Aug 4, 2009 - differentiation in the pig (Knight, et al. 1977). ...... Sterle JA, Boyd CK, Peacock JT, Koenigsfeld AT, Lamberson WB, Gerrard DE & Lucy MC 1998.
Page 1 of 37
Accepted Preprint first posted on 4 August 2009 as Manuscript JOE-09-0131
1
Responses to maternal growth hormone or ractopamine during early-mid pregnancy are similar in
2
primiparous and multiparous pregnant pigs.
3 4
Authors: Kathryn L Gatford*†‡, Miles J de Blasio*†‡, Claire T Roberts*†‡, Mark B Nottle†‡, Karen L
5
Kind*†#, William H E J van Wettere#, Robert J Smits¶, Julie A Owens*†‡
6 7
* Research Centre for Early Origins of Health and Disease, Robinson Institute, University of
8
Adelaide SA 5005, Australia
9
†
Research Centre for Reproductive Health, Robinson Institute, University of Adelaide SA 5005,
10
Australia
11
‡
12
University of Adelaide SA 5005, Australia
13
#
14
of Adelaide, Roseworthy SA 5371, Australia
15
¶
16
Australia
Discipline of Obstetrics and Gynaecology, School of Paediatrics and Reproductive Health,
Discipline of Agricultural and Animal Science, School of Agriculture, Food and Wine, University
Research and Development Unit, QAF Meat Industries Ltd., Redlands Rd, Corowa NSW 2646,
17 18
Key words: growth hormone, ractopamine, fetal growth, muscle, placenta, pig
19 20
Running head: Fetal responses to maternal GH, ractopamine and parity
1 Copyright © 2009 by the Society for Endocrinology.
Page 2 of 37
21
Abstract
22 23
Fetal growth is restricted in primiparous pigs (gilts) compared to dams who have had previous
24
pregnancies (sows), as in other species. In gilts, daily maternal pGH injections from d 25 to 50 of
25
pregnancy (term ~115 d) increase fetal growth and progeny muscularity, and responses in sows are
26
unknown. Whether feeding the β2-adrenergic agonist ractopamine during this period increases
27
progeny growth rates in either parity, and fetal responses in gilts, have not been investigated. We
28
hypothesised that fetal and placental growth and fetal muscle development would be increased more
29
by maternal pGH and/or ractopamine during early-mid pregnancy in gilts than sows, since fetal
30
growth is restricted in gilts causing lower birth weights. Large White x Landrace gilts and sows
31
were injected daily with water (controls) or pGH (~15 µg.kg-1.d-1), or were fed 20 ppm
32
ractopamine, between d 25 and 50 of pregnancy. Maternal pGH increased litter average fetal weight
33
(11%, P=0.007) and length (3%, P=0.022), but not placental weight, at d 50 of pregnancy,
34
irrespective of parity, and had greatest effects in the heaviest fetuses of each litter. Maternal
35
ractopamine increased average fetal weight (9%, P=0.018), but not length. Muscle fibre diameter
36
was increased by pGH in heavy littermates, and by ractopamine in median littermates. Similar fetal
37
growth responses to pGH and ractopamine in gilts and sows suggest that these hormones increase
38
fetal nutrient availability similarly in both parities. We therefore predict that sustained pGH
39
treatment will increase progeny birth weight, postnatal growth and survival, in both sows and gilts.
2
Page 3 of 37
40
Introduction
41 42
Primiparity and adolescent pregnancy each restrict fetal growth in humans and other species,
43
including pigs (Ritter, et al. 1984) (Bryan & Hindmarsh 2006; Rasmussen & Fischbeck 1987).
44
These may both contribute to the reduced birth weight and poorer subsequent performance of
45
progeny of primiparous pigs (first pregnancy, gilts) compared to multiparous pigs (sows), since
46
fetal growth and birth weight predict neonatal survival and postnatal growth and metabolism in pigs
47
as in other species (Campbell & Dunkin 1982; Dwyer, et al. 1993; Fahmy & Bernard 1971;
48
Wigmore & Stickland 1983; Winters, et al. 1947). Multiple factors can restrict fetal growth and
49
birth weight in pigs; competing maternal demands in growing adolescent animals, large and
50
variable litter size, and restricted maternal nutrition used in commercial pig production systems
51
during pregnancy. Interventions to increase fetal growth may therefore be more effective in the gilt
52
than the multiparous sow, where fetal growth and development are less constrained.
53 54
Manipulation of nutrition in the pig shows that early-mid pregnancy is a critical period for fetal
55
growth and development, when increasing fetal growth leads to improved postnatal growth and
56
muscle gain (Dwyer, et al. 1994). This corresponds to the period of most rapid placental growth and
57
differentiation in the pig (Knight, et al. 1977). We and others have shown that daily maternal
58
injections of gilts with porcine growth hormone (pGH) during early-mid pregnancy increases fetal
59
growth (Gatford, et al. 2000; Kelley, et al. 1995), numbers of muscle fibres in fetuses or progeny at
60
birth (Rehfeldt, et al. 1993; Rehfeldt, et al. 2001b) and increases growth rates and muscle size of
61
their postnatal progeny (Gatford, et al. 2003; Kelley et al. 1995). Some studies have reported the
62
greatest responses to pGH are greatest in the smallest piglets within each litter (Rehfeldt & Kuhn
63
2006; Rehfeldt, et al. 2001a; Sterle, et al. 1995), suggesting that the effects of pGH are greatest in
64
the most restricted fetuses. Limited evidence from pigs and sheep also suggests that GH may act at
65
least in part by increasing placental growth and/or function (Harding, et al. 1997; Rehfeldt et al. 3
Page 4 of 37
66
2001a; Sterle et al. 1995; Wallace, et al. 2004). Whether pGH promotes fetal growth in sows with
67
lower competing nutrient demands for maternal growth is not known.
68 69
Feeding pregnant pigs with β2-adrenergic agonists has been suggested as an alternate endocrine
70
strategy to increase fetal and progeny growth and development. Progeny growth rates and carcass
71
weight, but not birth weight, muscle size nor muscle fibre numbers, were increased in sows fed with
72
the β2-adrenergic agonist ractopamine from d 25 to 50 of pregnancy (Hoshi, et al. 2005a; Hoshi, et
73
al. 2005b). Similarly, feeding sows the β2-adrenergic agonist salbutamol in the first third of
74
pregnancy increased muscle size and altered muscle fibre types of progeny (Kim, et al. 1994).
75
Whether these changes in progeny muscle development are preceded by increased fetal and
76
placental growth is not known, and parity differences in responses to β-adrenergic agonists have
77
also not been explored.
78 79
We therefore tested the hypotheses that daily pGH injections or feeding ractopamine during early-
80
mid pregnancy (from d 25 to 50) would promote fetal and placental growth and fetal muscle
81
development to a greater extent in gilts than in sows, and that these effects would be greatest in the
82
smallest fetuses in each litter.
83 84
Materials and Methods
85 86
Animals
87
The study was designed in accordance with the Australian Code of Practice for the Care and Use of
88
Animals for Scientific Purposes (National Health and Medical Research Council of Australia 1997)
89
and approved by the University of Adelaide Animal Ethics Committee. The in vivo studies were
90
conducted at the University of Adelaide Roseworthy Piggery. Twenty-four Large White x Landrace
91
gilts were mated at 23 weeks of age, and twenty-four mature Large White x Landrace, 3rd to 5th 4
Page 5 of 37
92
parity sows were mated at the first post-weaning estrus. Pigs were mated twice by artificial
93
insemination using semen from Landrace or Large White boars, either in the morning and afternoon
94
of the same day, or on the afternoon and morning of consecutive days. The day of second
95
insemination was taken as d 0 of pregnancy. All animals were individually housed in stalls
96
throughout the study.
97 98
Nutrition and treatments
99
Gilts and sows were fed ~1 kg of a dry sow diet (13.0 MJ DE/kg, 15.2% total protein, 0.65% total
100
lysine) on the day of mating (d 0) and d 1. From d 2 until the end of the study, gilts were fed 2.2
101
kg/d and sows were fed 2.5 kg/d of the same diet. Pregnancy was confirmed by ultrasound scanning
102
at d 23 of pregnancy, and dams were randomly allocated within parity groups to control, pGH-
103
injected or ractopamine-fed treatment groups (n=8 per treatment and parity). Control dams were
104
injected i.m. daily with 1 mL sterile water from day 25 to day 50 of pregnancy. pGH-injected gilts
105
and sows were injected i.m. daily with 1 mL sterile water containing 2.0 or 3.5 mg pGH
106
respectively (recombinant porcine GH, Reporcin, OzBiopharm Pty Ltd, Knoxfield, Vic, Australia),
107
which was calculated to provide a dose of ~15 µg pGH.kg-1.d-1 in both parity groups, based on
108
previous live weight data from the herd. Ractopamine-fed dams were fed ractopamine at 20 ppm in
109
the diet (added to the daily ration as Paylean, 2.2 g/day in gilts and 2.5 g/day in sows, Elanco
110
Animal Health, USA). One gilt (pGH-injected) returned to oestrus in the first week of the treatment
111
period and was removed from the study. Pregnant dams were weighed, and backfat depth measured
112
by ultrasound at the P2 site (110 mm from the midline over the 13th rib, using B-mode live
113
ultrasound), at the start of treatments and on the day of post-mortem (Gatford et al. 2000).
114 115
Postmortem
116
Maternal blood was collected by jugular venepuncture into EDTA tubes on the morning of post-
117
mortem (day 50 of gestation), and placed on ice. Dams were then humanely killed, and a 5
Page 6 of 37
118
hysterectomy performed. Number of fetuses and visible resorption sites were counted, and
119
individual fetuses and their intact placentae were dissected from the uterus. Placentae were cut
120
lengthways, laid out flat, and digitally photographed on a board marked with a 1 cm grid for later
121
measurement of placental area, made by tracing the perimeter of the placenta in VideoPro software
122
(Leading Edge software, Adelaide). Fetal blood was collected into EDTA tubes from the umbilical
123
cord immediately after removal of each fetus, and placed on ice. Blood was centrifuged and plasma
124
collected and stored frozen at -20C for later analyses. Weights, crown-rump length, abdominal
125
circumference, head width and sex of each fetus were measured and recorded. The fetal liver was
126
dissected and weighed, and the left fetal hindlimb was removed and a mid-femur cross-section was
127
taken through the upper hindlimb. Fetal muscle sections were fixed in 4% paraformaldehyde in PBS
128
for 24 h, then washed in PBS, and embedded in paraffin wax. The maternal ovaries were removed
129
and number of corpora lutea counted as a measure of ovulation rate. The proportion of fetuses plus
130
visible implantations was calculated as the number of fetuses plus resorption sites divided by the
131
number of corpora lutea. Fetal survival was calculated as the number of fetuses divided by the
132
number of corpora lutea. Runt piglets were defined as fetuses with weight more than 2SD below the
133
mean weight calculated using the mean and SD of weight for their litter.
134 135
Hormone and metabolite analyses
136
Growth hormone concentrations in maternal plasma were measured by radioimmunoassay. Growth
137
hormone purified from porcine pituitaries (Lot# AFP10888C) was obtained from the National
138
Hormone and Peptide Program (NHPP), NIDDK and Dr Parlow, and used as standard and for
139
iodination as tracer. Rabbit anti-porcine GH (Lot# AFP422801), also obtained from the NHPP, was
140
used as the primary antibody. All maternal samples were analysed for pGH in a single RIA with an
141
intra-assay CV of 9.9%. IGFs were extracted from plasma by size exclusion high performance
142
liquid chromatography at pH 2·5, using a modification of the original procedure (Scott & Baxter
143
1986), as described previously (Owens, et al. 1990). IGF-I concentrations were determined by RIA 6
Page 7 of 37
144
of neutralised chromatography fractions devoid of IGF binding proteins, using the Conlon rabbit
145
polyclonal antibody to human IGF-I (Francis, et al. 1989). All maternal plasma samples were
146
extracted in a single HPLC run with a recovery of 125I-IGF-I of 87.5%, and assayed in a single RIA,
147
with an intra-assay CV of 13.7% for a HPLC eluate fraction 3 pool containing 18 ng/ml IGF-I.
148
Covariance for extraction and assay of a maternal porcine plasma QC containing 174 ng/ml IGF-I
149
and included at the start, middle, and end of the HPLC run of maternal samples was 7.2% (n = 1
150
HPLC runs, 3 measurements). Insulin in maternal plasma was measured in a single assay using a
151
commercially-available RIA kit (Human insulin-specific RIA kit, HI-14K, Linco Research Inc, St
152
Charles, USA) with an intra-assay CV of 6.4% and 100% cross-reactivity with porcine insulin.
153
Glucose and urea in maternal and fetal plasma were measured by colorimetric enzymatic analysis
154
on a Hitachi 912 automated metabolic analyser using Roche/Hitachi Glucose/HK and UREA/BUN
155
kits respectively (Roche Diagnostics GmbH, Mannheim, Germany).
156 157
Fetal muscle histology
158
Fetal muscle fibre density and size were measured on the lightest (non-runt), median weight, and
159
heaviest fetus of each litter. Cross-sections (7 µm) of the fetal muscle were cut from embedded
160
blocks and mounted on electrostatically charged slides (Superfrost Plus, Menzel GmbH & Co KG,
161
Germany). Sections were stained with haemotoxylin-eosin, and images digitally-captured using a
162
40x lens and the NanoZoomer slide capture system (Hitachi, Japan). M. semitendinosus muscle
163
cross-sectional area, muscle fibre density and muscle fibre diameters were measured on all sections
164
where the M. semitendinosus was complete and fibres approximately perpendicular to the section.
165
Cross-sectional area was measured by tracing the M. semitendinosus perimeter in the NanoZoomer
166
viewing software (NDP View, Hitachi, Japan). For each section, twelve fields were captured at 20x
167
resolution for later counting of muscle fibres using random-systematic sampling, starting in the
168
upper left-hand quadrant and capturing fields with constant horizontal and vertical spacing. All
169
fibres (primary and secondary combined) were counted in an area of 0.1055 mm2 per field using 7
Page 8 of 37
170
VideoPro image analysis software (Leading Edge, Adelaide). Fibre diameter of primary fibres was
171
measured at x40 resolution using NDP View software for twenty fibres per field and 12 fields per
172
M. semitendinosus (total 240 fibres per fetus), with diameter measured on primary fibres falling
173
closest to points on a sampling grid.
174 175
Statistical analyses
176
One gilt returned to oestrus within 1 week of commencing treatment and was excluded from the
177
study. The effects of maternal parity, treatment and their interaction on maternal and litter outcomes
178
were analysed using a two-way analysis of variance model, with litter size (total number of fetuses)
179
included as a covariate. Specific contrasts were performed to test the a priori hypotheses that
180
maternal pGH treatment or ractopamine treatment would increase fetal growth and survival.
181
Hormone and metabolite data were excluded for two pGH-treated sows that were not injected on
182
the morning of post-mortem. Mixed model procedures were used to further evaluate weights of
183
individual fetuses and placentae and fetal:placental weight ratios, to allow analysis of effects of
184
fetal sex and also to compare responses to maternal factors in littermates of varying weight within
185
each litter. All fetuses were ranked and assigned to a quartile of weight within their litter. Sire breed
186
did not affect outcomes and was excluded as a factor in the models. These mixed models included
187
maternal parity, maternal treatment, fetal sex, fetal weight quartile, and their interactions, and litter
188
size was included as a covariate. Measures on each fetus within a litter were treated as repeated
189
measures on the dam, and multiple comparisons between treatment groups for repeated measures
190
models used the Bonferroni correction method. Where effects of parity or treatment varied with
191
fetal weight quartile, effects of sex, maternal treatment and maternal parity were evaluated
192
separately for fetuses of each weight quartile. Fetal M. semitendinosus characteristics were similarly
193
evaluated, using a repeated measures model including maternal parity, maternal treatment, fetal size
194
group (lightest, median or heaviest of litter), and their interactions. Fetal sex and litter size did not
195
affect M. semitendinosus characteristics and were excluded from these models. Data were log8
Page 9 of 37
196
transformed before analysis where necessary to achieve equal variances and normality. All tests
197
were carried out using SPSS v17.0 for Windows (SPSS Inc. USA) and are presented as mean ±
198
SEM for reproductive data and estimated mean ± SEM where litter size was included in the model,
199
unless otherwise indicated.
200 Results
201 202 203
Maternal weight and backfat
204
Prior to the start of treatments, maternal weight and maternal P2 backfat depth were greater in sows
205
than in gilts (P