Open Access

Asian-Australas J Anim Sci

Vol. 33, No. 5:722-731 May 2020
https://doi.org/10.5713/ajas.19.0334
pISSN 1011-2367 eISSN 1976-5517

Oxidative stress status and reproductive performance of sows

during gestation and lactation under different thermal environments
Yan Zhao1,2 and Sung Woo Kim2,*

*C orresponding Author: Sung Woo Kim Objective: Two experiments were conducted using 28 healthy multiparous sows to evaluate
Tel: +1-919-513-1494, Fax: +1-919-515-6884,
E-mail: sungwoo_kim@ncsu.edu the oxidative stress status and reproductive performance of sows during gestation and lacta­
tion under different thermal environments.
1
Department of Animal Science and Technology, Methods: Fourteen multiparous sows were used in Exp. 1 under a high thermal environment,
Shanxi Agricultural University, Taigu, Shanxi 030801,
China
and the other 14 multiparous sows were used in Exp. 2 under a moderate thermal environ­
2
Department of Animal Science, North Carolina State ment. In both experiments, reproductive performances of sows were recorded. Plasma
University, Raleigh, NC 27606, USA samples were collected on d 35, 60, 90, and 109 of gestation, and d 1 and 18 of lactation
ORCID
for malondialdehyde, protein carbonyls, 8-hydroxy-deoxyguanosine, immunoglobulin g
Yan Zhao (IgG), and IgM analysis.
https://orcid.org/0000-0002-4424-8329 Results: For sows in Exp. 1, plasma malondialdehyde concentration on d 109 of gestation
Sung Woo Kim
https://orcid.org/0000-0003-4591-1943
tended to be greater (p<0.05) than it on d 18 of lactation. Plasma concentration of protein
carbonyl on d 109 of gestation was the greatest (p<0.05) compared with all the other days.
Submitted Apr 23, 2019; Revised Jun 18, 2019; Plasma concentrations of 8-hydroxy-deoxyguanosine on d 109 of gestation was greater (p<
Accepted Sept 18, 2019 0.05) than d 18 of lactation in Exp. 1. For sows in Exp. 2, there was no difference of malondi­
aldehyde and protein carbonyl concentration during gestation and lactation. In both Exp.
1 and 2, litter size and litter weight were found to be negatively correlated with oxidative
stress indicators.
Conclusion: Sows under a high thermal environment had increased oxidative stress during
late gestation indicating that increased oxidative damage to lipid, protein, and DNA could
be one of the contributing factors for reduced reproductive performance of sows in this
environment. This study indicates the importance of providing a moderate thermal environ­
ment to gestating and lactating sows to minimize the increase of oxidative stress during late
gestation which can impair reproductive outcomes.

Keywords: Gestation; High Thermal Environment; Lactation; Oxidative Stress; Sow

INTRODUCTION
Reactive oxygen species (ROS) including superoxide anion (O2•–), hydroxyl radical (•OH),
and hydrogen peroxide (H2O2) are produced during biochemical processes within an
animal’s body [1]. If the balance between the production of ROS and antioxidant defense
is disturbed, the animal can suffer from oxidative stress. Accumulation of ROS could react
with biological macromolecules such as proteins, lipids, and nucleic acids, resulting in lipid
peroxidation, protein damage, and DNA damage [2,3]. In sows, external factors such as
social stress and environmental stressors can lead to increased oxidative stress [4,5]. Zhao
et al [5] studied the oxidative stress status of gestating and lactating sows under different
gestational housing systems and social ranks. The results showed that sows under intense
social stress could increase oxidative stress level during gestation and lactation [5]. Bottje
et al [4] showed that environmental stressors such as high dust and ammonia levels in poultry

Copyright © 2020 by Asian-Australasian Journal of Animal Sciences

This is an open-access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and repro-
722 duction in any medium, provided the original work is properly cited. www.ajas.info
Zhao and Kim (2020) Asian-Australas J Anim Sci 33:722-731

house environment can cause oxidative stress in lung lining the other group of 14 multiparous sows (initial BW, 243.2±
fluid of broilers. 30.3 kg; parity, 5.1±1.3; Landrace×Large White) were select­
Heat stress is another important environmental stressor ed and assigned to a moderate thermal environment on d
to animal’s reproductive performance especially during sum­ 35 of gestation. Sows from both experiments were housed
mer months in tropical and subtropical regions of the world. in the same gestation and farrowing buildings and managed
It has a significant negative impact on reproductive perfor­ with standard procedures. A negative pressure system was
mance of sows including impaired embryonic development used to ventilate the gestation building with four 1,100-rpm,
[6], reduced feed intake and milk production in sows [7,8], 61-cm wall fans. The farrowing building had an evaporative
and delayed puberty of gilts [9], which result in summer in­ cooling system and two 1,100-rpm, 61-cm wall fans. Water
fertility and productivity losses that amount to millions of flow through the evaporative cooling system was controlled
dollars. In the last decades, genetic improvement of sows for by an on-off thermostat which was set at 25°C, and one gas
high reproductive performance results in increased metabolic heater was provided during winter season in the farrowing
heat production which makes sows more susceptible to the room. The thermal environment inside the facilities was
high thermal environment [10,11]. Previous studies found monitored hourly using data loggers (Tinytag Plus 2, TPG-
that sows were under severe catabolic status during late ges­ 4500, Gemini Data Loggers, Chichester, UK) to measure
tation and lactation [12,13], and their oxidative stress levels temperature and relative humidity. Average hourly ambient
increased during the same period [14]. However, the nega­ temperature and temperature-humidity index (THI) with­
tive impact of high thermal environment on sow’s oxidative in a day in the gestation and farrowing buildings in Exp. 1
stress status have not been well studied, and it is not known and 2 were showed in Figure 1 and 2, respectively. In Exp.
if oxidative stress status is related to reproductive performance 1, sows were exposed to an ambient temperature above
of sows under different thermal environments. Therefore, we 25°C for an average of 17 h and 14 h per day in gestation
hypothesized that oxidative stress status of sows could elevate building and farrowing building, respectively (Figure 1A).
during late gestation and lactation when sows were housed The hourly THI was ranged from 72 to 78 in the gestation
under high thermal environment, and the elevated oxidative building and from 69 to 83 in the farrowing building (Figure
stress could be detrimental to the reproductive performance 1B). On the other hand, sows in Exp. 2 were kept in a moder­
of sows. Thus, the objective of this study was to investigate ate thermal environment as shown in Figure 2. The hourly
oxidative stress status and its relationship with reproductive ambient temperature was from 12°C to 17°C in the gestation
performance of sows during gestation and lactation under building and 19°C to 21°C in the farrowing building (Figure
moderate and high thermal environments. 2A). The hourly THI was ranged from 54 to 61 in the ges­
tation building and from 65 to 67 in the farrowing building
MATERIALS AND METHODS (Figure 2B).
The estrus of weaning sows in these 2 experiments was
Animal care detected daily by the immobilization response as a reaction
Procedures used in this study were reviewed and approved to a heat check boar after 3 d back to the breeding barn. Sows
by the North Carolina State University Animal Care and Use were artificially inseminated twice (12 h apart) after estrus
Committee. onset. Each insemination was conducted within 18 h of se­
men collection (Duroc semen). Pregnancy was detected and
Animals and experimental design confirmed at d 30 post-breeding using an ultrasound scanner
Two experiments were carried out using 28 healthy multip­ (VSS700 EZ Preg Checker, Veterinary Sales & Service Inc.,
arous sows at the North Carolina State University Swine Stuart, FL, USA). In the gestation building, sows were indi­
Educational Unit (Raleigh, NC, USA), in the summer (Exp. vidually housed in gestation crates (2.0×0.64 m) with semi-
1, June to August with average daily maximum and minimum automatic feeders, individual drinkers, and slatted flooring.
temperature of 30.3°C±2.9°C and 24.8°C±2.2°C in the ges­ They were fed 2.2 kg gestation diet daily which contained
tation building, and 30.9°C±2.6°C and 22.1°C±1.8°C in the 13.3% crude protein (CP) and 3.3 Mcal metabolizable energy
farrowing building, respectively) and winter (Exp. 2, Novem­ (ME)/kg (Table 1). Water was ad libitum during gestation and
ber to January with the average daily maximum and minimum lactation. On d 108 of gestation, all of the sows were weighed
temperature of 16.7°C±°C and 11.9°C±3.0°C in the gesta­ and moved to farrowing crates (2.1×1.5 m) in an adjacent
tion building, and 22.3°C±2.0°C and 20.4°C±2.1°C in the farrowing building. Three sows in Exp. 1 and 2 sows in Exp
farrowing building, respectively). In Exp. 1, 14 multiparous 2. did not maintain their pregnancy and did not farrow. Dur­
sows (initial body weight [BW], 245.7±38.7 kg; parity, 5.8±3.2; ing lactation, sows were fed ad libitum by electronic feeders
Landrace×Large White) were selected and assigned to a (JYGA Technologies, Saint-Nicolas, QC, Canada) in the far­
high thermal environment on d 35 of gestation. In Exp. 2, rowing building. Lactation diet contained 15.8% CP and 3.5

www.ajas.info  723
 Zhao and Kim (2020) Asian-Australas J Anim Sci 33:722-731

(A)

(B)

Figure 1. Average hourly temperature (A) and temperature-humidity index (THI) (B) in a day from June to August for sows in a high thermal environment in Exp. 1.
Average daily minimum and maximum temperatures of 24.8°C±2.2°C, 30.3°C±2.9°C in gestation building, and 22.1°C±1.8°C, 30.9°C±2.6°C in farrowing building,
respectively.

Mcal ME/kg (Table 1). The gestation and lactation diets were Sarstedt, Newton, NC, USA; Air-Tite Product Co. Inc., Vir­
formulated to meet or exceed NRC nutrient requirements ginia Beach, VA, USA). Plasma samples were obtained by
[15], and formulations were the same in Exp. 1 and 2. In both centrifugation (5810 R, Eppendorf AG, Hamburg, Germany)
experiments, any feed left from the previous day was removed at 3,000 g, 15 min, 4°C, then allocated into 1.5 mL micro­
and weighed around 0800 h, and the difference between the centrifuge tubes, kept in liquid nitrogen for 1 h, and stored
amount of feed provided and remained was used as an esti­ at –80°C until analysis.
mate of daily feed intake. The BW of sows was measured on
d 35 and 109 of gestation, and d 1 and 18 of lactation. Backfat Analysis of oxidative stress indicators
thickness (10th rib, 6 cm off-midline) of each sow was mea­ Plasma samples from Exp. 1 and 2 were used to measure con­
sured on d 1 and 18 of lactation using an ultrasound scanner centrations of malondialdehyde (MDA), protein carbonyl,
(VSS700 EZ Preg Checker, Veterinary Sales & Service Inc., and 8-hydroxy-deoxyguanosine (8-OHdG). Concentrations
USA). After farrowing, individual weight of piglet and litter of MDA were measured using the thiobarbituric acid reactive
size of sows on d 1 (total number of piglets, born alive, still­ substances assay kit (Cell Biolabs, San Diego, CA, USA)
born, and mummy) and 18 of lactation were measured. according to the method described by Zhao et al [5]. Plasma
samples and MDA standards were incubated and reacted
Blood sampling with thiobarbituric acid at 95°C. After centrifuge and buta­
In both Exp. 1 and 2, blood was sampled via jugular veni­ nol extraction, samples and standards were read at 532 nm
puncture from all of the sows restrained by snout snare by a spectrophotometric plate reader (Synergy HT, BioTek
between 0900 and 1000 h on d 35, 60, 90, and 109 of gesta­ Instruments, Winooski, VT, USA). Samples were quantified
tion in the gestation building, and d 1 and 18 of lactation against the standard curve which was constructed by MDA
in the farrowing building. Blood was collected using 9 mL standards. The detective limit of MDA analysis was 0.98 μM.
ethylenediaminetetraacetic acid-coated syringes and dispos­ Protein concentration in each plasma sample was measured
able 16-gauge×0.1-mm hypodermic needles (MONOVETTE, by a bicinchoninic acid protein assay kit (Pierce Biotech­

www.ajas.info
724  
Zhao and Kim (2020) Asian-Australas J Anim Sci 33:722-731

(A)

Hours

(B)

Hours
Figure 2. Average hourly temperature (A) and temperature-humidity index (THI) (B) in a day from November to January for sows in a moderate thermal environment in
Exp. 2. Average daily minimum and maximum temperatures of 11.9°C±3.0°C, 16.7°C±3.5°C in gestation building, and 20.4°C±2.1°C, 22.3°C±2.0°C in farrowing
building, respectively.

nology, Rockford, IL, USA). Then all plasma samples were ma samples and 8-OHdG standards were first added into a
diluted with bicinchoninic acid to reach protein concentra­ 96-well plate. Then an anti-8-OHdG monoclonal antibody
tion at 10 μg/mL before measuring protein carbonyl. Protein was added, followed by adding a secondary antibody. A sub­
carbonyl concentration in each diluted plasma sample was strate solution and stop solution were added. The absorbance
measured via the protein carbonyl ELISA kit (Cell Biolabs, of each well was read at 450 nm by a spectrophotometric
USA) according to the method described by Shen et al [16]. plate reader (Synergy HT, BioTek Instruments, USA). Con­
The protein carbonyl presented in the sample or standard centrations of 8-OHdG in plasma samples were quantified
was first derivatized to dinitrophenyl hydrazine and probed against the standard curve which was constructed by 8-OHdG
with an anti-dinitrophenyl antibody, then incubated with a standards. The detective limit for 8-OHdG was 0.078 ng/mL.
secondary antibody. Finally substrates and stop solutions
were added. Standards and samples were read at 450 nm Immunoglobulin evaluation
by a spectrophotometric plate reader (Synergy HT, BioTek In both experiments, concentrations of immunoglobulin G
Instruments, USA). Protein carbonyl concentrations in (IgG) and IgM in sows’ plasma on d 109 of gestation, d 1 and
samples were quantified against the standard curve which 18 of lactation were measured by ELISA kits (Bethyl, Mont­
was drawn by protein carbonyl standards. The detective gomery, TX, USA) according to the method described by
limit for protein carbonyl was 0.375 nmol/mg. Chaytor et al [18]. Briefly, goat anti-pig IgG or IgM were used
Concentrations of 8-OHdG in plasma samples were measured as capture antibodies to coat wells. All of the samples were
using the oxidative DNA damage enzyme-linked immuno­ diluted to 1:100,000 for IgG and IgM measurement. Horse­
sorbent assay (ELISA) kit (Cell Biolabs, USA) according to radish peroxidase goat anti-pig IgG or IgM was used as the
the method described by Weaver and Kim [17]. Briefly, plas­ detection. The plate was read at 450 nm by an ELISA plate

www.ajas.info  725
 Zhao and Kim (2020) Asian-Australas J Anim Sci 33:722-731

Table 1. Composition of gestation and lactation diets (as-fed basis) in Exp. 1 reader (Synergy HT, BioTek Instruments, USA) and software
and 2 (KC4 Data Analysis Software, BioTek Tnstrument, USA).
Item Gestation Lactation Sample concentration was quantified against the standard
Ingredient curve which was drawn by standards. Detective limits were
Corn, yellow (%) 81.30 74.00 7.8 ng/mL for IgG, and 15.6 ng/mL for IgM, respectively.
Soybean meal, 48% CP (%) 13.85 19.60
Poultry fat (%) 1.00 3.99 Statistical analysis
L-Lys (%) 0.00 0.25 Exp. 1 and 2 were completely randomized designs. Data from
L-Thr (%) 0.00 0.01
oxidative stress indicators and immunoglobulin assay were
Limestone (%) 1.11 1.08
compared among different gestation and lactation days in
Dicalcium phosphate (%) 2.05 2.38
Salt (%) 0.50 0.50
both experiments using the MIXED procedure of SAS (SAS
Trace mineral permix1) (%) 0.15 0.15 Inst. Inc., Cary, NC, USA). Oxidative stress data from d 35 of
Vitamin permix2) (%) 0.04 0.04 gestation was used as a covariance when analyzing the oxida­
Total 100.00 100.00 tive stress indicators. The individual sow was the experimental
Calculated composition unit. Day was a fixed effect and the sow was a random effect.
DM (%) 89.7 90.1 The canonical correlation between reproductive performance
ME (Mcal/kg) 3.3 3.5 and oxidative stress indicators on different days in Exp. 1 and
CP (%) 13.3 15.8 2 were analyzed by the CANCORR procedure of SAS (SAS
Lys (%) 0.63 0.92 Inst., Inc., USA), respectively. Data was considered statisti­
Met (%) 0.49 0.51
cally different when probability values were less than 0.05,
Trp (%) 0.14 0.17
Thr (%) 0.49 0.55
and probability less than 0.1 and equal or greater than 0.05
Ca (%) 1.03 1.12 was considered as a trend.
Total P (%) 0.69 0.76
CP, crude protein; DM, dry matter; ME, metabolizable energy. RESULTS
1)
The trace mineral premix provided per kilogram of complete diet: 3.96 mg of
Mn as manganous oxide; 16.5 mg of Fe as ferrous sulfate; 16.5 mg of Zn as Oxidative stress parameters
zinc sulfate; 1.65 mg of Cu as copper sulfate; 0.30 mg of I as ethylenediamine
Plasma concentrations of MDA, protein carbonyl, and 8-OHdG
dihydroiodide; and 0.30 mg of Se as sodium selenite.
2)
The vitamin premix provided per kilogram of complete diet: 8,228 IU of vitamin were compared among days during gestation and lactation
A as vitamin A acetate; 1,173 IU of vitamin D3; 47 IU of vitamin E; 0.03 mg of in Exp. 1 and 2, respectively (Table 2). For sows in Exp. 1,
vitamin B12; 5.88 mg of riboflavin; 23.52 mg of D-pantothenic acid as calcium plasma MDA concentration on d 109 of gestation tended to
panthonate; 35.27 mg of niacin; 0.24 mg of biotin; 1.76 mg folic acid; 3.88 mg
menadione. be greater (p = 0.096) than it on d 18 of lactation (Table 2).

Table 2. Oxidative stress indicators of sows under a high thermal environment (HT)1) in Exp. 1 and a moderate thermal environment (MT)1) in Exp. 2
Gestation Lactation
Item 2)
SEM p-value
d 35 d 60 d 90 d 109 d1 d 18
Exp. 1
n3) 11 11 11 11 11 11
Malondialdehyde (μM) 6.14 6.39AB 7.51AB 8.54A 8.39AB 6.02B 1.05 0.096
Protein carbonyl (nmol/mg) 1.41 1.00a 1.06a 2.29c 1.40ab 1.76b 0.13 < 0.001
8-hydroxy-deoxyguanosine (ng/mL) 0.32 0.59ab 0.61ab 0.95b 0.62ab 0.29a 0.12 0.006
Exp. 2
n3) 12 12 12 12 12 12
Malondialdehyde (μM) 4.42 4.63 4.00 2.56 3.53 5.26 0.98 0.285
Protein carbonyl (nmol/mg) 1.59 1.50 0.72 1.03 1.29 0.76 0.22 0.122
8-hydroxy-deoxyguanosine (ng/mL) 0.71 1.00b 0.90b 1.07b 1.04b 0.48a 0.12 < 0.001
SEM, standard error of the mean.
1)
HT, high thermal environment (average daily minimum and maximum temperatures of 24.8°C ± 2.2°C, 30.3°C ± 2.9°C in gestation building, and 22.1°C ± 1.8°C,
30.9°C ± 2.6°C in farrowing building, respectively); MT, moderate thermal environment (average daily minimum and maximum temperatures of 11.9°C ± 3.0°C,
16.7°C ± 3.5°C in gestation building, and 20.4°C ± 2.1°C, 22.3°C ± 2.0°C in farrowing building, respectively).
2)
Oxidative stress data from d 35 of gestation was used as a covariance when analyzing the oxidative stress indicators.
3)
Initial number of sows was 14. There were 3 sows in Exp. 1 and 2 sows in Exp. 2 did not maintain pregnancy and did not farrow which were excluded from the study.
a-c
Means within a row with different superscripts differ (p < 0.05).
A,B
Means within a row tend to differ (0.05 ≤ p < 0.10).

www.ajas.info
726  
Zhao and Kim (2020) Asian-Australas J Anim Sci 33:722-731

Plasma concentration of protein carbonyl on d 109 of gesta­ Table 3. Reproductive performance of sows under a high thermal environment
tion was the greatest (p<0.05) compared with all the other (HT)1) in Exp. 1 and a moderate thermal environment (MT)1) in Exp. 2
days (Table 2). Besides, protein carbonyl concentration on d Item Exp. 1 SD2) Exp. 2 SD2)
18 of lactation was greater (p<0.05) than concentrations on n3)
11 12
d 60 and 90 of gestation. Plasma concentrations of 8-OHdG Parity 5.8 3.2 5.1 1.3
on d 109 of gestation was greater (p<0.05) than d 18 of lac­ Body weight of sows (kg)
tation in Exp. 1 (Table 2). For sows in Exp. 2, there was no d 35 of gestation 246 39 243 30
difference of MDA concentration among gestation and lac­ d 109 of gestation 281 35 290 27
d 1 of lactation 275 32 276 28
tation days (Table 2). Protein carbonyl concentration did not
d 18 of lactation 273 29 261 31
differ among days (Table 2). However, plasma concentration
Body weight changes (kg)
of 8-OHdG on d 18 of lactation was smaller (p<0.05) than Gestation (d 35 to 109) 35 17 47 14
all the other days in Exp. 2 (Table 2). Lactation (d 1 to 18) –5 13 –13 17
Backfat of sows4) (mm)
Reproductive performance of sows d 1 of lactation 17.0 2.5 15.8 3.9
In Exp. 1, sows gained 35.0 kg from d 35 to 109 of gestation, d 18 of lactation 16.6 2.9 14.4 3.8
but lost 1.5 kg from d 1 to 18 of lactation. Sows consumed Change from d 1 to 185) –0.4 2.7 –1.4 1.6
4.3±1.1 kg feed daily during lactation. Sows had 9.5±2.9 pigs ADFI of sows (kg) 4.3 1.1 4.6 1.1
born alive per litter, and weaned 7.4±2.1 pigs per litter at the Litter size (pig)
end of lactation. Litters gained 27.1±8.7 kg from d 1 to 18 of d 1, total born 12.5 2.9 13.6 2.2
d 1, born alive 9.5 2.9 11.9 1.8
lactation. Piglets gained 204.4±41.3 g daily during lactation
d 1, stillborn 2.5 1.9 1.2 1.2
(Table 3). d 1, mummy 0.5 0.8 0.4 0.9
In Exp. 2, sows gained 46.6 kg from d 35 to 109 of gestation, d 18 7.4 2.1 10.4 1.2
but lost 14.7 kg from d 1 to 18 of lactation. Sows consumed Change from d 1 to 185) –2.2 1.7 –1.5 1.7
4.6±1.1 kg feed daily during lactation. Sows had 11.9±1.8 Litter weight (kg)
pigs born alive per litter, and weaned 10.4±1.2 pigs per litter d 16) 15.0 3.7 17.9 2.9
at the end of lactation. Litters gained 37.5±9.5 kg from d 1 to d 18 42.1 11.7 55.4 11.2
18 of lactation. Piglets gained 199.5±41.2 g daily during lac­ Gain from d 1 to 185) 27.1 8.7 37.5 9.5
tation (Table 3). Piglet weight (kg)
d 17) 1.66 0.42 1.52 0.20
d 18 5.78 0.96 5.42 0.78
Immunoglobulin evaluation
ADG from d 1 to 18 (g/d) 204 41 200 41
Plasma concentrations of IgG and IgM were compared among
SD, standard deviation; ADFI, average daily feed intake; ADG, average daily gain.
d 109 of gestation, and d 1 and 18 of lactation in Exp. 1 and 1)
HT, high thermal environment (average daily minimum and maximum
2, respectively (Table 4). For sows in Exp. 1, there was no dif­ temperatures of 24.8°C ± 2.2°C, 30.3°C ± 2.9°C in gestation building, and
ference for plasma concentrations of IgG and IgM on d 109 22.1°C ± 1.8°C, 30.9°C ± 2.6°C in farrowing building, respectively); MT, moder-
of gestation, and d 1 and 18 of lactation (Table 4). For sows ate thermal environment (average daily minimum and maximum temperatures
of 11.9°C ± 3.0°C, 16.7°C ± 3.5°C in gestation building, and 20.4°C ± 2.1°C,
in Exp. 2, plasma concentration of IgM did not differ among 22.3°C ± 2.0°C in farrowing building, respectively).
different days. However, plasma concentration of IgG on d 2)
Results are expressed as mean with SD.
3)
18 of lactation was greater (p<0.05) than it on d 1 of lactation Initial number of sows was 14 for both Exp. 1 and 2. There were 3 sows in Exp.
and 2 sows in Exp. 2 did not maintain pregnancy and did not farrow which were
(Table 4). excluded from the study.
4)
Measured at the P2 position (locate at left side of the10th rib, and 6 cm away
Canonical correlation analysis between oxidative stress from the spine).
5)
Difference between d 1 and 18 of lactation
and reproductive performance 6)
Litter weight at birth includes piglets born alive only.
Three sets of canonical relationships between oxidative stress 7)
Weight of piglets born alive.
indicators and reproductive performance of sows were ob­
tained in Exp. 1 (Table 5). The first set of variables are oxidative
stress indicators including concentrations of MDA, protein the other two sets showed tendency (p = 0.093 and 0.059),
carbonyl, and 8-OHdG on d 60 and 90 of gestation, and d18 suggesting that there is a significant canonical correlation
of lactation (Md60, Pd60, Od60, Md90, Pd90, Od90, Md18, Pd18, and between oxidative stress and reproductive performance of
Od18). The second set of variables are BW and backfat of sows, sow in Exp. 1 (Table 5). From the linear expression of canoni­
and piglet weight on d 3 and 18 of lactation (X1, X2, X3, X4, cal variable composition in the first set of canonical relationship,
X5, and X6). The correlation coefficients in the first set of ca­ Md60, Od60, and X2 were relatively large (Table 5), showing that
nonical relationship reached a significant level (p<0.05), and plasma concentration of MDA and 8-OHdG on d 60 of ges­

www.ajas.info  727
 Zhao and Kim (2020) Asian-Australas J Anim Sci 33:722-731

Table 4. Immunological parameters of sows under a high thermal environment concentration of MDA and protein carbonyl on d 90 of
(HT)1) in Exp. 1 and a moderate thermal environment (MT)1) in Exp. 2 gestation are negatively correlated with backfat of sows on
Plasma d 18 of lactation. In the third set of canonical relationship,
Item d 109 of d 1 of d 18 of SEM p value there was a large load on Md18, Pd18, and X6 (Table 5), sug­
gestation lactation lactation gesting that plasma concentration of MDA and protein
Exp. 1 carbonyl on d 18 of lactation are negatively correlated with
n2) 11 11 11 - - piglet weight on d 18 of lactation.
IgM (mg/mL) 3.25 2.55 2.33 0.41 0.419 Two sets of canonical relationships between oxidative stress
IgG (mg/mL) 19.38 17.98 19.67 1.69 0.898 indicators and reproductive performance of sows in Exp. 2
Exp. 2
and the composition of canonical variables are shown in
n2) 12 12 12 - -
Table 6. The first set of variables are oxidative stress indica­
IgM (mg/mL) 2.96 2.36 2.53 0.28 0.621
IgG (mg/mL) 19.20ab 16.22a 20.07b 1.19 0.031
tors including concentrations of MDA, protein carbonyl,
and 8-OHdG on d 90 of gestation, and d3 of lactation (Md90,
SEM, standard error of the mean; IgM, immunoglobulin M; IgG, immunoglobulin
G. Pd90, Od90, Md3, Pd3, and Od3). The second set of variables are
1)
HT, high thermal environment (average daily minimum and maximum tem- piglet weight on d 3 and 18 of lactation, litter size of born
peratures of 24.8°C ± 2.2°C, 30.3°C ± 2.9°C in gestation building, and alive and wean of sows (X5, X6, X7, and X8). The correlation
22.1°C ± 1.8°C, 30.9°C ± 2.6°C in farrowing building, respectively); MT, moder-
ate thermal environment (average daily minimum and maximum temperatures
coefficients in the first and second set of canonical relation­
of 11.9°C ± 3.0°C, 16.7°C ± 3.5°C in gestation building, and 20.4°C ± 2.1°C, ships reached a significant level (p<0.05), suggesting that
22.3°C ± 2.0°C in farrowing building, respectively).
2)
there is a significant canonical correlation between oxida­
Initial number of sows was 14. There were 3 sows in Exp. 1 and 2 sows in Exp.
tive stress and reproductive performance of sow in Exp. 2
2 did not maintain pregnancy and did not farrow which were excluded from the
study. (Table 6). From the linear expression of canonical variable
a,b
Means within a row with different superscripts differ (p < 0.05). composition in the first set of canonical relationship, Md3,
Od3, and X6 were relatively large (Table 6), showing that MDA
and 8-OHdG on d 3 of lactation are negatively correlated
tation are negatively correlated with BW of sows on d 18 of with piglet weight on d 18 of lactation. As can be seen from
lactation. As can be seen from the linear expression of canoni­ the linear expression of canonical variables in the second
cal variables in the second set of canonical relationship, Md90, set of canonical relationship, there was a large load on Pd90
Pd90, and X4 were relatively large (Table 5), showing that plasma and X8 (Table 6), suggesting that protein carbonyl on d 90

Table 5. Canonical correlations between oxidative stress indicators and reproductive performance of sows in Exp. 1
The second set of Canonical correlation
The first set of variables1) p-value Canonical variable composition
variables2) coefficient
Oxidative stress indicators Sow BW 0.990 0.023 V 1 = 0.924M d60 + 0.340P d60 + 0.652O d60
W 1 = 0.364X 1 – 1.222X 2
Oxidative stress indicators Sow backfat 0.879 0.093 V 1 = 1.211M d90 + 1.143P d90 + 0.621O d90
W 1 = 0.277X 3 – 1.026X 4
Oxidative stress indicators Piglet weight 0.960 0.059 V 1 = -0.988M d18 – 1.448P d18 – 0.413O d18
W 1 = 0.269X 5 + 0.821X 6
BW, body weight.
1)
The first set of variables are oxidative stress indicators including concentrations of malondialdehyde, protein carbonyl, and 8-hydroxy-deoxyguanosine on d 60 and 90 of
gestation, and d18 of lactation (M d60, P d60, O d60, M d90, P d90, O d90, M d18, P d18, and O d18).
2)
The second set of variables are BW and backfat of sows, and piglet weight on d 3 and 18 of lactation (X 1, X 2, X 3, X 4, X 5, and X 6).

Table 6. Canonical correlations between oxidative stress indicators and reproductive performance of sows in Exp. 2
The second set of Canonical correlation
The first set of variables1) p-value Canonical variable composition
variables2) coefficient
Oxidative stress indicators Piglet weight 0.999 0.024 V 1 = 0.482M d3 – 0.215P d3 + 0.823O d3
W 1 = –0.187X 5 – 0.986X 6
Oxidative stress indicators Sow litter size 0.999 0.023 V 1 = 0.181M d90 – 1.029P d90 + 0.077O d90
W 1 = 0.257X 7 + 0.889X 8
1)
The first set of variables are oxidative stress indicators including concentrations of malondialdehyde, protein carbonyl, and 8-hydroxy-deoxyguanosine on d 90 of gestation,
and d 3 of lactation (M d90, P d90, O d90, M d3, P d3, and O d3).
2)
The second set of variables are piglet weight on d 3 and 18 of lactation, and litter size of born alive and wean of sows (X 5, X 6, X 7, and X 8).

www.ajas.info
728  
Zhao and Kim (2020) Asian-Australas J Anim Sci 33:722-731

of gestation is negatively correlated with litter size of sows hand, feed intake of lactating sows was not the main factor
on d 18 of lactation. for the reduced milk production in Exp.1. Because feed in­
take in Exp. 1 was similar to that in Exp. 2, and it did not
DISCUSSION show significant reduction as some other studies have found
[23]. This may be because it took a few days for sows to adjust
In this study we have confirmed that sows under a high ther­ the electronic feeders in the farrowing room.
mal environment had increased oxidative stress during late Besides the reduced reproductive performance of sows in
gestation, and the increased oxidative damage to lipid, pro­ Exp. 1, heat stress environment also affected the oxidative
tein, and DNA could be one of the contributing factors for stress status of sows indicated by the enhanced oxidative
reduced reproductive performance of sows under the high damage to lipid, protein, and DNA during late gestation. The
thermal environment. Hence, we accepted the hypothesis plasma MDA, protein carbonyl, and 8-OHdG concentra­
that oxidative stress status of sows could elevate during late tion increased during late gestation which showing that sows
gestation and lactation when sows are housed under high could suffer severe oxidative damage during late gestation
thermal environment, and the elevated oxidative stress could under heat stress environment. The current results indicate
be detrimental to the reproductive performance of sows. that oxidative stress could be one of stress responses caused
In this study, sows in Exp. 1 were under heat stress condi­ by high thermal environment. On the other hand, sow in
tion indicated by the observed ambient temperatures and THI. Exp. 2 did not show the similar pattern with Exp. 1. Although
We observed that the ambient temperature in Exp. 1 were sows had increased oxidative DNA damage during gesta­
above 25°C for an average of 17 h/d in the gestation building tion and early lactation compared with late lactation, other
and 14 h/d in the farrowing building. Environmental tem­ oxidative indicators for protein and lipid did not enhanced,
perature remaining above 25°C could cause heat stress to sows indicating that sows in Exp. 2 did not suffer severe oxidative
[19]. Meanwhile, high levels of THI (72 to 78 in the gestation stress during gestation and lactation when they were housed
building, and 69 to 83 in the farrowing building) was also under moderate thermal environment. Previous studies
observed in Exp. 1. The THI is a combination of tempera­ showed that sows were under severe catabolic status during
ture and humidity which has been used to indicate heat stress late gestation and lactation which could induce oxidative
environment. It was reported that the THI value less than damage and decrease antioxidants concentration at the same
74 was classified as a safe, whereas critical when greater than period [12-14]. Studies also found that the ROS production
74 in sows [20]. The observed high THI in Exp. 1 can be detri­ increased under the high temperature treatment [24,25],
mental to sows, because the high temperature and high which could explain why sows in Exp. 1 showed enhanced
humidity together can limit their heat loss through evapo­ oxidative stress. However, the molecular mechanisms respon­
ration [21]. These data indicated that sows in Exp. 1 were sible for excessive ROS production or lowered antioxidant
under heat stress condition. On the other hand, based on defenses under heat stress are still not known in sows. There
the average daily maximum and minimum temperatures are some reports in poultry which found that heat stressed
and THI recorded in Exp. 2, it seems that sows in Exp. 2 chickens exhibit overproduction of mitochondrial ROS in
were kept within their thermos-neutral environment. skeletal muscle, which might result from enhanced sub­
Our Exp. 1 was conducted from June to August in North strate oxidation and downregulation of avian uncoupling
Carolina, which we observed the reduced reproductive per­ protein [26,27].
formance of sows indicated by the less number of total born, As discussed above, heat stress environment negatively
born alive, more still born piglets, and small litter size and affected the reproductive performance and enhanced the
litter weight gain. These results indicated that the reproduc­ oxidative damage to sows during late gestation and lactation.
tive performance of sows in Exp. 1 was negatively affected We hypothesized that the elevated oxidative stress could be
by the heat stress environment. The smaller litter size of sow detrimental to the reproductive performance of sows. There­
in Exp. 1 contributed to reduce mobilizing of their body re­ fore, we investigated the relationship between oxidative stress
serve for less milk production, which may explain why sows indicators and reproductive performance of sows during
in Exp. 1 lost less BW and backfat during lactation com­ gestation and lactation under both moderate and high ther­
pared with sows in Exp. 2. Similar result was reported that mal environments. The canonical correlation coefficients were
the over-all reproductive performance of sows reduced in close to 1 for all the sets of canonical relationships, and the
large confinement units in North Carolina during hot months linear expression of canonical variable compositions showed
from June to October [22].These findings were also consis­ that there were negative correlations between oxidative stress
tent with other studies which showed that sows under high indicators and piglet weight, BW, backfat, and litter size of
thermal environment during gestation farrowed fewer born sows in current studies. These results indicates a strong cor­
alive with increased number of stillborn [8]. On the other relation between oxidative stress indicators and reproductive

www.ajas.info  729
 Zhao and Kim (2020) Asian-Australas J Anim Sci 33:722-731

performance of sows. Similar results were reported in pre­ REFERENCES

vious study, which showed that oxidative damages to lipid,
protein, and DNA were negatively correlated with repro­ 1. Betteridge DJ. What is oxidative stress? Metabolism 2000;49:
ductive performance of gestating sows under social stress 3-8. https://doi.org/10.1016/S0026-0495(00)80077-3
[5]. Excessive ROS could negatively affect multiple physio­ 2. Gutteridge JM. Lipid peroxidation and antioxidants as bio­
logical processes from oocyte maturation to fertilization, markers of tissue damage. Clin Chem 1995;41:1819-28. https://
embryo development and pregnancy [28]. Study on pregnant doi.org/10.1093/clinchem/41.12.1819
women showed that oxidative stress can increase membrane 3. Hall ED. Efficacy and mechanisms of action of the cytopro­
damage which is related with fetal growth retardation and tective lipid peroxidation inhibitor tirilazad mesylate in sub­
cause a higher risk of prenatal mortality [29]. arachnoid haemorrhage. Eur J Anaesthesiol 1996;13:279-89.
In Exp. 1, plasma concentrations of IgG and IgM were not https://doi.org/10.1046/j.1365-2346.1996.00980.x
different among days, indicating that sows did not suffer im­ 4. Bottje WG, Wang S, Kelly FJ, Dunster C, Williams A, Mudway
mune challenge during the study. The plasma concentration I. Antioxidant defenses in lung lining fluid of broilers: impact
of IgG on d 1 of lactation in Exp. 2 was smaller than it on d 18 of poor ventilation conditions. Poult Sci 1998;77:516-22. https://
of lactation, but both of them were within the normal range doi.org/10.1093/ps/77.4.516
[30]. The decreased plasma concentration of IgG on d 1 of 5. Zhao Y, Flowers B, Saraiva A, Yeum KJ, Kim SW. Effect of
lactation may be because more IgG in the blood were secreted social ranks and gestation housing systems on oxidative stress
into the colostrum. status, reproductive performance, and immune status of sows.
J Anim Sci 2014;91:5848-58. https://doi.org/10.2527/jas.2013-
CONCLUSION 6388
6. Renaudeau D, Noblet J. Effects of exposure to high ambient
Taken together, sows under a high thermal environment had temperature and dietary protein level on sow milk produc­
increased oxidative stress during late gestation. These oxidative tion and performance of piglets. J Anim Sci 2001;79:1540-8.
stress indicators were negatively correlated with reproductive https://doi.org/10.2527/2001.7961540x
performance of sows, suggesting that increased oxidative 7. Black JL, Mullan BP, Lorschy ML, Giles LR. Lactation in the
damage to lipid, protein, and DNA could be one of the con­ sow during heat stress. Livest Prod Sci 1993;35:153-170. https://
tributing factors for reduced reproductive performance of doi.org/10.1016/0301-6226(93)90188-N
sows under the high thermal environment. This study in­ 8. Renaudeau D, Quiniou N, Noblet J. Effects of exposure to
dicates the importance of providing a moderate thermal high ambient temperature and dietary protein level on per­
environment to gestating and lactating sows to minimize formance of multiparous lactating sows. J Anim Sci 2001;79:
the increase of oxidative stress during late gestation which 1240-9. https://doi.org/10.2527/2001.7951240x
can impair reproductive outcomes. Our findings in this study 9. Flowers B, Day BN. Alterations in gonadotropin secretion
also shows the importance for future studies to regulate oxi­ and ovarian function in prepubertal gilts by elevated environ­
dative stress status of sow during gestation and lactation mental temperature. Biol Reprod 1990;42:465-71. https://doi.
under heat stress. org/10.1095/biolreprod42.3.465
10. Zhang H, Wang Z, Liu G, He J, Su C. Effect of dietary fat
CONFLICT OF INTEREST sup­plementation on milk components and blood parameters
of early-lactating cows under heat stress. Slovak J Anim Sci
We certify that there is no conflict of interest with any financial 2011;44:52-8.
organization regarding the material discussed in the manu­ 11. Wegner K, Lambertz C, Das G, Reiner G, Gauly M. Climatic
script. effects on sow fertility and piglet survival under influence of
a moderate climate. Animal 2014;8:1526-33. https://doi.org/
ACKNOWLEDGEMENTS 10.1017/S1751731114001219
12. Ji F, Wu G, Blanton JR, Kim SW. Weight and compositional
This study was supported by grants from North Carolina Pork changes in pregnant gilts and its implication to nutrition. J
Council (10M69MF1), Shanxi Agriculture University Science Anim Sci 2005;83:366-75.
& Technology Innovation Fund (2017YJ07), and Key Projects 13. Kim SW, Wu G. Regulatory role for amino acids in mammary
of Shanxi Key R&D Program (201703D211001). The authors gland growth and milk synthesis. Amino Acids 2009;37:89-
wish to thank Dr. Alysson Saraiva and Yunlong Shi for their 95. https://doi.org/10.1007/s00726-008-0151-5
appreciable contribution to sampling. 14. Berchieri-Ronchi CB, Kim SW, Zhao Y, Correa CR, Yeum
KJ, Ferreira ALA. Oxidative stress status of highly prolific
sows during gestation and lactation. Animal 2011;5:1774-9.

www.ajas.info
730  
Zhao and Kim (2020) Asian-Australas J Anim Sci 33:722-731

https://doi.org/10.1017/S1751731111000772 for animal production. London, UK: Butterworths; 1981. p.

15. Committee on Nutrient Requirements of Swine, National 3-17.
Research Council. Nutrient requirements of swine. 11th ed. 24. Ozawa M, Hirabayashi M, Kanai Y. Developmental com­
Washington, DC, USA: National Academy Press; 2012. petence and oxidative state of mouse zygotes heat-stressed
16. Shen YB, Weaver AC, Kim SW. Effect of feed grade L-methio­ maternally or in vitro. Reproduction 2002;124:683-9. https://
nine on growth performance and gut health in nursery pigs doi.org/10.1530/rep.0.1240683
compared with conventional DL-methionine. J Anim Sci 25. Matsuzuka T, Ozawa M, Nakamura A, Ushitani A, Hirabayashi
2014;92:5530-9. https://doi.org/10.2527/jas.2014-7830 M, Kanai Y. Effects of heat stress on the redox status in the
17. Weaver AC, Kim SW. Supplemental nucleotides high in ino­sine oviduct and early embryonic development in mice. J Reprod
5-monophosphate to improve the growth and health of nursery Dev 2005;51:281-7. https://doi.org/10.1262/jrd.16089
pigs. J Anim Sci 2014;92:645-51. https://doi.org/10.2527/jas. 26. Mujahid A, Akiba Y, Toyomizu M. Acute heat stress induces
2013-6564 oxidative stress and decreases adaptation in young white
18. Chaytor AC, See MT, Hansen JA, de Souza ALP, Middleton leghorn cockerels by downregulation of avian uncoupling
TF, Kim SW. Effects of chronic exposure of diets with reduced protein. Poult Sci 2007;86:364-71. https://doi.org/10.1093/
concentrations of aflatoxin and deoxynivalenol on growth ps/86.2.364
and immune status of pigs. J Anim Sci 2011;89:124-35. https:// 27. Mujahid A, Akiba Y, Warden CH, Toyomizu M. Sequential
doi.org/10.2527/jas.2010-3005 changes in superoxide production, anion carriers and substrate
19. Huynh TTT, Aarnink AJA, Verstegen MWA, et al. Effects of oxidation in skeletal muscle mitochondria of heat-stressed
increasing temperatures on physiological changes in pigs at chickens. FEBS Lett 2007;581:3461-7. https://doi.org/10.1016/
different relative humidities. J Anim Sci 2005;83:1385-96. j.febslet.2007.06.051
https://doi.org/10.2527/2005.8361385x 28. Karowicz-Bilinska A, Suzin J, Sieroszewski P. Evaluation of
20. Botto L, Lendelova J, Strmenova A, Reichstadterova T. The oxidative stress indices during treatment in pregnant women
effect of evaporative cooling on climatic parameters in a stable with intrauterine growth retardation. Med Sci Monit 2002;8:
for sows. Res Agric Eng 2014;60:S85-91. https://doi.org/10. CR211-6.
17221/40/2013-RAE 29. Toy H, Camuzcuoglu H, Arioz DT, Kurt S, Celik H, Aksoy
21. Seedorf J, Hartung J, Schroder M, et al. Temperature and N. Serum prolidase activity and oxidative stress markers in
moisture conditions in livestock buildings in Northern Europe. J pregnancies with intrauterine growth restricted infants. J
Agric Eng Res 1998;70:49-57. https://doi.org/10.1006/jaer. Obstet Gynaecol Res 2009;35:1047-53. https://doi.org/10.
1997.0284 1111/j.1447-0756.2009.01063.x
22. Britt JH, Szarek VE, Levis DG. Characterization of summer 30. McGlone JJ, Von Borell EH, Deen J, et al. Reviews: Compil­
infertility of sows in large confinement units. Theriogenology ation of the scientific literature comparing housing systems
1983;20:133-40. https://doi.org/10.1016/0093-691X(83) for gestating sows and gilts using measures of physiology,
90032-8 behavior, performance, and health. Prof Anim Sci 2004;20:
23. Robertshaw D. The environmental physiology of animal pro­ 105-17. https://doi.org/10.15232/S1080-7446(15)31285-7
duction. In: Clark JA, editor. Environmental aspects of housing

www.ajas.info  731