Content uploaded by Kevin D Sinclair
Author content
All content in this area was uploaded by Kevin D Sinclair on Nov 06, 2014
Content may be subject to copyright.
1624
BIOLOGY OF REPRODUCTION 64, 1624–1632 (2001)
Effect of Dietary Energy and Protein on Bovine Follicular Dynamics
and Embryo Production In Vitro: Associations with the Ovarian Insulin-Like
Growth Factor System
1
D.G. Armstrong,
2,3
T.G. McEvoy,
4
G. Baxter,
3
J.J. Robinson,
4
C.O. Hogg,
3
K.J. Woad,
3
R. Webb,
5
and K.D. Sinclair
4
Division of Integrative Biology,
3
Roslin Institute (Edinburgh), Roslin, Midlothian EH25 9PS, United Kingdom
Department of Applied Physiology,
4
Scottish Agricultural College, Craibstone Estate, Bucksburn,
Aberdeen AB21 9YA, United Kingdom
University of Nottingham,
5
Division of Agriculture and Horticulture, School of Biosciences, Loughborough,
Leicestershire LE12 5RD, United Kingdom
ABSTRACT
Heifers were assigned either low or high (HE) levels of energy
intake and low or high concentrations of dietary crude protein.
The effect of these diets on the plasma concentrations of insulin,
insulin-like growth factor (IGF)-I, and urea on follicular growth
and early embryo development is described. We propose that
the observed dietary-induced changes in the ovarian IGF system
increase bioavailability of intrafollicular IGF, thus increasing the
sensitivity of follicles to FSH. These changes, in combination
with increased peripheral concentrations of insulin and IGF-I in
heifers offered the HE diet, contribute to the observed increase
in growth rate of the dominant follicle. In contrast to follicular
growth, increased nutrient supply decreased oocyte quality, due
in part to increased plasma urea concentrations. Clearly a num-
ber of mechanisms are involved in mediating the effects of di-
etary energy and protein on ovarian function, and the formu-
lation of diets designed to optimize cattle fertility must consider
the divergent effects of nutrient supply on follicular growth and
oocyte quality.
fertilization, follicle, follicular development, gene regulation,
granulosa cells, growth factors, IGF receptor, insulin, IVF/ART,
oocyte development, ovary, ovum, theca cells
INTRODUCTION
Nutritional status is a major factor controlling fertility in
cattle [1–3]. Poor nutrition results in delayed puberty, ab-
errant estrous cycles, lowered conception rates, and reduced
birth weight. Specific effects of undernutrition on ovarian
follicular growth have also been reported. For example the
diameters of preovulatory follicles have been negatively
correlated to weight loss in Bos indicus [4], and in post-
partum dairy cattle the extent of the energy balance deficit
influences follicular growth [5]. Details of mechanisms in-
volved in these changes remain to be characterized fully.
Recently, short-term changes in the plane of nutrition
have been shown to affect ovarian follicular dynamics in
1
Ministry of Agriculture Fisheries and Food (contract DSO206) and Bio-
technology and Biological Sciences Research Council supported this
study. Scottish Agricultural Colleges receives funding from the Scottish
Executive Rural Affairs Department. The Medical Research Council fund-
ed K.J.W.
2
Correspondence: FAX: 0131 440 0434; e-mail: david.armstrong@bbsrc.ac.uk
Received: 19 July 2000.
First decision: 25 August 2000.
Accepted: 16 January 2001.
Q2001 by the Society for the Study of Reproduction, Inc.
ISSN: 0006-3363. http://www.biolreprod.org
cattle without any changes in the circulating concentrations
of gonadotropin [6]. This is in direct contrast to the situa-
tion in monogastric species, such as primates, that show
significant changes in LH pulse frequency in response to
short-term changes in nutrient intake [7]. In the case of
ruminants, it is hypothesized that nutritionally induced
changes in circulating metabolic hormones act directly on
the ovary to bring about the observed changes in patterns
of follicular growth [6]. Supporting this hypothesis is the
observation that insulin can influence ovulation rate in en-
ergy-deprived beef heifers [8] and ewes [9]. Moreover,
treatment of lactating cows with growth hormone (GH) in-
creases the number of small and medium-sized follicles [10,
11], an effect most likely due to GH-induced changes in
the circulating concentrations of insulin-like growth factor
(IGF)-I and/or insulin. The observations that insulin and
IGF-I have potent effects on cultured granulosa cells from
cattle [12] and sheep [13, 14] provide further evidence sup-
porting this hypothesis.
Endocrine and metabolic signals that regulate follicular
growth also are expected to influence oocyte development
either through changes in hormone/growth factor concen-
trations in follicular fluid or via granulosa-oocyte interac-
tion [15]. For example, as well as regulating follicular
growth [6], short-term changes in dietary energy intake in-
fluence both oocyte morphology and developmental poten-
tial [16, 17]. In addition, high levels of highly degradable
protein, as well as increasing plasma ammonia concentra-
tions, increase the concentration of ammonia in bovine fol-
licular fluid [18]. This has been associated with altered fol-
licular growth patterns and a reduction in both the number
of ova that cleave following insemination and the propor-
tion that develop to the blastocyst stage. Clearly there are
a number of mechanisms through which nutrition can act
to influence both follicle dynamics and the developmental
competence of oocytes. Although the precise mechanisms
are unknown, the observations that the intraovarian IGF
system can regulate the response of granulosa and theca
cells to gonadotropin [19–21] implicates this system as a
candidate for mediating the effects of metabolic hormones
on ovarian function.
In this study these mechanisms were analyzed by ma-
nipulating the intake of dietary energy and crude protein to
produce defined changes in circulating concentrations of
metabolic hormones and urea. The resultant changes in the
growth of antral follicles and the developmental compe-
tence of their oocytes were analyzed. To assess the involve-
ment of the intrafollicular IGF system in this process, di-
etary-induced changes in the steady-state concentrations of
1625
DIETARY REGULATION OF OVARIAN FUNCTION
FIG. 1. Experimental protocol. See text for details.
mRNA encoding components of the IGF system in bovine
antral follicles was measured and correlated with the ob-
served changes in follicle dynamics and oocyte quality.
MATERIALS AND METHODS
Experimental Design and In Vivo Procedures
The experiments described in this paper were approved
by the Animal Experiments Committee of the Scottish Ag-
ricultural College and were conducted under the auspices
and in accordance with the requirements of the Home Of-
fice Animals (Scientific Procedures) Act 1986.
Hereford
3
Friesian heifers (n
5
24) weighing 439
6
19.9 kg (mean
6
SD) with a body condition score (scale
1–5) of 2.7
6
0.14 (mean
6
SD) were used in this study.
Animals were allocated on the basis of live weight in a 2
3
2 balanced factorial design. They received diets provid-
ing daily metabolizable energy (ME) intakes of either 408
(low energy, LE) or 816 (high energy, HE) kJ/kg W
0.75
per
day (equivalent approximately to 0.8 and 1.6 times main-
tenance ME requirements, respectively). This was obtained
from rations containing dietary crude protein concentra-
tions of either 20 or 27 g/MJ ME (low protein, LP; and
high protein, HP, respectively). The energy density of all
diets was held constant at 10.4 MJ ME/kg dry matter.
Therefore, the two levels of daily ME intake involved of-
fering these diets at two levels of intake on a metabolic
live-weight basis consistent with the levels of ME intake
detailed above. Protein levels in the diets were adjusted by
altering the level of soya bean meal in the pelleted concen-
trate rations that also consisted of nutritionally improved
straw, barley grain, molassed sugar beet pulp, distillers’
dark grains, and soya bean meal. Diets were offered daily
as two meals at 0800 and 1600 h. The heifers also were
offered 2 kg of barley straw daily, as a source of long
roughage.
For 4 wk prior to the experiment, all 24 heifers received
the LELP diet. Then, during the study (Days 0–32; see Fig.
1), half of the animals received the LE diets and half the
HE diets. All animals received LP rations from Days 0 to
11 inclusive, with 12 heifers (6 from each of the two levels
of dietary energy intake) switching to the HP diet on Day
12. The delayed introduction of the HP ration was intended
to preclude any confounding effects due to possible phys-
iological adaptation by the heifers assigned to this more
extreme dietary formulation.
During the study, the heifers were managed in two
groups, 2 days apart, with each group comprising half of
the animals on each dietary energy
3
protein combination.
All were accommodated in individual pens on sawdust-bed-
ded floors within the same building. Intravaginal applica-
tion of a CIDR-B device (controlled internal drug-release;
SmithKline Beecham, Hertfordshire, UK), containing 1.9 g
progesterone for a 10-day period from Day 0, together with
an intramuscular administration of prostaglandin F
2
a
ana-
logue (15 mg Luprostiol; Prosolvin, Intervet, UK) on Day
6, ensured onset of a reference estrus on Day 12 of the
experimental program. Fourteen days later, a second injec-
tion of the prostaglandin analogue induced a second estrus
that occurred on Day 28. All heifers were artificially insem-
inated on this and the following day (Day 29) using semen
from a single Simmental sire (Fig. 1).
Ovarian follicular growth (diameter) was monitored dai-
ly from Day 17 of the experimental program (Day 5 after
the reference estrus) by transrectal real time B-mode ultra-
sonography using an Aloka SSD-500v scanner (Aloka Co.
Ltd., Tokyo, Japan) equipped with a 5-MHz linear array
transducer. Blood samples were collected at CIDR-B inser-
tion and removal, at Day 12, and daily from Day 17 onward
by jugular venipuncture into lithium heparin tubes on ice.
On each occasion samples were collected at 0800 h, just
before the animals were fed.
The heifers were slaughtered at a local abattoir on Day
32 of the experimental program (4 days after the second
induced estrus and first insemination), and reproductive
tracts were recovered within 25 min postmortem. The ova-
ries from each animal were examined and the site of the
recent ovulation noted before a small section of one ovary
(approximately 25% of its mass; alternating left or right
ovary from consecutive animals) was removed and imme-
diately processed for in situ hybridization studies. The re-
maining ovarian tissue was kept warm by prompt place-
ment in an insulated flask (30–35
8
C) while the reproductive
tracts were placed in an insulated, prewarmed (30–35
8
C)
container before transport to the laboratory.
Collection of Tissue for In Situ Hybridization
A portion of the ovary was collected, as described in the
previous section, taking care not to rupture any exposed
follicles. It was then oriented on a flat surface to expose
the maximum number of follicles. A map of the ovarian
surface was drawn, showing the location of all follicles.
Follicle diameters were recorded before coating with opti-
mal cutting temperature (OCT) compound (Merck, Poole,
Dorset, UK) and frozen above liquid nitrogen before im-
mersion in liquid nitrogen. The ovarian blocks were stored
at
2
70
8
C until required for subsequent in situ hybridization.
After cutting ovarian sections (14
m
m), follicles were
classified, morphologically, as healthy, atretic, or grossly
atretic. Healthy follicles had an intact basement membrane
and a healthy granulosa cell layer with only a few pycnotic
nuclei. Atretic follicles had fewer granulosa cells with local
disruption in the basement membrane and cells with pyc-
notic nuclei were identified within the granulosa layer. Fol-
licles with a more extensively disrupted basement mem-
brane, and with a significant reduction in the number of
granulosa cells and increase in the number of pycnotic nu-
clei were classified as grossly atretic [22].
Riboprobes
Riboprobes for IGF binding protein (BP)-2 and IGFBP-4
[23] and IGF-II, and type 1 IGF receptor [24] were labeled
with
35
S-UTP according to the method described by Arms-
trong et al. [23]. The insulin receptor probe was obtained
1626
ARMSTRONG ET AL.
by reverse transcription-polymerase chain reaction using
RNA isolated from bovine luteal tissue. The forward and
reverse primers were: 5
9
-aactcttcttccactataaccc-3
9
and 5
9
-
gcaatgtcgtttctctcc-3
9
, respectively, and amplified a 100-base
pair cDNA corresponding to positions 1493–1592 of the
human insulin receptor cDNA [25].
In Situ Hybridization
Frozen sections (14
m
m) were dehydrated, fixed, and
probed with
35
S-labeled riboprobes according to the method
described by Xu et al. [22]. After the final high stringency
wash the sections were dipped in autoradiographic K2 pho-
tographic emulsion (Ilford Limited, Mobberley, Cheshire,
UK) and exposed for 3 wk at 4
8
C. Sections were then de-
veloped (Kodak D-19; H.A. West Watson Cres., Edinburgh,
UK) and fixed using Hypam fixer (Ilford Limited) before
staining in hematoxylin and eosin. The sections were finally
mounted in DPX (R.A. Lamb, London, UK) before micro-
scopic examination using both light- and dark-field illu-
mination.
Image Analysis
The intensity of the in situ hybridization signal was an-
alyzed using an NIH-Image analysis system (NIH, Bethes-
da, MD) as described previously [23]. Briefly, the number
of graphic pixels occupied by silver grains (identified by a
set gray threshold) within a defined area of the tissue sec-
tion was counted and presented as a percentage of the total
pixel number within the defined area. The hybridization
intensity is therefore presented as the percentage of occu-
pied pixels to total pixels within a defined area of the tissue.
Background hybridization intensity, measured with the
sense RNA probes, was subtracted from the measurements
obtained with the antisense probes to give the final hybrid-
ization signal. Within each follicle three separate fields
were analyzed for each probe. There was no significant
difference (P
.
0.05) in hybridization intensity obtained
with antisense and sense RNA probes within a nonexpress-
ing region of a tissue section. Under the conditions de-
scribed here the hybridization signal was proportional to
the length of time the slides were exposed to photographic
emulsion for up to 3 wk.
In Vivo Zygote Collection and Evaluation
On arrival at the laboratory, and not more than 2 h post-
mortem, the uterine horn and oviduct ipsilateral to the cor-
pus luteum were each flushed with 10 ml phosphate-buff-
ered fluid containing 0.4% w/v BSA (Ovum Culture Me-
dium; Imperial Laboratories, Andover, Hampshire, UK) to
retrieve recently ovulated and fertilized eggs for evaluation
of their development in vivo.
Oocyte Retrieval and Embryo Production In Vitro
The collection of ovaries was timed to coincide with
presumptive emergence of the first follicular wave and to
be in advance of establishment of follicular dominance. The
status of all ovaries at slaughter was determined with re-
spect to the numbers of follicles in 1- to 4-mm, 4- to 8-
mm, and
.
8-mm diameter categories. Follicles in the 1- to
4-mm and 4- to 8-mm categories were aspirated separately,
and all retrieved oocyte-cumulus complexes (OCC) classi-
fied as competent, on the basis of conventional qualitative
evaluation criteria [26] were retained for in vitro oocyte
maturation (within-category). The OCCs were matured on
4-day-old bovine granulosa cell layers in 50-
m
l droplets of
Medium 199 with Earles salts (Life Technologies, Paisley,
UK) supplemented with 10% v/v heat-inactivated steer se-
rum (Globepharm, Esher, UK). The maturation medium
was overlaid with mineral oil. After 24 h maturation at
38.5
8
C (5% CO
2
in air), oocytes were fertilized in vitro,
using swim-up-derived frozen-thawed spermatozoa from
the same Simmental sire (20 h) as used for artificial insem-
ination in modified Tyrode, albumin, lactate, pyruvate me-
dium containing 0.6% w/v fatty acid-free BSA (pH
5
7.8;
290–310 mOsm). They were then transferred (Day 1) for
further culture, until Day 8, in a modified version of syn-
thetic oviductal fluid [27] containing 0.99 mmol L
2
1
sodi-
um pyruvate and 9.90 mmol L
2
1
sodium lactate and sup-
plemented with 10% v/v heat-inactivated steer serum. Post-
fertilization culture was carried out in 50-
m
l droplets under
mineral oil (5% CO
2
,5%O
2
, 90% N
2
; 38.5
8
C), and eggs
were transferred to fresh droplets at 48-h intervals. Inci-
dence of cleavage (Day 3) and blastocyst development
(Days 7 and 8) were determined by visual evaluation (up
to 400
3
magnification). Embryos at the appropriate stage
of development for the day of culture and of good mor-
phological quality, as determined by the criteria of Lindner
and Wright [28] were classified as viable. All oocyte and
embryo evaluations were carried out by operators who were
unaware of the treatments from which the eggs were de-
rived.
Metabolite and Hormone Assays
Urea.
Plasma samples were analyzed for urea on a
BMD/Hitachi 705 autoanalyzer using a commercially avail-
able kit (Boehringer Mannheim [Diagnostics and Biochem-
icals], Lewes, East Sussex, UK).
Progesterone.
Plasma progesterone concentrations were
measured without prior extraction, using an
125
I-labeled
progesterone double-antibody RIA [29]. The nonextraction
assay was modified and validated to enable the use of a
rabbit antiprogesterone first antibody, donkey antirabbit
IgG, and normal rabbit serum that were obtained as gifts
from Diagnostics Scotland, Carluke, Lanarkshire, UK. The
sensitivity of the assay at an 80% effective dose (ED
80
)
was 0.51 ng/ml. The inter- and intraassay coefficients of
variation for low, medium, and high controls were, 15.7%
and 9.5%, 9.3% and 7.8%, and 6.9% and 4.9%, respectively.
Insulin.
Plasma insulin concentrations were measured us-
ing an
125
I-labeled insulin double-antibody RIA [30]. The
assay was modified to use porcine insulin (I-3505, 24 IU/
mg) obtained from Sigma; guinea pig antiporcine insulin,
normal guinea pig serum, and sheep antiguinea pig IgG
obtained as a gift from Diagnostics Scotland, Law Hospital,
Carluke, UK. The sensitivity of the assay at ED
80
was 2.50
mIU/L. The inter- and intraassay coefficients of variation
for low, medium, and high controls were 12.6% and 10.8%,
7.6% and 7.5%, and 7.1% and 6.6%, respectively.
Insulin-like growth factor-I.
Plasma IGF-I was measured
by RIA after acid gel filtration to remove IGFBPs [31],
using a rabbit polyclonal antibody raised against human
recombinant IGF-I [32]. The sensitivity of the assay was
22 pg, and the inter- and intraassay coefficients of variation
were 13% and 8%, respectively.
Follicle-stimulating hormone.
Plasma FSH was mea-
sured using an
125
I-labeled FSH double-antibody RIA [33].
The assay used bovine FSH (USDA-bFSH-1-2, for stan-
dards and USDA-b-1-2, for radioiodination), obtained as a
gift from Dr. D.J. Bolt through the USDA Animal Hormone
1627
DIETARY REGULATION OF OVARIAN FUNCTION
FIG. 2. Changes in mean (6SEM) circulating insulin (a) and IGF-I (b)
concentrations throughout the experimental period. High and low energy
intakes (grouped over both dietary protein concentrations) are represented
by open and solid symbols, respectively. The CIDR insertion was between
Day 0 and 10, prostaglandin injections were carried out on Days 6 and
26, and the reference estrus was recorded on Day 12.
Program, Beltsville, MD. Rabbit antiovine FSH-1 antise-
rum was obtained as a gift from Dr. A.F. Parlow through
the program of the National Institute of Diabetes and Di-
gestive and Kidney Diseases, Harbor-UCLA Medical Cen-
ter, Torrance, CA. Normal rabbit serum and second anti-
body, donkey antirabbit IgG were obtained as gifts from
Diagnostics Scotland, Law Hospital, Carluke, UK. The sen-
sitivity of the assay at ED
80
was 69 ng/L. The inter- and
intraassay coefficients of variation for low, medium, and
high controls were 10.3% and 8.8%, 5.7% and 5.2%, and
6.6% and 6.1%, respectively.
Statistical Analyses
Data relating to plasma insulin, IGF-I, urea, FSH, and
follicular growth were analyzed using repeated-measures
(split-plot) ANOVA (Genstat 5, Version 3.2; Lawes, Ro-
thamstead, 1993). Level of feeding (energy intake) and pro-
tein concentration in the diet (CP/MJ ME) and their asso-
ciated interactions were used as the between-animal stra-
tum. Sample date and associated interactions with energy
intake and dietary protein concentration formed the within-
animal stratum.
Live-weight change and growth rate of the dominant fol-
licle of the second follicular wave following the reference
estrus were analyzed by regressing heifer live weight and
follicle diameter, respectively, to day of experiment foreach
animal and then analyzing the regression coefficients by a
two-way ANOVA.
Oocyte quality and embryo production data, subjected
to square root or arcsine transformation as appropriate,
were also analyzed using split-plot ANOVA with the block-
ing structure comprising slaughter date, animal, and follicle
size category. Thus, the between-animal stratum examined
effects of energy intake and dietary protein concentration
and associated interactions. The within-animal stratum ex-
amined effects of follicle size category and associated in-
teractions with energy intake and dietary protein concen-
tration. In addition, within each follicle size category, sin-
gle-factor ANOVA was used to compare embryo produc-
tion data for animals with contrasting mean plasma urea
concentrations (moderate,
,
6.0 mmol/L; high,
$
6.0 mmol/
L) over the 15-day period prior to oocyte collection (Days
17–31 inclusive).
The effects of energy intake and dietary protein concen-
tration on the expression of mRNA encoding components
of the IGF system were analyzed, after arcsine transfor-
mation, using an unbalanced-linear-mixed model according
to the method of residual maximum likelihood (REML;
Genstat 5, Version 3.2; Lawes, Rothamstead, 1993). The
model was constructed using random effects described by
the fixed block structure in which replicate observations
were nested within follicles that were nested within cows
and fixed effects described by the main effects of dietary
energy levels, dietary protein concentrations, and follicle
size plus their interactions. Individual animals were used as
an absorbing factor in the analysis so as to reduce the size
of the matrices involved in the calculation.
RESULTS
Growth Rate
During the course of the experiment, heifers offered the
HE diets gained weight (1.12 kg/day), whereas those of-
fered the LE diets lost weight (
2
0.14 kg/day; SED
5
0.12;
P
,
0.001). Growth rate was unaffected by protein level
in the diet. Energy intake and dietary protein concentration
had no effect on body condition score (P
.
0.1).
Plasma Insulin and IGF-I Concentrations
Mean plasma insulin and IGF-I concentrations (Fig. 2)
were higher for heifers offered the HE compared to the LE
diets (P
,
0.001) but were unaffected by the protein con-
centration in the diet (P
.
0.05). Both hormones showed
similar changes in concentrations during the estrous cycle
with maximum concentrations on the day of ovulation. The
ovulatory increases in plasma insulin and IGF-I concentra-
tions were more pronounced in cattle offered the HE diet.
Plasma FSH and Progesterone Concentrations
Plasma FSH concentrations did not differ significantly
(P
.
0.05) among dietary treatments. Mean values ranged
from 0.14 ng/ml on the day of the reference estrus up to
maximum values of 0.22 ng/ml during the first and second
waves of follicle growth following the reference estrus.
Mean plasma progesterone concentrations (Fig. 3a) be-
tween Days 17 and 26 of the experiment (Days 5 and 14
of the estrous cycle) were significantly higher in heifers
offered the HE than the LE diets (6.2 vs. 4.8 ng/ml; SED
5
0.6; P
,
0.05). However, there was a significant inter-
action (P
,
0.05) between dietary energy level and protein
concentration with day of cycle for plasma progesterone
levels during this period. As Day 14 of the estrous cycle
approached, progesterone levels in heifers offered the
LEHP and HEHP diets converged, whereas progesterone
levels in heifers offered the LELP and HELP diets di-
verged. Mean plasma progesterone levels between Days 5
1628
ARMSTRONG ET AL.
FIG. 3. Changes in plasma progesterone concentrations (a) and the di-
ameter of the dominant follicle (b) during the first follicle wave of the
synchronized estrous cycle for heifers offered the LELP (m), LEHP (m),
HELP (□), and HEHP (n) diets and during the second follicle wave for
heifers offered the HE (V)orLE(v) diets (averaged over both dietary
protein concentrations). A prostaglandin injection was given on Day 26
of the cycle. The bar represents a pooled standard error of the difference
(SED).
FIG. 4. Plasma urea concentrations during the experimental period are
shown for heifers offered the LELP (m), LEHP (m), HELP (□), and HEHP
(n) diets. The CIDR insertion was between Day 0 and 10, prostaglandin
injections were carried out on Days 6 and 26, and the reference estrus
was recorded on Day 12. The bar represents a pooled standard error of
the difference (SED).
and 14 of the estrous cycle for heifers offered the LELP,
LEHP, HELP, and HEHP diets were 4.3, 5.3, 6.6, and 5.7
ng/ml, respectively (SED
5
0.8).
Follicle Dynamics
The mean diameter of the dominant follicle (DF) of the
first follicular wave after the induced reference estrus on
Day 12 of the experimental program (Fig. 3b) was greater
for heifers offered HE rather than LE diets (11.0 vs. 8.1
mm; SED
5
0.73; P
,
0.01). There was a significant in-
teraction (P
,
0.001) between dietary protein concentration
and day of cycle on the diameter of the DF of the first
wave. This indicated that maximum size and subsequent
initiation of atresia of the DF occurred 1–2 days earlier in
heifers offered HP rather than LP diets. Growth rate of the
DF of the second follicular wave was significantly greater
in heifers offered the HE than the LE diets (1.78 vs. 1.22
mm/day; SED
5
0.20; P
,
0.05). The diameter of this
follicle at the point of induced ovulation was also greater
in heifers offered HE rather than LE diets (15.6 vs. 12.1
mm; SED
5
0.8; P
,
0.001). Dietary protein concentration
had no effect on the growth rate of the DF of the second
follicular wave. Energy intake and dietary protein concen-
tration had no effect on the number of 1- to 4-mm diameter
follicles (P
5
0.57) or 4- to 8-mm diameter follicles (P
5
0.61) on the day of slaughter.
Plasma Urea Concentrations
Plasma urea levels increased between Days 0 and 12 of
the experimental program in heifers offered the HE but not
the LE diets (Fig. 4). Plasma urea concentrations increased
further between Days 12 and 17 of the program in heifers
offered the HP but not the LP diets. Mean plasma urea
levels for heifers offered the LELP, LEHP, HELP, and
HEHP diets between Days 17 and 31 of the program were
4.3, 6.3, 6.1, and 7.4 mmol/L, respectively (SED
5
0.4; P
,
0.001).
Fertilization In Vivo
Examination of ovaries postmortem indicated that ovar-
ian follicle populations were similar for all diets and that
all 24 heifers had ovulated recently, one (HELP diet) with
twin ovulations. Native zygotes, fertilized in vivo and col-
lected by flushing the reproductive tracts postmortem, were
retrieved from 19 heifers. Among these, 4 LELP (67%), 3
LEHP (60%), 2 HELP (50%), and 2 HEHP (50%) donor-
derived ova, respectively, had reached the 8-cell stage of
development at collection. The remainder were less ad-
vanced.
In Vitro Fertilization and Blastocyst Production
Yields of oocytes and the proportions classified as suit-
able for in vitro embryo production, together with subse-
quent postfertilization development data, are presented in
Table 1. Oocyte yields from animals on the contrasting diets
were not significantly different, although follicle size cat-
egory influenced the proportions subsequently classified as
suitable for in vitro embryo production (P
,
0.005), with
proportionately more OCCs from medium-sized follicles
(4–8 mm) selected. All 24 donors yielded a minimum of
two suitable oocytes each from their 4- to 8-mm follicle
cohorts, whereas three animals, all receiving the HP diet (1
5
LEHP; 2
5
HEHP), failed to yield suitable oocytes from
small (1- to 4-mm) follicles. Zygote cleavage data, recorded
on Day 3 (48 h post-in vitro fertilization) and reflecting
presumptive fertilization incidence among selected oocytes,
were unaffected by either donor diet or follicle size cate-
gory. In contrast, blastocyst production data, expressed as
proportions of zygotes cleaved by Day 3 and classified in
terms of their viability (Table 1), were influenced by dietary
energy (LE
.
HE; P
5
0.032). Moreover, within-animal
data analysis indicated that cleaved ova derived from 1- to
4-mm follicles were less capable of development to blas-
1629
DIETARY REGULATION OF OVARIAN FUNCTION
TABLE 1. In vitro oocyte evaluation and blastocyst production from heifers offered rations containing low (LE) and high (HE) dietary energy and low
(LP) and high (HP) protein levels.
a
Diets over follicle size (mm)
LELP
1–4 4–8
LEHP
1–4 4–8
HELP
1–4 4–8
HEHP
1–4 4–8
SED
Within
animal
Between
animal
No. of aspirated follicles
No. of selected oocytes
Selected oocytes; proportion of col-
lected ova
Cleaved ova; proportion of selected
oocytes (n, if ,6)
Blastocysts; proportion of cleaved ova
(n, if ,6)
Viable blastocysts; proportion of
cleaved ova (n, if ,6)
18.8
10.0
0.59
0.73
0.53
0.49
26.3
11.7
0.78
0.75
0.60
0.56
13.0
8.0
0.63
0.69
(5)
0.44
(5)
0.39
(5)
19.3
11.8
0.80
0.79
(5)
0.66
(5)
0.57
(5)
14.0
7.2
0.69
0.64
0.36
0.31
17.8
9.8
0.75
0.80
0.61
0.53
8.5
2.8
0.34
0.63
(4)
0.10
(3)
0.07
(3)
19.3
13.0
0.80
0.77
(4)
0.53
(3)
0.53
(3)
3.94
2.39
0.11
0.13
0.17
0.16
6.78
4.34
0.12
0.15
0.15
0.15
a
Data, based on n 56 heifers per diet unless specified otherwise, are presented as untransformed mean values with corresponding within- and
between-animal SED values.
tocysts in vitro than those derived from 4- to 8-mm follicles
(P
5
0.023). For animals yielding 1- to 4-mm follicle-de-
rived ova that cleaved following fertilization in vitro, blas-
tocysts were generated from five of six LELP (83%), four
of five LEHP (80%), four of six HELP (67%), and one of
three HEHP (33%) donors, respectively. The proportions of
blastocysts classified as viable were higher (P
5
0.056)
among ova from the medium-size follicle size category,
while dietary protein concentration tended to influence the
proportions of viable blastocysts that were graded as ex-
cellent or good (grades 1
1
2/viable: LP
.
HP, P
5
0.062).
A retrospective analysis of the developmental compe-
tence of blastocysts derived from donors exhibiting mod-
erate or high plasma urea levels in the 14 days prior to
oocyte collection was performed. In respect of blastocysts
produced from selected oocytes derived from small folli-
cles, proportionately more viable blastocysts (0.31
6
0.081
vs. 0.14
6
0.051; P
,
0.08) were produced from donors
with moderate (
,
6 mmol/L) plasma urea concentrations (n
5
10) than from those exhibiting high (
.
6 mmol/L) plasma
urea concentrations (n
5
11).
Ovarian IGF System
The spatial distribution of mRNA encoding components
of the IGF system within bovine antral follicles is shown
in Figure 5. The IGF-II and IGFBP-4 mRNA were confined
to thecal tissue; in contrast, mRNA encoding IGFBP-2 was
located in granulosa cells. Both type 1 IGF receptor mRNA
and insulin receptor mRNA were detected in both granulosa
and thecal tissue with expression in granulosa tissue greater
than in thecal tissue. In addition to granulosa and thecal
tissue, mRNA encoding type 1 IGF receptor was also de-
tected in oocytes from preantral and antral follicles. The
IGF-I mRNA was not detected in bovine ovarian follicular
tissue. The HE diet reduced the steady-state concentrations
of mRNA encoding all the components of the IGF system
in healthy follicles. The reduction was significant (P
,
0.05) in the case of IGFBP-2, IGFBP-4, and insulin recep-
tor mRNA. No effect of diet was observed in atretic or
grossly atretic follicles (data not shown). All subsequent
statistical analyses were carried out using data from healthy
follicles alone. The interaction between follicle size and
dietary energy and protein is shown in Figure 6. Significant
(P
,
0.05) reductions in the steady-state concentrations of
mRNA encoding IGFBP-2 and -4 and insulin receptor in
small (
,
4-mm) follicles were induced by the HE diet.
There was no direct effect of dietary protein on the ex-
pression of mRNA encoding components of the IGF sys-
tem.
DISCUSSION
The overall aim of this work was to analyze mechanisms
through which dietary energy and protein influence follic-
ular dynamics and the developmental competence of oo-
cytes in cattle. It is the first study to demonstrate a direct
effect of dietary intake on the expression of mRNA encod-
ing components of the ovarian IGF system and supports the
hypothesis that the nutritional regulation of follicular
growth is mediated, at least in part, by the action of cir-
culating metabolic hormones on the ovarian IGF system.
The high level of energy intake described in this study
significantly increased plasma insulin and IGF-I concentra-
tions relative to the LE diet. In contrast, dietary protein
concentration had no effect on these metabolic hormones.
The absence of any significant changes in FSH concentra-
tions supports the results of other studies that failed to show
any effects of short-term changes in the plane of nutrition
on FSH concentrations in ruminants [2]. The ovulatory in-
crease in plasma IGF-I concentrations supports results of
related studies in sheep [34] and is probably a consequence
of increased plasma estradiol stimulating hepatic IGF-I pro-
duction because treatment of ovariectomized cattle with es-
tradiol was shown to increase serum IGF-I concentrations
significantly [35].
Although there is no direct evidence of an endocrine role
for insulin and IGF-I in the control of follicular growth,
increasing indirect evidence indicates that such a role ex-
ists. For example, plasma insulin and IGF-I concentrations
have been positively correlated with follicular growth in
postpartum dairy cattle [5, 36], and twinning rate in cattle
is also associated with elevated plasma IGF-I concentration
[37]. The stimulation of antral follicle growth following
recombinant GH treatment of nonlactating heifers was sim-
ilarly associated with increased circulating concentrations
of insulin and IGF-I [38, 39]. Collectively, the available
evidence suggests that the increased size of the dominant
follicle in cattle fed the HE diet in the present study was
due to endocrine actions of both these metabolic hormones
on follicular growth.
Luteal progesterone production during the induced es-
trous cycle was significantly increased in cattle offered the
HE diet and is probably a reflection of increased follicle
1630
ARMSTRONG ET AL.
FIG. 5. Light-field illumination (a,d,g,j, and m) and dark-field illumination of bovine ovarian tissue sections (14 mm) hybridized with antisense (b,
e,h,k, and n) and sense (c,f,i,l, and o) riboprobes. The probes were designed to detect bovine IGF-II (a–c), IGFBP-2 (d–f), IGFBP-4 (g–i), type 1 IGF
receptor (j–l), and insulin receptor (m–o). T and G represent theca and granulosa tissue, respectively. The arrow indicates expression of mRNA encoding
type 1 IGF receptor in an oocyte. Bar 5180 mm.
1631
DIETARY REGULATION OF OVARIAN FUNCTION
FIG. 6. The interaction of dietary energy (a,c,e,g, and i) and protein
(b,d,f,h, and j) intake with follicle size on the expression of mRNA
encoding IGF-II (a,b), IGFBP2 (c,d), IGFBP4 (e,f), type 1 IGF receptor
(g,h), and insulin receptor (i,j). The solid and open bars (a,c,e,g, and
i) represent high and low energy diets (grouped over both protein diets),
respectively. The solid and open bars (b,d,f,h, and j) represent high and
low protein diets (grouped over both energy diets), respectively. Messen-
ger RNA expression is expressed as the percent pixels within a defined
area occupied by a silver grain. SED represents pooled standard error of
the difference within follicles of the same size. *
P
,0.05 within follicles
of the same size.
size prior to ovulation. This has relevance to embryo sur-
vival in the pregnant cow as increases in luteal progesterone
production as early as Day 5 of pregnancy have been shown
to significantly increase embryo growth and associated in-
terferon-
t
production [40].
The spatial patterns of expression of mRNA encoding
components of the ovarian IGF system in antral follicles
support previous observations [24, 26, 41, 42]. The results
also clearly demonstrate the presence of mRNA encoding
type 1 IGF receptor in oocytes from preantral and antral
follicles, confirming that IGF-I can regulate oocyte matu-
ration directly through binding to its own receptor.
Although nutritional status has been shown to regulate
the expression of mRNA encoding components of the he-
patic IGF system [43], the results presented here show, for
the first time, a direct effect of dietary intake on the ovarian
IGF system. Specifically, increased dietary energy signifi-
cantly decreased the steady-state concentration of mRNA
encoding IGFBP-2 and -4 and insulin receptor in small an-
tral follicles. Previous studies [23] have shown that during
the development of dominance there is a significant de-
crease in the steady-state concentration of mRNA encoding
IGFBP-2 in granulosa cells from the dominant follicle. A
reduction in the local level of IGFBP-2 would increase IGF
bioactivity in the dominant follicle that, in turn, would be
expected to increase the sensitivity/response of the domi-
nant follicle to FSH, thus allowing its continued growth in
an environment of decreasing systemic FSH concentrations.
Similarly, in heifers offered the HE compared to the LE
diet in the present study, the reduction in the steady-state
concentration of mRNA encoding IGFBP-2 and -4 in small
antral follicles is expected to increase the bioavailability of
intrafollicular IGF (both locally produced IGF-II and sys-
temically derived IGF-I) in these follicles. The consequent
increase in the sensitivity/response toward FSH would be
expected to result in an increased rate of follicular growth.
The factors involved in regulating IGFBP-2 and -4
mRNA expression in the bovine ovary are not understood.
Although we have shown that FSH decreases IGFBP-2
mRNA expression in cultured granulosa cells [23], the roles
of insulin and IGF-I, the most likely candidates for medi-
ating the effects of dietary energy on the ovarian IGF sys-
tem have not been examined. The decrease in insulin re-
ceptor mRNA expression in cattle offered the HE diet is
most probably due to insulin-mediated receptor downreg-
ulation. The functional significance of this latter observa-
tion, however, is not known.
Oocyte quality in small follicles was negatively corre-
lated with plasma urea concentrations (Table 1 and Fig. 4).
Exposure of follicle-enclosed oocytes to high levels of am-
monia and/or urea has previously been observed to com-
promise their capacity to develop to the blastocyst stage
following a period of in vitro culture [18]. In that study,
however, the detrimental effect of ammonia and/or urea on
oocyte quality was greater for oocytes from medium than
from small follicles. The mechanisms by which oocyte
competence from specific follicle size categories is com-
promised by these nitrogen moieties are not understood,
and the most vulnerable follicle size category may be
linked to the energy status. In the present experiment the
highest levels of urea reflected the combined effects of
feeding a high protein diet at a high level (HEHP). It is
therefore not possible, with the present experimental de-
sign, to ascertain whether energy intake per se may have
modified the known detrimental effect of high plasma urea
on the oocyte.
The ovarian IGF system also has the potential to interact
directly with the oocyte through the type 1 IGF receptor.
Small follicles from heifers offered the HEHP diet in the
present study had significantly reduced levels of mRNA
encoding IGFBP-2 and -4, and as discussed, we expect this
to increase the bioactivity of IGF in these follicles, which
is probably a critical factor controlling oocyte developmen-
tal capacity. Indeed our results indicate that overstimulation
by IGF may be detrimental to oocyte development.
In conclusion we have shown that dietary energy and pro-
tein can directly affect the expression of mRNA encoding
components of the ovarian IGF system. We hypothesize that
the resultant changes in the ovarian IGF system increase the
sensitivity of follicles toward FSH and, in combination with
dietary-induced increases in the concentration of circulating
insulin and IGF-I, contribute to the observed increase in
growth rate of the dominant follicle. Dietary protein concen-
tration also influenced oocyte quality, with developmental
1632
ARMSTRONG ET AL.
competence being negatively correlated with plasma urea
levels. In addition, we hypothesize that nutritionally induced
changes in the ovarian IGF system may play a key role in
regulating oocyte quality. Finally, the data presented here
indicate that future studies concerned with the formulation
of diets designed to optimize cattle fertility must take into
account the possibility of divergent actions of nutrient supply
on follicular growth and oocyte quality.
REFERENCES
1. Robinson JJ. Nutrition in the reproduction of farm animals. Nutr Res
Rev 1990; 3:253–276.
2. O’Callaghan D, Boland MP. Nutritional effects on ovulation, embryo
development and the establishment of pregnancy in ruminants. Anim
Sci 1999; 68:299–314.
3. Robinson JJ, Sinclair KD, McEvoy TG. Nutritional effects on foetal
growth. Anim Sci 1999; 68:315–332.
4. Rhodes FM, Fitzpatrick LA, Entwistle KW, De’ath G. Sequential
changes in ovarian follicular dynamics in Bos indicus heifers before
and after nutritional anoestrus. J Reprod Fertil 1995; 104:41–49.
5. Beam SW, Butler WR. Effects of energy balance on follicular devel-
opment and first ovulation in postpartum dairy cows. J Reprod Fertil
1999; 54(suppl):411–424.
6. Gutierrez CG, Oldham J, Bramley TA, Gong JG, Campbell BK, Webb
R. The recruitment of ovarian follicles is enhanced by increased die-
tary intake in heifers. J Anim Sci 1997; 75:1876–1884.
7. Cameron JL. Regulation of reproductive hormone secretion in primates
by short-term changes in nutrition. Rev Reprod 1996; 1:117–126.
8. Harrison LM, Randel RD. Influence of insulin and energy intake on
ovulation rate, luteinizing hormone and progesterone in beef heifers.
J Anim Sci 1986; 63:1228–1235.
9. Downing JA, Joss J, Scaramuzzi RJ. The effect of a direct arterial
infusion of insulin and glucose on the ovarian secretion rates of an-
drostenedione and oestradiol in ewes with an autotransplanted ovary.
J Endocrinol 1999; 163:531–541.
10. Gong JG, Bramley T, Webb R. The effect of recombinant bovine so-
matotropin on ovarian function in heifers: follicular populations and
peripheral hormones. Biol Reprod 1991; 45:941–949.
11. Gong JG, Bramley TA, Webb R. The effect of recombinant bovine
somatotrophin on ovarian follicular growth and development in heif-
ers. J. Reprod Fertil 1993; 97:247–254.
12. Gutierrez CG, Campbell BK, Webb R. Development of a long-term
bovine granulosa cell culture system: induction and maintenance of
oestradiol production, response to follicle stimulating hormone and
morphological characteristics. Biol Reprod 1997; 56:608–616.
13. Campbell BK, Scaramuzzi RJ, Webb R. Induction and maintenance
of oestradiol and immunoreactive inhibin production with FSH by
ovine granulosa cells cultured in serum-free media. J Reprod Fertil
1996; 106:7–16.
14. Armstrong DG, Campbell BK, Hogg CO, WebbR. Insulin-like growth
factor binding protein production by primary cultures of ovine gran-
ulosa and theca cells. The effect of IGF-I gonadotropin and follicle
size. Biol Reprod 1996; 55:1163–1171.
15. Webb R, Gosden RG, Telfer EE, Moor RM. Factors affecting folli-
culogenesis in ruminants. Anim Sci 1999; 68:257–284.
16. McEvoy TG, Robinson JJ, Aitken RP, Findlay PA,Palmer RM, Robertson
IS. Dietary-induced suppression of preovulatory progesterone concentra-
tions in superovulated ewes impairs the subsequent in vivo and in vitro
development of their ova. Anim Reprod Sci 1995; 39:89–107.
17. O’Callaghan D, Yaakub H, Hyttel P, Spicer LJ, Boland MP. Effect of
nutrition and superovulation on oocyte morphology, follicular fluid
composition and systemic hormone concentrations in ewes. J Reprod
Fertil 2000; 118:303–313.
18. Sinclair KD, Kuran M, Gebbie FE, Webb R, McEvoy TG. Nitrogen
metabolism and fertilility in cattle: II. Development of oocytes recov-
ered from heifers offered diets differing in their rate of nitrogen re-
lease in the rumen. J Anim Sci 2001; 78:2670–2680.
19. Armstrong DG, Webb R. Ovarian follicular dominance: the role of intra-
ovarian growth factors and novel proteins. Rev Reprod 1997; 2:139–146.
20. Webb R, Armstrong DG. Control of ovarian function; effect of local
interactions and environmental influences on follicular turnover in cat-
tle: a review. Livestock Prod Sci 1998; 53:95–112.
21. Webb R, Campbell BK, Garverick HA, Gong JG, Gutierrez CG,
Armstrong DG. Molecular mechanisms regulating follicular recruit-
ment and selection. J Reprod Fertil 1999; 54(suppl):33–48.
22. Xu ZZ, Garverick HA, Smith GW, Smith MF, Hamilton SA, Youn-
quist RS. Expression of follicle-stimulating hormone and luteinizing
hormone receptor messenger ribonucleic acids in bovine follicles dur-
ing the first follicular wave. Biol Reprod 1995; 53:951–957.
23. Armstrong DG, Baxter G, Gutierrez CG, Hogg CO, Glazyrin AL,
Campbell BK, Bramley TA, Webb R. Insulin-like growth factor bind-
ing protein-2 and -4 mRNA expression in bovine ovarian follicles:
effect of gonadotropins and developmental status. Endocrinology
1998; 139:2146–2154.
24. Armstrong DG, Gutierrez CG, Baxter G, Glazyrin AL, Mann GE,
Woad KJ, Hogg CO, Webb R. Expression of mRNA encoding IGF-I,
IGF-II and type 1 IGF receptor in bovine ovarian follicles. J Endo-
crinol 2000; 165:101–113.
25. Ebina Y, Ellis L, Jarnagin K, Edery M, Graf L, Clause E, Ou JH,
Masiarz F, Kan YH, Goldfine ID, Roth RA. The human insulin re-
ceptor cDNA: the structural basis for hormone-activated transmem-
brane signalling. Cell 1985; 40:747–758.
26. Hawk HW, Wall RJ. Improved yields of bovine blastocysts from in-
vitro produced oocytes. 1. Media and coculture of cells. Theriogenol-
ogy 1994; 41:1585–1594.
27. Tervit HR, Whittingham DG, Rowson LEA. Successful culture in vi-
tro of sheep and cattle ova. J Reprod Fertil 1972; 30:493–497.
28. Lindner GM, Wright RW Jr. Bovine embryo morphology and evalu-
ation. Theriogenology 1983; 20:407–416.
29. Corrie J, Hunter W, Macpherson J. A strategy for radioimmunoassay
of plasma progesterone with the use of homologous site
125
I-labelled
radioligand. Clin Chem 1981; 27:594–599.
30. Starr JI, Horwitz DL, Rubenstein AH, Mako ME. Insulin, proinsulin
and c-peptide. In: Jaffer BM, Behman HR (eds.), Methods of Hor-
mone Radioimmunoassay, 2nd ed. New York: Academic Press; 1979.
31. Gutierrez CG, Campbell BK, Armstrong DG, Webb R. Insulin-like
growth factor-I (IGF-I) production by bovine granulosa cells in vitro
and peripheral IGF-I measurement in cattle serum: an evaluation of
IGFBP extraction protocols. J Endocrinol 1997; 153:231–240.
32. Armstrong DG, Duclos MJ, Goddard C. Biological activity of insulin-
like growth factor-I purified from chicken serum. Domest Anim En-
docrinol 1990; 7:383–393.
33. Crowe MA, Padmanabhan V, Hynes N, Sunderland SJ, Beitins IZ,
Enright WJ. Validation of a sensitive RIA for measurement of serum
FSH in cattle, and its correlation with FSH bioassay. In: Program of
the XX meeting of the Joint Societies of Reproduction and Fertility;
1995; Dublin. Abstract series 15, Abstract 112.
34. Leeuwenberg BR, Hudsen NL, Moore LG, Hurst PR, McNatty KP.
Peripheral and ovarian IGF-I concentrations during the ovine oestrous
cycle. J Endocrinol 1996; 148:281–289.
35. Richards MW, Welleman RP, Spicer LJ, Morgan CL. Nutritional an-
estrous in beef cows: effect of body condition and ovariectomy on
serum luteinizing hormone and insulin-like growth factor-I. Biol Re-
prod 1991; 44:961–966.
36. Beam SW, Butler WR. Energy balance, metabolic hormones and early
postpartum follicle development in dairy cows fed prilled lipid. J
Dairy Sci 1998; 81:121–131.
37. Echternkamp SE, Spicer LJ, Gregory KE, Canning SF, Hammond JM.
Concentration of insulin-like growth factor-I in blood and ovarian fol-
licular fluid of cattle selected for twins. Biol Reprod 1990; 43:8–14.
38. Gong JG, Bramley T, Webb R. The effect of recombinant bovine so-
matotropin on ovarian function in heifers: follicular populations and
peripheral hormones. Biol Reprod 1991; 45:941–949.
39. Gong JG, Baxter G, Bramley T, Webb R. Enhancement of ovarian fol-
licular development in heifers by treatment with recombinant bovine
somatotropin: a dose-response study. J Reprod Fertil 1997; 110:91–97.
40. Mann GE, Lamming GE, Robinson RS, Wathes DC. The regulation
of interferon-
t
production and uterine hormone receptors during early
pregnancy. J Reprod Fertil 1999; 54(suppl):33–48.
41. Yuan W, Bao B, Garverick HA, Youngquist R, Lucy MC. Follicular
dominance in cattle is associated with divergent patterns of ovarian
gene expression for insulin-like growth factor (IGF)-I, IGF-II andIGF
binding protein-2 in dominant and subordinate follicles. Domest Anim
Endocrinol 1998; 15:55–63.
42. Perks CM, Peters AR, Wathes DC. Follicular and luteal expression of
insulin-like growth factors I and II and the type 1 IGF receptor in the
bovine ovary. J Reprod Fertil 1999; 116:157–165.
43. Thissen JP, Ketelslegers JM, Underwood LE. Nutritional regulation of
the insulin-like growth-factors. Endocrinol Rev 1994; 15:80–101.
