affecting oocyte quality and embryo de-
velopment. For example, high hCG dosages are associated
with increased oocyte degeneration and decreased IVF suc-
cess in the domestic cat [9]. It also is possible that anes-
thesia and/or the laparoscopic procedure itself may interfere
with successful fertilization and embryo development in
vivo.
It now is possible to study the effects of exogenous go-
nadotropins and laparoscopic AI on embryo, endocrine, lu-
165
ROTH ET AL.
teal, and uterine characteristics in the domestic cat because
comparative data are available in the naturally estrual, mat-
ed queen [17, 33-35]. The primary objective of this study
was to determine the proportion of developmentally com-
petent preimplantation embryos produced by eCG-treated
cats. Additionally, the effects of eCG alone, the combined
eCG-hCG regimen, and laparoscopic Al on embryo quality,
circulating reproductive steroids, and luteal progesterone
content were determined. These new data were compared
to similar historical data reported for the naturally estrual,
mated queen.
MATERIALS AND METHODS
Animals
Adult (1-3 yr old) female domestic cats (n = 32) were
housed in stainless steel cages (1-2 cats per cage) or in
communal pens (2-10 cats per pen). Two proven breeder
males used in previous studies [17, 33-35] were housed
singly in separate pens. All cats were provided a commer-
cial feline diet (Purina Cat Chow; Ralston-Purina Co., St.
Louis, MO) and water ad libitum and were maintained in
a controlled ambient environment under artificial fluores-
cent illumination (12L:12D daily) during the -2-yr study
period.
Treatment Groups
Individual queens were monitored daily for signs of be-
havioral estrus [17, 34]. Females not showing estrus for -
3 consecutive days were given an i.m. injection of 100 IU
eCG (Sigma Chemical Company, St. Louis, MO) and were
assigned randomly to one of four treatments: 1) natural
breeding 80 h later (eCG-NB), 2) NB 79 h later with 75
IU hCG (Sigma) injected i.m. 1 h after the first copulation
(eCG-NB-hCG), 3) NB 80 h later and sham AL (see below)
36-38 h after the first copulation (eCG-NB-sham AL), and
4) hCG 80 h later followed by AI 36-38 h after hCG (eCG-
hCG-AI). To minimize the number of animals in the study,
historical data reported for naturally estrual, mated queens
served as a nontreated control data set [17, 33-35].
Natural Breeding and Al
The natural breeding regimen was one that has proven
highly effective for inducing ovulation and producing em-
bryos in naturally estrual cats [17, 33]. Each queen was
mated three times a day at 3-h intervals for 2 days with
two proven breeder males on a rotating basis. Laparoscopic
AI was conducted according to previously described meth-
ods [1]. Briefly, 36-38 h after hCG injection or first cop-
ulation, queens were induced into a surgical plane of an-
esthesia with a ketamine hydrochloride (Vetalar; Parke-Da-
vis, Morris Plains, NJ) plus acepromazine maleate (Ayerst
Laboratories, Rouses Point, NY) mixture (10.0 mg/kg and
0.2 mg/kg BW, respectively, i.m.). Anesthesia was main-
tained by delivering isoflurane gas/oxygen via a face mask.
A 7-mm laparoscope was inserted -4 cm cranial to the
umbilicus and used to examine ovaries for the presence of
fresh corpora lutea (CL). Ovulating females were insemi-
nated by inserting a 20-g indwelling catheter (Sherwood
Medical, Tullamore, Ireland) percutaneously into the cranial
portion of each uterine horn and depositing washed sperm
(100 l/horn) directly into the uterine lumen [1, 2, 20].
Sham AL procedures were conducted in the same fashion
except that Ham's F-10 medium (HF10) containing no sper-
matozoa was deposited into the uterine lumen (100 1/
horn).
Semen Collection and Processing
The two males used for breeding also served as sperm
donors for Al. These cats were induced into a surgical plane
of anesthesia with tiletamine/zolazepan (Telazol; A.H. Rob-
ins Company, Richmond, VA; -4 mg/kg BW, i.m.), and
semen was collected by electroejaculation [36]. Semen was
diluted with an equal volume of HFO1 and centrifuged (150
x g; 8 min). After removal of the supernatant, the sperm
pellet was resuspended in 210 il of HF10 and maintained
in room atmosphere protected from light for 1-3 h until
used for AI. Immediately before AL, a 5-p1 aliquot was
examined for percentage motile spermatozoa. Using this
value along with the sperm concentration (determined using
a hemacytometer) and sample volume, the total number of
motile spermatozoa inseminated was calculated.
Ovariohysterectomy and Embryo Collection and Culture
A total of 26 ovulating queens were ovariohysterecto-
mized 144-148 h after the first copulation or hCG injection
at laparotomy under a surgical plane of anesthesia (as de-
scribed above for AL). Within 30 min of excision, oviducts
and uterine horns were flushed repeatedly with 1-5 ml of
warm (38?C) HF10 supplemented with 5% fetal calf serum
(Irvine Scientific, Princeton, NJ), 0.011 mg/ml pyruvate
(Sigma), and 0.284 mg/ml glutamine (Sigma), and the re-
covered fluid was examined for embryos and oocytes. Un-
cleaved oocytes were classified as unfertilized and were
discarded. Embryos were washed through two dishes con-
taining 2 ml of HF10, evaluated for developmental stage
and quality grade (grade 1: dark, homogenous coloration
and uniformly shaped blastomeres; grade 2: lighter in color,
some abnormally shaped blastomeres, slight vacuolation;
grade 3: degenerate, pale, fragmenting blastomeres), and
cultured in 100-1 HF10 drops under oil in a humidified
incubator (38?C; 5% CO 2 and air). In the naturally estrual,
mated cat, in vivo-produced embryos typically develop to
the morula stage by 144-148 h after the first copulation
[33]. Therefore, only embryos - 9-16 cells at recovery
and classified as grade 1 or 2 were considered "good qual-
ity." Embryos were assessed every 24 h for up to 8 days
(or until degeneration), and cleavage stage and blastocyst
formation were recorded.
Serum Estradiol-173 and Progesterone Concentration
Blood samples were collected via jugular venipuncture
from anesthetized queens immediately before laparoscopy,
and recovered serum was stored at -80?C until analyzed.
Estradiol-17 and progesterone concentrations were mea-
sured in unextracted serum using solid-phase 125I RIA kits
(Coat-a-Count; Diagnostic Products Corporation, Los An-
geles, CA). This assay has been standardized and tested
thoroughly for domestic cat serum [34]. All serum samples
were evaluated simultaneously in a single RIA for each
hormone. Assay sensitivities (based on 90% of maximum
binding) for estradiol-173 and progesterone were 5 pg/ml
and 0.05 ng/ml, respectively. The intraassay coefficients of
variation were < 10% for both assays.
Uterine and Ovarian Tissue Processing
For each queen, both ovaries were examined for the total
number of follicles (- 2 mm diameter) and CL. Arbitrarily,
166
eCG, hCG, AND AI EFFECTS ON CAT EMBRYO QUALITY
TABLE 1. Ovarian structures and embryo location and recovery from eCG-treated queens subjected to NB and/or Al, with or without hCG adminis-
tration.
No. of cats with embryos in the
Cats Corpora Follicles Oviducts and Recovery
Treatment group (n) lutea* (-2 mm)* Oviducts Uterus uterus (%)t
NBt 6 6.5 + 1.1 2.6 ? 1.6 1 5 0 56
eCG-NB 5 24.2 + 3.0 6.4 ? 5.5 2 2 1 96
eCG-NB-hCG 8 20.9 + 1.7 7.5 ? 1.6 2 3 3 70
eCG-NB-sham Al 5 13.6 + 4.1 4.4 ? 2.5 1 3 1 60
eCG-hCG-AI 6 27.2 + 5.6 8.7 ? 1.5 2 2 2 66
* Values are means SEM.
tRecovery (%) = total number of oocytes + embryos/CL number x 100.
tHistorical data on naturally estrual, mated queens adapted from previous reports [17, 33]; data not included in statistical analyses.
the right ovary was chosen for CL progesterone content
analysis. CL were excised from the ovary and bisected.
Each hemi-CL was weighed, snap-frozen in liquid nitrogen,
and stored at -80?C until analyzed. To quantify luteal pro-
gesterone concentrations (ng/mg), hemi-CL were processed
according to a previously described protocol [34]. Briefly,
CL were individually homogenized in 1 ml PBS. Homog-
enates were diluted with ethanol to 5 ml total volume,
boiled for 20 min, and centrifuged (500 x g, 20 min). After
the supernatants were decanted, residual luteal pellets were
extracted a second time in ethanol (2 ml) and centrifuged.
The supernatants were combined and dried under air. Fi-
nally, extracts were resuspended in 1 ml methanol, diluted
1:400 with PBS, and stored at -80 0C. Aliquots (100 1l) of
these diluted extracts were analyzed using a 125I solid-phase
RIA as described for serum. This assay has been validated
for quantifying cat luteal progesterone [34].
Uterine endometrial hyperplasia is associated with poor
embryo quality in domestic cats [35]. To ensure that it was
not a factor associated with poor embryo quality in this
study, uteri from all queens were examined histologically.
After each uterine horn was flushed with warm (38C), ster-
ile HFIO to recover embryos, uteri were placed in 50-ml
conical tubes containing 10% buffered formalin and later
processed for histological evaluation. Two transverse sec-
tions were sampled from each uterine horn, embedded in
paraffin, sectioned at 7 pim, and stained with hematoxylin
and eosin. Data for queens diagnosed with endometrial hy-
perplasia were discarded from the data set.
Statistical Analysis
Average values are reported as means + SEM. Numbers
of ovarian structures (follicles and CL), oocytes and em-
bryos collected, blastocysts produced, and all endocrine
data were compared among treatment groups by analysis
of variance and least significant difference mean compari-
sons [37]. The proportions of unfertilized oocytes, good-
quality embryos, and blastocysts produced per total embry-
os collected for each treatment group were compared by
chi-square. Correlation coefficients were calculated between
follicle number and serum estradiol-171 concentrations.
Additionally, CL number was correlated with total luteal
progesterone content and serum progesterone concentra-
tions. Finally, correlation coefficients were calculated for
serum endocrine and luteal traits in cats producing good-
versus poor-quality embryos.
RESULTS
Breeding, Ovarian Characteristics, and Embryo Recovery
Of the original 32 queens, eight were excluded from fur-
ther analysis for the following reasons: five estrual queens
scheduled for natural breeding refused to mate with males;
one failed to ovulate; one was diagnosed histologically with
severe endometrial hyperplasia; and the eighth was insem-
inated with < 2 x 106 motile spermatozoa, far fewer than
the number received by all other cats in the eCG-hCG-AI
group (range, 8-14 x 106).
Therefore, data were analyzed from a total of 24 queens
with a minimum of five per treatment group as follows:
eCG-NB, n = 5; eCG-NB-hCG, n = 8; eCG-NB-sham AI,
n = 5; eCG-hCG-AI, n = 6. For these groups, numbers of
follicles (- 2 mm) and CL at ovariohysterectomy ranged
from 0 to 28 and 2 to 42 per cat, respectively, and treatment
means did not differ (p 0.05) among groups (Table 1).
Embryo recovery (total number of oocytes plus embryos
divided by CL number and multiplied by 100) ranged from
60% to 96%, and embryos were recovered from the ovi-
ducts and uterine horns in all treatment groups (Table 1).
Embryo recovery and location were similar to those re-
ported for naturally estrual, mated queens in a previous
study [33].
Embryo Number, Quality, and Development to
Blastocysts
Mean numbers of oocytes plus embryos recovered per
female did not differ (p - 0.05) among treatment groups
(Table 2), primarily because of extreme variation among
cats within groups. For example, naturally bred eCG-treated
cats produced from 15 to 36 embryos and oocytes; cats
treated with the eCG-hCG combined regimen and artifi-
cially inseminated produced from 4 to 40 embryos and oo-
cytes. For all treatment groups, the number of embryos plus
oocytes recovered was more than twice that reported for
naturally estrual, mated queens [33].
Proportions of unfertilized oocytes were greater (p <
0.05) for cats given hCG and/or subjected to AI than for
those naturally bred (Fig. 1). The proportion of degenerate
embryos was greater (p < 0.05) in cats artificially insem-
inated after eCG-hCG administration than in those that
were naturally bred after eCG treatment (Fig. 1). Cats in
the hCG-AI group also produced the fewest (p < 0.05)
morulae (Fig. 1). Furthermore, the proportions of good-
quality embryos produced by eCG-treated, naturally bred
queens (with or without hCG) were greater (p < 0.05) than
for cats given eCG and hCG and then artificially insemi-
nated (Table 2). Although oocytes and/or embryos were
collected from all 24 eCG-treated queens, not all cats in
each group produced good-quality embryos except those
that were naturally bred without hCG administration (5 of
5). In contrast, 5 of 8, 2 of 5, and 3 of 6 cats in the eCG-
NB-hCG, eCG-NB-sham AI, and eCG-hCG-AI groups, re-
spectively, produced good-quality embryos. Mean numbers
167
ROTH ET AL.
TABLE 2. Embryo quality and development after collection from eCG-treated queens subjected to NB and/or Al, with or without hCG administration.
Good-quality No. embryos Blastocysts
No. embryos embryos/total developing formed/total
Treatment group recovered/cat* collected (%)* to blastocysts* embryos collected (%)
NB* 3.7 + 0.7 18/22 (81.8) 2.2 + 0.7 13/22 (59.1)
eCG-NB 23.2 + 3.8 81/116 (69.8)a 11.8 + 1.8 59/116 (50.9)'
eCG-NB-hCG 14.5 + 3.3 73/116 (62.9)a 8.4 + 3.2 67/116 (57.8)'
eCG-NB-sham Al 8.2 + 2.6 22/41 (53.7)ahb 4.4 + 2.9 22/41 (53.7)'
eCG-hCG-AI 17.8 + 5.6 37/107 (34.6)h 4.8 + 3.2 31/107 (29.0)b
* Values are means SEM.
t Only embryos -9-16 cells at recovery and classified as grade 1 or 2 were assigned "good-quality" status.
* Historical data on naturally estrual, mated queens adapted from previous reports [17, 331; data not included in statistical analyses.
f.b Within columns, values with different superscripts differ (p < 0.05).
of embryos developing to blastocysts after collection did
not differ among groups (p < 0.05), but proportionally few-
er (p < 0.05) blastocysts were produced in the eCG-h-
CG-AI group compared to any other group (Table 2). Still,
even in this group, the mean number of blastocysts was
twice that reported for naturally estrual, mated controls
[17, 33].
Endocrine Characteristics
Serum progesterone and estradiol-173 concentrations did
not differ (p - 0.05) among treatments (Table 3), again
largely due to variation among individuals, and were not
unlike those reported previously for naturally estrual, mated
queens [34]. Luteal progesterone concentration (ng/mg)
was less (p < 0.05) for cats naturally bred and given hCG
than for the other three groups. However, luteal mass also
was high in these individuals, so the total luteal progester-
one per ovary did not differ (p - 0.05) among groups.
When data were analyzed for cats across all treatments, CL
number was positively correlated (p < 0.01) with serum
progesterone concentration (r = 0.77) and total luteal pro-
gesterone (r = 0.81). Similarly, serum progesterone con-
centration was highly correlated (p < 0.01) with total luteal
progesterone (r = 0.89). There was no significant relation-
ship (p - 0.05) between serum estradiol-17P concentration
and follicle number. Across all groups, there were no cor-
relations (p - 0.05) between serum endocrine or luteal
traits and queens producing poor-quality embryos.
DISCUSSION
The general hypothesis of this study was that the exog-
enous gonadotropin eCG is responsible for recruiting poor-
quality oocytes and/or producing poor-quality embryos that
ultimately result in reduced pregnancy success and small
litter sizes following assisted reproduction in the domestic
cat. Results clearly allowed rejecting this hypothesis.
Queens that were treated with eCG and naturally bred con-
sistently produced many good-quality embryos, at least half
of which were capable of developing into blastocysts in
vitro. These findings were comparable to those reported in
an extensive recent study that examined embryo develop-
ment and quality in naturally estrual cats (no eCG) mated
using the same copulatory regimen described here [17, 33-
35].
Mean CL and embryo numbers for all treatment groups
were at least twice those reported for naturally estrual, mat-
ed queens ovariohysterectomized at the same stage of preg-
nancy in a previous study [33]. Additionally, more blasto-
cysts were produced in all treatment groups as compared
to the number reported for the naturally estrual, mated cat
FIG. 1. Within treatment groups, percent-
age of recovered embryos at each devel-
opmental stage. Different superscripts de-
note differences among treatments within
developmental stage (p < 0.05)
168
eCG, hCG, AND Al EFFECTS ON CAT EMBRYO QUALITY
TABLE 3. Serum and luteal endocrine traits for eCG-treated queens subjected to NB and/or Al, with or without hCG administration (means + SEM).
Serum Serum Luteal Total luteal
Treatment progesterone estradiol progesterone/ Luteal progesterone/
group (ng/ml) (pg/ml) CL (ng/mg) mass (mg) ovary (mg)
NB* 27.7 + 2.8 18.3 + 1.8 174.6 + 11.0 16.8 + 1.1 29.8 _ 2.6
eCG-NB 46.4 + 12.2 7.3 ? 1.5 157.5 + 8.8 a 15.4 + 0.9' 23.2 + 5.4
eCG-NB-hCG 32.9 + 2.9 23.9 + 9.4 92.5 + 4.9b 19.0 + 0.9b 16.4 - 2.9
eCG-NB-sham Al 25.8 + 5.2 5.7 + 0.4 140.4 + 11.6a 16.7 + 1.2,b 16.6 + 3.6
eCG-hCG-AI 51.3 + 12.7 18.1 + 2.3 153.3 + 7.0a 15.3 + 0.8a 36.2 + 10.8
* Historical data on naturally estrual, mated queens adapted from a previous report [341; data not included in statistical analyses.
a,b Within columns, values with different superscripts differ (p < 0.05).
[17, 33]. The greatest difference was observed in the eCG-NB
group; queens in this group produced > 4 times as many
blastocysts as the untreated, historical controls. However,
in some species, the negative impact of exogenous gonad-
otropins becomes most apparent later in pregnancy. For ex-
ample, eCG is associated with implantation failure and re-
tarded fetal growth in mice and hamsters, ultimately re-
sulting in few live fetuses [27, 38].
When studying eCG effects on pregnancy success in the
cat, it may be more appropriate to examine the proportion
of queens producing blastocysts than the overall average
number of blastocysts produced. All of the eCG-treated cats
that were naturally mated produced embryos that developed
into blastocysts, in contrast to all other groups, in which at
least 40% of females failed to produce blastocysts. Of the
cats treated with the standardized AI protocol
(eCG+hCG+AI), half produced blastocysts-exactly the
incidence of pregnancy reported after laparoscopic AI in
this species [1]. Individual variation in response to exoge-
nous gonadotropins is not surprising, since timing of go-
nadotropin administration relative to natural, cyclic fluctu-
ations in ovarian activity can markedly alter follicular and
ovulatory response [26, 39]. In most species, variation is
minimized by synchronizing the estrous cycle prior to go-
nadotropin treatment, but luteal regulation is poorly under-
stood in the cat and conventional estrous synchronization
protocols have not been developed. In any case, ovarian
responses of queens given eCG and then naturally bred
were consistently high, suggesting that some other factor(s)
must be responsible for ovarian/embryo variations observed
across treatment groups. Candidates include hCG, anesthe-
sia, and/or the AI procedure itself.
The impact of hCG on embryo quality and/or pregnancy
success appears to be species specific. In humans, hCG has
been reported to increase the number of oocytes, transfer-
able embryos, and resulting pregnancies [40]. However,
compared to naturally mated mice, hCG-treated counter-
parts produce more abnormal embryos and experience
higher postimplantation mortality [27]. In the rabbit, an in-
duced ovulator like the cat, hCG increases the proportion
of poor-quality and degenerate or fragmenting embryos
[41, 42]. Information in the cat is limited, but in one study,
high hCG dosages were associated with increased numbers
of degenerate oocytes [9]. In the present study, eCG-treated
queens that were naturally bred and given hCG produced
more unfertilized oocytes than naturally bred queens re-
ceiving no hCG. Nonetheless, hCG treatment had no neg-
ative effect on the overall number of morulae produced.
Furthermore, more than half of these embryos developed
to blastocysts in culture, almost double the number of blas-
tocysts produced by either of the two groups subjected to
AI. Therefore, these data did not suggest that hCG is pri-
marily responsible for poor embryo quality in gonadotro-
pin-treated queens.
Of the queens subjected to sham or actual AI, only
-45% (5 of 11) produced embryos that formed blastocysts
compared to -77% (10 of 13) of naturally bred cats. These
data suggested that the AI procedure itself may have ad-
versely affected embryo quality. Laparoscopic AI was de-
veloped in cats because transcervical AI is relatively un-
successful, presumably due to compromised sperm trans-
port [1]. Furthermore, postovulatory laparoscopic intrauter-
ine AI is necessary in the cat because anesthesia inhibits
ovulation [1]. To confirm the presence of fresh CL, a Verres
needle is used to elevate and gently remove the fimbria
from the ovarian surface before AI commences. Manipu-
lating the reproductive tract at this sensitive time in the
periovulatory interval could disrupt oocyte capture by the
fimbria. In the rat, ovarian bursa displacement results in
significant loss of oocytes [43]. Additionally, laparoscopic
intrauterine Al in the ewe has been suspected of adversely
affecting subsequent embryo quality and recovery [44], but
primarily as a result of inappropriate timing relative to ovu-
lation [44, 45] and not directly related to "procedure." The
actual disruption of ovum fertilizability resulting from la-
paroscopic AI seems unlikely considering that similar tech-
niques are routinely successful (> 70% pregnancies) in fer-
rets and deer [46, 47].
It also is useful to consider the effects of the AI proce-
dure in the context of hCG treatment. Previous studies in
the naturally estrual cat have demonstrated that an endog-
enous LH surge peaks -9 h after the first of a series of
matings [48]. However, there is considerable individual
variation, and ovulation can occur 24-64 h after first cop-
ulation [33, 48, 49]. The breeding schedule used in the pres-
ent study ensured that sperm were present in the reproduc-
tive tract 28 h before ovulation began, with the last natural
insemination occurring after (or as many as 34 h before)
ovulation. In eCG-treated cats that were naturally mated,
follicular development was initiated by eCG, whereas final
oocyte maturation and ovulation presumably were caused
by mating-induced endogenous LH surges. It seems logical
that the sequence of events in these females (LH surge,
ovulation, fertilization) was similar to that for naturally es-
trual, mated queens [33, 48, 49], thereby explaining the
similarity in good-quality embryo production. Superimpos-
ing a sham AI procedure on the eCG-NB regimen resulted
in more unfertilized oocytes but a comparable proportion
of recovered morulae. The increased number of unfertilized
oocytes perhaps resulted from anesthesia and reproductive
tract manipulation during the final hours of ovulation, just
as fertilization commenced.
Queens given hCG after the onset of natural breeding
again produced more unfertilized oocytes than queens that
169
ROTH ET AL.
mated but did not receive hCG. These results may be ex-
plained, in part, by the recent confirmation that hCG alone
induces both follicular development and ovulation in the
cat [50]. Recovered embryos likely originated from eCG-
induced follicles that ovulated in response to hCG and/or
NB, whereas the unfertilized oocytes possibly were pro-
duced by hCG-induced secondary follicles that may have
ovulated later but failed to fertilize in the absence of viable
sperm. This theory is supported by the lower luteal
progesterone/CL in this group; CL formed from the sec-
ondary ovulating follicles would have been less mature at
ovariohysterectomy.
Finally, cats given eCG and hCG and then artificially
inseminated produced a high proportion of unfertilized oo-
cytes and fragmenting or degenerate embryos. For this
group, the timing of ovulation relative to laparoscopy and
sperm deposition is critical. It is possible that 1) some of
the queens ovulated early and there were no sperm present
to fertilize the oocytes; 2) queens ovulated at the time of
AI and oocytes were somehow disrupted during manipu-
lation of the fimbria and reproductive tract; or 3) some of
the follicles did not rupture prior to AI, and anesthesia de-
layed ovulation of these follicles beyond the functional lon-
gevity of in vitro-processed, inseminated sperm. Although
we have not determined exactly why queens in the eCG-
hCG-AI group produced many poor-quality embryos, it
seems reasonable that the problem stems from an inability
to consistently ensure that fertilizable oocytes and fully
functional sperm are present in the reproductive tract si-
multaneously.
Aberrant serum estradiol levels in rats and cats following
exogenous gonadotropin treatment have been implicated in
poor embryo recovery and development, respectively [51,
52]. Although there were numerous follicles (- 2 mm) on
the ovaries of queens in all groups, circulating estradiol
concentrations at ovariohysterectomy were within the range
of those reported for naturally estrual, mated queens at the
same stage of pregnancy [34]. Percentage embryo recovery(based upon CL number) was equivalent to, or better than,
the 56% recovery rate reported for naturally estrual, mated
queens [33]. These two findings are important for the fol-
lowing reasons. First, recovery of 70-96% of embryos pro-
duced by queens in the eCG-hCG-NB and eCG-NB groups,
respectively, provided strong evidence against the hypoth-
esis that follicular luteinization in the absence of oocyte
release occurred, a finding reported after gonadotropin
treatment in other species [53, 54]. Second, these findings
clearly indicated that embryos were not being lost due to
accelerated transport through the reproductive tracts of go-
nadotropin-treated females, a phenomenon reported in rats
[51]. In fact, the time of ovariohysterectomy was specifi-
cally chosen because it coincided with embryo transport
through the uterotubal junction in naturally estrual and mat-
ed cats [33]. Because we found embryos in the oviducts
and uterine horns in all treatment groups, it appeared that
the various treatments were allowing embryo progression
through the reproductive tract at a rate mimicking that for
naturally estrual, mated queens.
Luteal insufficiency following gonadotropin treatment is
recognized as one cause of pregnancy failure in humans
after assisted reproduction [55]. In contrast, hyperelevated
circulating progesterone has been associated with poor em-
bryo quality in cats [34]. Serum progesterone concentra-
tions for queens after all four treatments reported here were
variable but generally similar to those for naturally estrual,
mated queens [34]. Furthermore, mean luteal mass and pro-
gesterone concentrations were similar to those reported for
the naturally estrual, mated queen [34], and no aberrant
serum endocrine or luteal traits were found associated with
queens producing poor-quality embryos. Therefore, it ap-
peared that luteal function, at least during preimplantation
embryo development, was not compromised in queens re-
ceiving eCG alone or in conjunction with hCG. Rather, do-
mestic cat embryos appeared relatively tolerant to a variety
of endocrine environments, at least during the early preim-
plantation stages of development.
Overall, these results were important for two primary
reasons. First, data from eCG-treated cats that were natu-
rally bred with or without receiving supplemental hCG pro-
duced fertilizable oocytes that were developmentally com-
petent to the blastocyst stage. Therefore, the partial morula-
to-blastocyst developmental block routinely encountered in
this species after IVF [15-18] is likely the result of inap-
propriate culture conditions during fertilization or early em-
bryo development. Second, these results suggested that
some other factor, perhaps the timing of insemination or an
unknown component of the AI technique, is responsible for
reduced pregnancy success and small litter sizes following
AI. In any case, these data support the use of eCG as an
ovarian-stimulating gonadotropin for assisted reproduction
in felids.
ACKNOWLEDGMENTS
The authors thank Jennifer Buff, Rachael Weiss, and Lena May Bush
for technical support. The authors also thank Laura Graham, Dr. Bill
Swanson, and Dr. Janine Brown for advice and assistance with the endo-
crine analyses.
REFERENCES
1. Howard JG, Barone MA, Donoghue AM, Wildt DE. The effect of
pre-ovulatory anaesthesia on ovulation in laparoscopically inseminat-
ed domestic cats. J Reprod Fertil 1992; 96:175-186.
2. Howard JG, Donoghue AM, Barone MA. Goodrowe KL, Blumer E,
Snodgrass K, Starnes D, Tucker M, Bush M, Wildt DE. Successful
induction of ovanan activity and laparoscopic Intrauterine artificial
insemination in the cheetah (Acnonyx jubatus). J Zoo Wildl Med
1992; 23:288-300.
3. Howard JG, Byers AP, Brown JL, Barrett SJ, Evans MZ, Schwartz
RJ, Wildt DE. Successful ovulation induction and laparoscopic intra-
uterine artificial insemination in the clouded leopard (Neofelis nebu-
losa). Zoo Biol 1996; 15:55-70.
4. Donoghue AM, Johnson LA, Armstrong DL, Simmons LG, Wildt DE.
Birth of a Siberian tiger cub (Panthera tigrs altaicia) following la-
paroscopic intrauterine artificial insemination. J Zoo Wildl Med 1993:
24:185-189.
5. Swanson WE Howard JG. Roth TL, Brown JL, Alvarado T, Burton
M, Starnes D, Wildt DE. Responsiveness of ovaries to exogenous
gonadotrophins and laparoscopic artificial insemination with frozen-
thawed spermatozoa in ocelots (Felis pardalis). J Reprod Fertil 1996;
106:87-94.
6. Barone, MA, Wildt DE, Byers AP, Roelke ME, Glass CM, Howard
JG. Gonadotrophin dose and timing of anaesthesia for laparoscopic
artificial insemnation in the puma (Fells concolor). J Reprod Fertil
1994; 101:103-108.
7. Roth TL, Armstrong DL, Bame MT, Wildt DE. Seasonal effects on
ovarian responsiveness to exogenous gonadotropins and successful ar-
tificial insemination in the snow leopard (Panthera uncia). Reprod
Fertil Dev 1997; 9:(in press).
8 Howard JG, Barone M, Byers A, Roth T. Wildt D. Ovulation induction
sensitivity and laparoscopic intrauterine insemination in the cheetah,
puma and clouded leopard. J Androl 1993; (suppl 1): (abstract P-55).
9. Goodrowe KL, Wall RJ, O'Brien SJ, Schmidt PM, Wildt DE. Devel-
opmental competence of domestic cat follicular oocytes after fertiliza-
tion in vitro. Biol Reprod 1988; 39:355-372.
10. Goodrowe KL, Miller AM, Wildt DE. In vitro fertilization of gonad-
otropin-stimulated leopard cat (Felis bengalensis) follicular oocytes. J
Exp Zool 1989; 252:89-95.
170
eCG, hCG, AND AI EFFECTS ON CAT EMBRYO QUALITY
11. Donoghue AM, Johnston LA, Seal US, Armstrong DL, Tilson RL,
Wolff P. Petrini K, Simmons LG, Gross T, Wildt DE. In vitro fertil-
ization and embryo development in vitro and in vivo in the tiger
(Panthera tigris). Biol Reprod 1990; 43:733-744.
12. Pope CE, Keller GL, Dresser BL. In vitro fertilization in domestic
and non-domestic cats including sequences of early nuclear events,
development in vitro, cryopreservation and successful intra- and in-
terspecies embryo transfer. J Reprod Fertil Suppl 1993; 47:189-201.
13. Roth TL, Howard JG, Donoghue AM, Swanson WF, Wildt DE. Func-
tion and culture requirements of snow leopard (Panthera uncia) sper-
matozoa in vitro. J Reprod Fertil 1994; 101:563-569.
14. Swanson WF, Godke RA. Transcervical embryo transfer in the do-
mestic cat. Lab Anim Sci 1994; 44:288-291.
15. Johnston LA, Donoghue AM, O'Brien SJ, Wildt DE. Culture medium
and protein supplementation influence in vitro fertilization and em-
bryo development in the domestic cat. J Exp Zool 1991; 257:350-
359.
16. Johnston LA, Donoghue AM, O'Brien SJ, Wildt DE. Influence of
temperature and gas atmosphere on in vitro fertilization and embryo
development in domestic cats. J Reprod Fertil 1991; 92:377-382.
17. Roth TL, Swanson WE Wildt DE. Developmental competence of do-
mestic cat embryos fertilized in vivo versus in vitro. Biol Reprod
1994; 51:441-451.
18. Roth TL, Donoghue AM, Byers AP. Munson L, Wildt DE. Influence
of ovlductal cell monolayer co-culture and the presence of corpora
hemorrhagica at the time of oocyte aspiration on gamete interaction
in vitro in the domestic cat. J Assist Reprod Genet 1994; 10:523-529.
19. Swanson WF, Roth TL, Godke RA. Persistence of the developmental
block of in vitro fertilized domestic cat embryos to temporal variations
in culture conditions. Mol Reprod Dev 1996; 43:298-305.
20. Phillips LG, Simmons LG, Bush M, Howard JG, Wildt DE. Gonad-
otropin regimen for inducing ovarian activity in captive-wild felids. J
Am Vet Med Assoc 1982; 181:1246-1250.
21. Wildt DE, Kinney GM, Seager SWJ. Gonadotropin-induced repro-
ductive cyclicity in the domestic cat. Lab Anim Sci 1978; 28:301-
307.
22. Goodrowe KL, Wildt DE. Ovarian response to human chorionic go-
nadotropin or gonadotropin releasing hormone in cats in natural or
induced estrus. Theriogenology 1987; 27:811-817.
23. Moor RM, Osborn JC, Crosby IM. Gonadotrophin-induced abnor-
malities in sheep oocytes after superovulation. J Reprod Fertil 1985,
74:167-172.
24. Yun YW, Yuen BH, Moon YS. Effects of superovulatory doses of
pregnant mare serum gonadotropin on oocyte quality and ovulatory
and steroid hormone responses in rats. Gamete Res 1987; 16:109-
120.
25. Yun YW, Yu FH, Yuen BH, Moon YS. Effects of a superovulatory
dose of pregnant mare serum gonadotropin on follicular steroid con-
tents and oocyte maturation in rats. Gamete Res 1989; 23:289-298.
26. Goulding D, Williams DH, Roche JF, Boland MP. Factors affecting
superovulation in heifers treated with PMSG. Theriogenology 1996;
45:765-773.
27. Ertzeid G, Storeng R. Adverse effects of gonadotrophin treatment on
pre- and postimplantation development in mice. J Reprod Fertil 1992:
96:649-655.
28. Armstrong DT. Recent advances in superovulation of cattle. Theri-
ogenology 1993; 39:7-24.
29. Elsden RP, Nelson LD, Seidel GE Jr. Superovulating cows with fol-
licle stimulating hormone and pregnant mare's serum gonadotrophin.
Theriogenology 1978; 9:17-26.
30. Kusuma PSH, Tainturier D. Comparison of induction of oestrus in
dogs using metergoline, metergoline plus human chorionic gonadotro-
phin, or pregnant mares' serum gonadotrophin. J Reprod Fertil Suppl
1993; 47:363-370.
31. Takedomi T, Aoyagi Y, Konishi M, Kishi H, Taya K, Watanabe G,
Sasamoto S. Superovulation in Holstein heifers by a single injection
of porcine FSH dissolved in polyvinylpyrrolidone. Thenogenology
1993; 39:327 (abstract).
32. Fernandez M, Sanchez L, Alvarez FS, Vazquez C, Iglesias A. Super-
ovulation in Rubia Gallega cows with a single subcutaneous injection
of FSH. Theriogenology 1993; 39:217 (abstract).
33. Swanson WE Roth TL, Wildt DE. In vivo embryogenesis, embryo
migration and embryonic mortality in the domestic cat. Biol Reprod
1994; 51:452-464.
34. Swanson WE Roth TL, Brown JL, Wildt DE. Relationship of circu-
lating steroid hormones, luteal luteinizing hormone receptor and pro-
gesterone concentration, and embryonic mortality during early em-
bryogenesis in the domestic cat. Biol Reprod 1995; 53:1022-1029.
35. Roth TL, Munson L, Swanson WE, Wildt DE. Histological character-
istics of the uterine endometrium and corpus luteum during early em-
bryogenesis and the relationship to embryonic mortality in the do-
mestic cat. Biol Reprod 1995; 53:1012-1021.
36. Howard JG, Bush M, Wildt DE. Semen collection, analysis, and cry-
opreservatlon in nondomestic mammals. In: Morrow DA (ed.), Cur-
rent Therapy in Theriogenology. Philadelphia: WB Saunders Co;
1986: 1047-1053.
37. Steel RGD, Torrie JH. In: Napier C, Maisel JW (eds.), Principles and
Procedures of Statistics. New York: McGraw-Hill Book Company;
1980.
38. McKiernan SH, Bavister BD. Effect of super-stimulation with preg-
nant mare's serum gonadotropin (PMSG) on hamster 1-cell embryo
pre- and post-implantation development. Biol Reprod 1996; 54(suppl
1):170 (abstract 454).
39. Boland MP, Goulding D, Roche JE Alternative gonadotrophins for
superovulatlon in cattle. Theriogenology 1991; 35:5-17.
40. Wang TA, Armant DR, Taymor ML, Seibel MM. The influence of
exogenous human chorionic gonadotropin cycles with spontaneous lu-
teinizing hormone surges on the outcome of in vitro fertilization. Fertil
Steril 1987; 48:613-616.
41. Peinado JA, Molina I, Pla M, Tresguerres JAE Romeu A. Recombi-
nant-human luteinizing hormone (r-hLH) as ovulatory stimulus in su-
perovulated does. J Assist Reprod Genet 1995; 12:61-64.
42. Molina I, Pla M, Vicente JS, Martin A, Romeu A. Induction of ovu-
lation in rabbits with pure urinary luteinizing hormone and human
chorionic gonadotrophin: comparison of oocyte and embryo quality.
Hum Reprod 1991; 6:1449-1452.
43. Vanderhyden BC, Rouleau A, Armstrong DT. Effect of removal of
the ovarian bursa of the rat on infundibular retrieval and subsequent
development of ovulated oocytes. J Reprod Fertil 1986; 77:393-399.
44. Robinson JJ, Wallace JM, Aitken RE Fertilization and ovum recovery
rates in superovulated ewes following cervical insemination or lapa-
roscopic intrauterine insemination at different times after progestogen
withdrawal and in one or both uterine horns. J Reprod Fertil 1989;
87:771-782.
45. Maxwell WMC. Artificial insemination of ewes with frozen-thawed
semen at a synchronised oestrus. 1. Effect of time of onset of oestrus,
ovulation and insemination on fertility. Anim Reprod Sci 1986; 10:
301-308.
46. Howard JG, Bush M, Morton C, Morton E Wildt DE. Comparative
semen cryopreservation in ferrets (Mustela putorius furo) and preg-
nancies after laparoscopic intrauterine insemination with frozen-
thawed spermatozoa. J Reprod Fertil 1990; 92:109-118.
47. Monfort SL, Asher GW, Wildt DE, Wood TC, Schiewe MC, William-
son LR, Bush M, Rall WE. Successful intrauterine insemination of
Eld's deer (Cervus eldi thamin) with frozen-thawed spermatozoa. J
Reprod Fertil 1993; 99:481-488.
48. Wildt DE, Chan SYW, Seager SWJ, Chakraborty PK. Ovarian activity,
circulating hormones and sexual behavior in the cat. I. Relationships
during the coitus-induced luteal phase and the estrous period without
mating. Biol Reprod 1981; 25:15-28.
49. Shille VM, Munro C, Farmer SW, Papkoff H, Stabenfeldt GH. Ovar-
ian and endocrine responses in the cat after coitus. J Reprod Fertil
1983; 68:29-39.
50. Swanson WE Wolfe BA, Brown JL, Roth TL, Wildt DE. Pharmaco-
kinetics and ovarian stimulatory effects of exogenous gonadotropins
administered singly and in combination in the domestic cat. Biol Re-
prod 1996; 54(suppl 1):189 (abstract 530).
51. Akira S, Sanbuissho A, Lin YC, Araki T. Acceleration of embryo
transport in superovulated adult rats. Life Sci 1993; 53:1243-1251.
52. Goodrowe KL, Howard JG, Wildt DE. Comparison of embryo recov-
ery, embryo quality, oestradiol-173 and progesterone profiles in do-
mestic cats (Felis cactus) at natural or induced oestrus. J Reprod Fertil
1988; 82:553-561.
53. Dieleman SJ, Bevers MM. Effects of monoclonal antibody against
PMSG administered shortly after the preovulatory follicles. Anim Re-
prod Sci 1987; 15:37-52.
54. Westfahl PK. Comparison of luteinized unruptured follicles and cor-
pora lutea: steroid hormone production and response to luteolytic and
luteotropic agents. Biol Reprod 1993; 48:807-814.
55. Babalioglu R, Varol FG, Ilhan R, Yalcin O, Cizmecioglu E Proges-
terone profiles in luteal-phase defects associated with recurrent spon-
taneous abortions. J Assist Reprod Genet 1996; 13:306-309.
171