Dietary energy sources during late gestation and lactation of sows: effects on performance, glucolipid metabolism, oxidative status of sows, and their offspring

Abstract

In this study, the effects of maternal energy sources during late gestation and lactation on the performance, glucolipid metabolism, and oxidative status of sows and their offspring were investigated using a total of 75 (2 to 6 of parity) Landrace × Large White sows at day 85 of gestation under 3 different dietary treatments: SO diet (basal diet plus 3.0% and 5.0% soybean oil during late gestation and lactation, respectively), FO diet (basal diet plus 3.0%/5.0% fish oil during late gestation and lactation, respectively), and CS diet (basal diet plus 32%/42% corn starch during late gestation and lactation, respectively). All the 3 groups showed no obvious differences (P > 0.05) in the number of total piglets born, born alive, after cross-fostering, and at weaning, whereas the CS group exhibited a shorter farrowing duration (P < 0.05) and lower stillbirth rate (P < 0.05) when compared with the SO group. In addition, litter weight at birth was significantly higher in the CS group than in the SO or FO group (P < 0.05). Despite no notable differences in the ADG of suckling piglets among dietary treatments (P > 0.05), the CS group had greater feed intake than the SO group during the lactation period (P < 0.05). In neonatal piglets with normal birth weight (NBW, 1.3 to 1.5 kg), the CS group was lower than the SO group in the content of liver glycogen (P < 0.05) and the mRNA abundances of fatty acid synthase, acetyl-CoA carboxylase, fatty acid-binding protein 1, and acyl-CoA oxidase (P < 0.05). Interestingly, compared with the SO group, the FO group had a lower preweaning mortality rate (P < 0.05), but greater liver glycogen pools (P < 0.05) in neonatal piglets with low birth weight (LBW, <1.1 kg). Compared with the CS group, the FO group showed an increase in the plasma malondialdehyde levels (P < 0.05) of sows, as well as an increase of 8-hydroxy-deoxyguanosine (P < 0.05) and a decrease of ferric reducing ability of plasma (P < 0.05) in NBW piglets. Overall, the diet rich in starch decreased the stillbirth rate and increased the litter weight of neonatal piglets, the dietary supplementation with fish oil decreased preweaning mortality rate, and the diet with a low n6:n3 ratio increased the oxidative status of sows and their offspring.

Introduction

Reproductive performance target is defined as the number of piglets weaned per sow per year, and it is affected by litter size and preweaning mortality (Vanderhaeghe et al., 2010). In the past decades, genetic selection and management changes have significantly increased the litter size of sows. However, the incidence of stillbirth rate and postnatal mortality are still high, with stillbirth and preweaning mortality accounting for 3% to 8% and 10% to 13% of all pigs born, respectively (Vanderhaeghe et al., 2013; Van den Bosch et al., 2019). Therefore, it is critical to develop feed strategies to reduce stillbirth rate and preweaning mortality rate for the improvement of sow reproductive performance.

The increase in the mortality of piglets can be attributed to inadequate energy supply, with the mortality rate being greatest for low birth weight (LBW) piglets during the first few days of life. Generally, newborn piglets rely on glycogen from liver and muscles to maintain glucose homeostasis until the intakes of colostrum and milk are sufficient to meet their energy demand. The health status of LBW piglets directly affects the survival rate of sucking piglets (Theil et al., 2011). Pregnant sows are commonly fed a diet rich in starch or fat for the rapid growth and glycogen deposition of fetuses (Theil et al., 2011). Utilization of starch depends on its conversion to glucose via hydrolysis of glycosidic bonds as well as glucose degradation by anaerobic glycolysis and aerobic oxidation, whereas fat is based on hydrolysis of lipase to supply energy mainly by fatty acid (FA) oxidation. Glycogen synthesis is closely related to gluconeogenesis. However, the effects of energy sources (starch or fat) in maternal diets on the mortality and the characteristics related to glycogen storage and fat metabolism of new born piglets still remain highly controversial (Van der Peet-Schwering et al., 2004; Quiniou et al., 2008; Almond et al., 2015). Hence, the first purpose of the present study was to evaluate the effect of different energy source (fat and starch) inclusion in diet during late gestation and lactation on the glucolipid metabolism in newborn piglets and the viability of suckling piglets.

Furthermore, fish oil is regarded as highly enriched in n-3 PUFA with a low n6:n3 ratio, whereas soybean oil is mainly composed of n-6 PUFA with a high n6:n3 ratio. The supplementation of fish oil and soybean oil has been reported to affect the performance of sows (Cools et al., 2011; Jin et al., 2017). However, PUFA can provide more substrates for lipoperoxidation and further lead to oxidative stress because of their double bonds in structure (Shen et al., 2015; Su et al., 2017). To date, the appropriate n6:n3 ratio for minimizing oxidative stress damages of dam and progeny has not been reported. Therefore, the second aim of this study was to investigate the effect of maternal diets with a different n6:n3 ratio on the oxidative status of dam and progeny.

Materials and Methods

The experimental designs and procedures presented in this study were carried out in accordance with the Animal Care and Use Committee of the South China Agricultural University.

Animals, Diets, and Housing

A total of 75 Landrace × Large White sows with 2 to 6 parity were allotted to the experiment. At day 84 of gestation, sows were allocated across 3 dietary treatment groups based on parity number and BW. The experiment was performed from day 85 of gestation to the end of weaning (on day 21 of lactation). Table 1 shows the 3 different diets formulated according to the nutrient requirements recommended by the NRC (2012): SO diet (basal diet plus 3.0% and 5.0% soybean oil during late gestation and lactation, respectively), FO diet (basal diet plus 3.0%/5.0% fish oil during late gestation and lactation, respectively), and CS diet (basal diet plus 32%/42% corn starch during late gestation and lactation, respectively). The 3 experimental diets were calculated to be isonitrogenous and isoenergetic. Fish oil was obtained from a commercial source (Rongcheng Ayers Ocean Bio-Technology Co., Ltd, Rongcheng, China); soybean oil and corn starch were acquired from a commercial source (Guangzhou Jingtang Bio-Technology Co., Ltd, Guangzhou, China).

Table 1.

Ingredient and nutrient composition of experimental gestation diets and lactation diets (as-fed basis)

	Gestation diet1			Lactation diet2
Item	SO	FO	CS	SO	FO	CS
Ingredient, %
Corn	51.55	51.55	19.35	52.90	52.90	9.10
Soybean meal (43% CP)	10.80	10.80	20.00	23.50	23.50	28.50
Corn starch	—	—	32.00	—	—	42.00
Wheat bran	19.70	19.70	19.70	9.75	9.75	10.55
Corn protein powder	1.50	1.50	3.50	1.50	1.50	4.50
Extruded soybean	8.00	8.00	—	2.00	2.00	—
Soybean oil	3.00	—	—	5.00	—	—
Fish oil	—	3.00	—	—	5.00	—
Dicalcium phosphate	2.00	2.00	2.00	2.00	2.00	2.00
Limestone	1.35	1.35	1.35	1.35	1.35	1.35
Salt	0.30	0.30	0.30	0.30	0.30	0.30
Sodium sulfate	0.30	0.30	0.30	0.30	0.30	0.30
Choline chloride	0.20	0.20	0.20	0.10	0.10	0.10
Lysine sulfate (70%)	0.20	0.20	0.20	0.20	0.20	0.20
Premix3	1.00	1.00	1.00	1.00	1.00	1.00
Mildewcide4	0.10	0.10	0.10	0.10	0.10	0.10
Calculated compositon5
DE, Mcal/kg	3.21	3.20	3.19	3.38	3.37	3.36
NE, Mcal/kg	2.36	2.36	2.33	2.46	2.46	2.42
CP, %	15.59	15.59	15.59	17.58	17.58	17.57
EE6, %	7.28	7.28	2.07	8.06	8.06	1.58
CF7, %	3.55	3.55	3.27	3.19	3.19	2.80
Starch, %	37.58	37.58	49.25	36.46	36.46	51.05
NDF, %	14.32	14.32	12.18	11.52	11.52	8.89
Ca, %	1.07	1.07	1.07	1.09	1.09	1.09
NPP8, %	0.50	0.50	0.46	0.49	0.49	0.45
Lys, %	0.87	0.87	0.90	1.06	1.06	1.09
Met, %	0.28	0.28	0.26	0.31	0.31	0.29
Thr, %	0.59	0.59	0.58	0.70	0.70	0.69
Trp, %	0.20	0.20	0.20	0.23	0.23	0.23
Analyzed composition
CP, %	15.77	15.73	15.64	18.60	18.49	18.63
CF, %	3.42	3.24	3.21	3.59	3.65	3.22
EE, %	6.98	7.22	2.04	7.60	6.95	1.96

¹SO = 3.0% soybean oil diet group; FO = 3.0% fish oil diet group; CS = 32% corn starch diet group.

²SO = 5.0% soybean oil diet group; FO = 5.0% fish oil diet group; CS = 42% corn starch diet group.

³Provided per kg of diet: Cu, 10.0 mg; Fe, 130 mg; Mn, 45 mg; Zn, 60 mg; I, 0.30 mg; Se, 0.27 mg; Co, 0.1 mg; vitamin A, 6,760 IU; vitamin D₃, 4,992 IU; vitamin E, 209.8 mg; vitamin K₃, 3.7 mg; thiamin, 3.7 mg; riboflavin, 12 mg; pyridoxine, 7.4 mg; niacin, 50 mg; folic acid, 3.7 mg; vitamin C, 200 mg.

⁴Mildewcide contains 60 % potassium propionate.

⁵Calculated chemical concentrations using values for feed ingredients from NRC (2012).

⁶EE = ether extract.

⁷CF = crude fiber.

⁸NPP = non-phytin phosphorus.

Feed was offered twice a day at 0700 and 1430 h during the experiment. Sows were fed 3.0 kg/d from day 85 of gestation to farrowing, and half of the daily feed was given at each meal. During lactation, all sows were allowed to consume each of the 3 diets ad libitum (Table 1). Sows and piglets were given free access to drinking water. During the experimental period, we eliminated the sows with serious lameness, death, and reproductive failure such as premature birth or nonpregnant in late gestation. Furthermore, sows with the number of piglets per litter <9 or >12 after cross-fostering were also excluded (Table 2). Feed samples were analyzed in terms of CP (ISO 5983-2), crude fiber (ISO 6865-2000), and ether extract (ISO 6492-1999). The FA composition of feed was analyzed by gas chromatography (Tanghe et al., 2015), using 100 mg of feed (Table 3).

Table 2.

The number of sows during the experimental period

	Diet
Item	SO1	FO1	CS1
At day 85 of gestation	25	25	25
Reproductive failure2	4	3	5
Excluded during gestation3	1	2	2
At day 109 of gestation	20	20	18
Farrowing	20	20	18
Excluded during lactation4	4	4	—
Weaning	16	16	18

¹SO = 3.0%/5.0% soybean oil diet group; FO = 3.0%/5.0% fish oil diet group; CS = 32%/42% corn starch diet group.

²Sow returned to estrus, miscarried, or nonpregnant.

³Sows were excluded because of illness, death, or serious lameness.

⁴Sows with the number of piglets per litter < 9 or > 12.

Table 3.

Fatty acid composition of experimental gestation diets and lactation diets (as-fed basis)

	Gestation diet1			Lactation diet2
Item	SO	FO	CS	SO	FO	CS
Fatty acid composition3, mg/g total fatty acids
SFA
C16:0	121.17	108.99	154.52	118.67	101.91	158.12
C18:0	33.24	28.06	25.61	35.57	26.25	33.61
Total	157.87	144.73	180.13	159.39	144.38	198.35
MUFA
C18:1n-9	238.74	185.40	231.19	247.03	162.75	271.76
Total	239.86	203.70	231.19	248.70	177.52	273.97
n-6 PUFA
C18:2n-6	538.60	363.40	538.44	515.80	274.85	483.56
C18:3n-6	1.71	—	—	2.91	—	—
C20:3n-6	3.56	1.92	—	3.79	1.46	3.29
C20:4n-6	—	8.25	—	—	12.26	—
Total	543.87	373.57	538.44	522.49	288.57	486.85
n-3 PUFA
C18:3n-3	51.66	44.72	37.76	47.84	39.05	40.43
C20:3n-3	—	0.77	—	—	1.10	—
C20:5n-3	—	77.29	—	6.05	114.28	—
C22:6n-3	—	115.53	—	9.23	172.20	—
Total	51.66	238.32	37.76	63.11	326.61	40.43
Ratio of n-6:n-3 PUFA	10.53	1.57	14.26	8.28	0.88	12.04

¹SO = 3.0% soybean oil diet group; FO = 3.0% fish oil diet group; CS = 32% corn starch group.

²SO = 5.0% soybean oil diet group; FO = 5.0% fish oil diet group; CS = 42% corn starch diet group.

³Data are given as means of 2 replicate batches.

Performance Measurement

Body weight and backfat thickness of sows were measured on days 85 and 109 of gestation and at weaning. Backfat thickness was measured at the P2 position (left side of the 10th rib and 6 cm lateral to the spine) with a mode ultrasound scanner (PIGLOG105, SFK Technology A/S Helver, Denmark). The numbers of total piglets born, born alive, and stillborn, as well as farrowing duration, were recorded at farrowing. Cross-fostering was kept within diet treatments to adjust litter size within 48 h post-parturition. The number of piglets was recorded after cross-fostering and at weaning. Average daily gain of piglets was measured by weighing within 24 h of birth before colostrum consumption (day 0), after cross-fostering, on day 10, and at weaning (day 21). Feed refusals were measured every morning during lactation, and the ADFI was calculated. Furthermore, both the stillbirth rate during parturition and the preweaning mortality rate during weaning were recorded.

Sample Collection

Briefly, a total of 24 sows (8 sows per dietary treatment with a similar parity and BW) were selected for blood sampling after an overnight fast (12 h) during gestation. Blood samples were collected in 10-mL ordinary centrifuge tubes and centrifuge tubes containing sodium heparin from the ear vein of the sows with a minimum amount of stress. After farrowing, blood samples were collected from the jugular vein and 12 piglets per treatment were slaughtered before suckling. The 12 piglets were chosen from 6 different litters, including one with a normal birth weight (NBW, 1.3 to 1.5 kg) and one with a LBW (<1.1 kg) from each litter (Wu et al., 2010). Blood samples were centrifuged at 4 °C, 3,000 × g for 15 min to obtain the serum and plasma, followed by storage at −20 °C for further analysis. Meanwhile, the duodenums, jejunums, ileums, stomachs, hearts, spleens, lungs, kidneys, livers, and pancreases of the piglets were removed and weighed, and the lengths of small intestine were also measured. Finally, the livers and longissimus dorsi (LD) muscles were snap-frozen in liquid nitrogen before storage at −80 °C for further analysis.

Analysis of Glucolipid Metabolism

The blood samples of sows and piglets were measured in terms of serum triglyceride, FFA, and glucose using specific commercially available enzymatic assays (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). A reference sample was included in all assay kits and the intra- and interassay coefficient of variation was always below 10% and 15%, respectively. The glycogen concentrations of the liver tissues and LD muscles were measured with the anthracenone method using a commercial kit (Nanjing Jiancheng Bioengineering Institute). The glycogen pool in liver was calculated by the equation: glycogen pool in liver (g) = glycogen content of liver (g/g) × liver weight (g) as reported by Theil et al. (2011).

Assessment of Oxidative Status

Systemic oxidative status was determined by measuring the oxidant and antioxidant parameters in the plasma of sows and piglets, including malondialdehyde (MDA), 8-hydroxy-deoxyguanosine (8-OHdG), and ferric reducing ability of plasma (FRAP), using specific assay kits (Nanjing Jiancheng Bioengineering Institute). Malondialdehyde, as a product of lipid peroxidation, was measured by reaction with thiobarbituric acid at 95 °C and expressed as nmol MDA per milliliter of plasma according to the method of Grotto et al. (2007). The concentration of 8-OHdG, the most common biomarker for oxidative damage to DNA, was measured by using an anti-8-OHdG monoclonal antibody in an ELISA kit and expressed as ng 8-OHdG per milliliter of plasma. To further assess the total antioxidative capacity of the plasma in sows and piglets, the FRAP values were colorimetrically measured by the reduction of the ferric ion (Fe³⁺) to ferrous ion (Fe²⁺) via the reaction of ferrous–tripyridyltriazine complex in relation to antioxidant-based ascorbic acid standards. The results were expressed in mmol Fe²⁺ formed per liter of plasma.

RNA Isolation and Quantitative Real-Time PCR

Total RNA was isolated from frozen liver samples using Trizol Reagent (Thermo Fisher, Wilmington, DE) according to the manufacturer's instructions. All samples were measured at optical density of 260 and 280 nm, using the NanoDrop ND-1000 spectrophotometer (Thermo Fisher). All the RNA samples had a 260/280 nm ratio between 1.8 and 2.0. The integrity of total RNA was determined using 1.0% agarose gel electrophoresis by visualizing the clear 5S, 18S, and 28S rRNA bands. According to the manufacturer's instructions, cDNA was synthesized with 1 µg RNA in a 20-µL reaction volume using PrimeScript first-strand cDNA synthesis kit (Takara, Osaka, Japan), followed by storage at −80 °C for further analysis. The primer sequences used for PCR are shown in Supplemental Table 1. The total reaction volume (15 µL) comprised 2 µL of cDNA template solution, 7.5 µL of SYBR Green PCR master mix (Thermo Fisher Scientific, Waltham, MA), 4.9 µL of water, and 0.6 µL of each primer. Quantitative real-time PCR was tested with the real-time fluorescent quantitative PCR (Thermo Fisher Scientific) on a QuantStudio 6&7 Real-Time PCR System and all measurements were done in duplicate. The RT-PCR program started with denaturation at a 95 °C for 10 min, followed by 40 cycles at 95 °C for 15 s, annealing for 30 s at 60 °C, and an extension for 30 s at 72 °C. Melting curves of all samples were analyzed to test the specificity of amplification by fluorescence measurement. The mRNA abundances of the selected genes were normalized using 18S ribosomal RNA (18S), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and beta-actin (β-actin) as reference genes. The relative mRNA abundances were expressed as the value of ∆Ct (Hu et al., 2017), which was estimated by the equation: ∆Ct = 2^{−∆∆Ct (sample − control)}, where ∆∆Ct_{(sample − control)} = (Ct_{target gene} − Ct_{reference genes}) treated − (Ct_{target gene} − Ct_{reference genes}) control.

Statistical Analyses

The data related to the reproduction performance of sows and piglets, glucolipid metabolism, and oxidative status were statistically analyzed using an ANOVA in Mixed Model procedure of SAS 8.2 (SAS Institute., Cary, NC). Multiple comparisons were performed between the 3 groups with Duncan multiple range test. The piglet stillbirth rate and preweaning mortality rate were analyzed using the chi-square test. Differences between mean values were considered statistically significant at P < 0.05, with a trend toward significance at 0.05 ≤ P ≤ 0.10.

Results

Sow Performance

Table 4 shows the effects of the 3 dietary treatments on sow performance. It can be seen that the 3 dietary treatments had no differences (P > 0.05) in their effects on sow BW or backfat thickness, but the FO diet group showed more lactation weight loss of 9.2 kg (P < 0.05) when compared with the CS diet group. In addition, the CS and FO groups had greater (P < 0.05) ADFI than the SO group during the second week and days 1 to 21 of lactation.

Table 4.

Effects of dietary supplementation with soybean oil, fish oil and corn starch on body weight, backfat thickness, and feed intake during lactation of sows

	Diet
Item	SO1	FO1	CS1	SEM	P-value
No. of sows	20	20	18
BW of sows, kg
Day 85 of gestation	262.0	273.8	270.8	4.03	0.46
Day 109 of gestation	279.7	292.1	286.6	4.11	0.46
Gain during gestation	17.6	18.8	15.8	0.89	0.52
Parturition	253.8	265.6	258.9	4.05	0.49
Weaning	230.6	238.7	241.2	3.86	0.51
Loss during lactation	23.2ab	26.9a	17.7b	1.31	0.01
Sow backfat thickness, mm
Day 85 of gestation	21.0	21.6	19.4	0.50	0.16
Day 109 of gestation	22.1	21.9	20.7	0.44	0.44
Weaning	17.6	17.9	16.8	0.48	0.69
Average daily feed intake2, kg
First week of lactation	3.0	3.1	3.6	0.13	0.08
Second week of lactation	5.3b	5.9a	6.3a	0.13	<0.01
Third week of lactation	5.3b	6.1ab	6.5a	0.18	0.02
Mean of first week to third week	4.5b	5.0a	5.5a	0.11	<0.01

¹SO = 3.0%/5.0% soybean oil diet group; FO = 3.0%/5.0% fish oil diet group; CS = 32%/42% corn starch diet group.

²The number of sows in SO, FO, and CS is 16, 16, and 18, respectively.

^a,bMean values with different superscript letters in the same row differ significantly (P < 0.05).

Piglet Performance

Table 5 shows the piglet performance in the 3 dietary treatments. There were no differences (P > 0.05) in the numbers of total piglets born, born alive, after cross-fostering, and at weaning in the 3 groups, while the number of stillbirths showed a reduction tendency (P = 0.08) in the CS group vs. the SO or FO groups. Furthermore, the CS group showed an increase of 0.12 and 0.15 kg in the BW at birth vs. the SO and FO group, respectively (P = 0.10). Litter weight at birth was greater in the CS group than in the SO or FO group (P < 0.05). The daily gain of piglets was not affected by the 3 dietary treatments (P > 0.05). The CS group showed the lowest farrowing duration among the 3 dietary treatments (P < 0.05). The stillbirth rate exhibited a decrease (P < 0.05) in the CS group vs. the SO or FO group, whereas preweaning mortality rate was lower in the FO group than in the SO group (P < 0.05).

Table 5.

Effects of dietary supplementation with soybean oil, fish oil and corn starch on litter performance of sows

Item	Diet
	SO1	FO1	CS1	SEM	P-value
No. of sows	16	16	18
No. of pigs per litter
Total piglets born2	14.2	14.1	13.9	0.46	0.98
Piglets born alive2	12.0	12.6	13.5	0.41	0.31
Stillbirth2	2.0	1.3	0.4	0.27	0.08
After cross-foster	10.1	9.9	10.4	0.15	0.38
Pigs weaned	9.4	9.8	9.9	0.15	0.34
Piglet mean BW, kg
At birth2	1.3	1.3	1.4	0.23	0.10
After cross-foster	1.6	1.5	1.6	0.05	0.11
On day 10	3.6	3.5	3.6	0.09	0.21
On day 21	6.2	6.2	6.4	0.14	0.20
Litter weight, kg
At birth2	15.9a	16.9a	19.7b	0.55	0.02
After cross-foster	15.9	15.5	16.4	0.61	0.82
On day 10	34.3	34.3	35.4	1.16	0.91
On day 21	58.1	60.8	63.3	1.88	0.52
ADG, g/d
Days 1 to 10	202.2	193.1	197.6	6.60	0.86
Days 11 to 21	230.2	244.7	255.0	8.38	0.48
Days 1 to 21	216.9	220.2	227.7	5.45	0.71
Farrowing duration2, min	341.4a	267.1b	168.1c	17.45	<0.01
Stillbirth rate3, %	13.8a	8.9a	3.2b		<0.01
Preweaning mortality rate3, %	7.4a	1.3b	4.8ab		0.03

¹SO = 3.0%/5.0% soybean oil diet group; FO = 3.0%/5.0% fish oil diet group; CS = 32%/42% corn starch diet group.

²The number of sows in SO, FO, and CS is 20, 20, and 18, respectively.

³Stillbirth rate and preweaning mortality rate were analyzed using the chi-square test.

^a–cMean values with different superscript letters in the same row differ significantly (P < 0.05).

Relative Weight of Internal Organs and Small Intestine

Table 6 shows the relative weight of internal organs and small intestine of piglets. LBW piglets showed no difference in the weight of duodenum, jejunum, ileum, stomach, heart, spleen, lung, and kidney among the 3 dietary treatments (P > 0.05). However, the liver weight of LBW piglets was greater in the FO group than in the SO group (P < 0.05), and the pancreas weight of LBW piglets showed an increase (P < 0.05) in the CS group vs. the SO group. For NBW piglets, the organ weight exhibited no differences in the 3 dietary treatments (P > 0.05), but the stomach weight was greater in the SO group than in the other 2 groups (P < 0.05).

Table 6.

Effects of dietary supplementation with soybean oil, fish oil and corn starch on weight of internal organs and intestinal weight per length of newborn piglets

	Diet
Item	SO1	FO1	CS1	SEM	P-value
No. of piglets	6	6	6
Piglets with LBW2
Duodenum, g/cm3	0.09	0.08	0.09	0.03	0.91
Jejunum, g/cm	0.06	0.08	0.06	0.01	0.59
Ileum, g/cm	0.06	0.07	0.06	0.00	0.12
Stomach, g/kg BW4	4.99	4.78	5.48	0.26	0.54
Heart, g/kg BW	6.92	6.70	6.90	0.16	0.85
Liver, g/kg BW	21.58b	27.11a	24.63ab	0.94	0.04
Spleen, g/kg BW	0.84	1.05	1.05	0.05	0.12
Lung, g/kg BW	19.09	16.99	15.12	1.20	0.42
Kidney, g/kg BW	7.64	7.47	8.74	0.36	0.31
Pancreas, g/kg BW	0.68b	1.00ab	1.13a	0.05	0.03
Piglets with NBW2
Duodenum, g/cm3	0.11	0.09	0.12	0.01	0.58
Jejunum, g/cm	0.09	0.09	0.09	0.00	0.95
Ileum, g/cm	0.09	0.08	0.10	0.00	0.23
Stomach, g/kg BW4	5.57a	4.13b	4.37b	0.23	0.01
Heart, g/kg BW	6.34	6.74	7.09	0.20	0.33
Liver, g/kg BW	27.49	29.40	28.34	1.22	0.83
Spleen, g/kg BW	0.89	0.95	0.92	0.04	0.84
Lung, g/kg BW	15.69	17.51	16.78	0.75	0.64
Kidney, g/kg BW	6.62	8.14	7.42	0.35	0.21
Pancreas, g/kg BW	0.98	1.09	1.06	0.03	0.45

¹SO = 3.0%/5.0% soybean oil diet group; FO = 3.0%/5.0% fish oil diet group; CS = 32%/42% corn starch diet group.

²LBW = low birth weight; NBW = normal birth weight.

³Intestinal weight per length = intestine weight, g/intestine length, cm.

⁴Weight of internal organ = organ weight, g/piglet birth weight, kg.

^a,bMean values with different superscript letters in the same row differ significantly (P < 0.05).

Parameters Related to Glucolipid Metabolism

The results of serum metabolites and characteristics related to glycogen storage are displayed in Table 7. There was no difference in the concentrations of total triglyceride and FFA in neonatal piglets (P > 0.05) among the 3 dietary treatments. The liver glycogen concentrations showed a decrease (P < 0.05) in the NBW piglets of the CS group vs. the SO or FO group, whereas the liver glycogen pools were affected by the dietary supplementation in LBW piglets of the FO group relative to the SO group (P < 0.05). The indices of LD muscle glycogen storage were not affected by fish oil or corn starch supplementation (P > 0.05).

Table 7.

Effects of dietary supplementation with soybean oil, fish oil, and corn starch on serum metabolites and characteristics related to glycogen storage

	Diet
Item	SO1	FO1	CS1	SEM	P-value
No. of piglets	6	6	6
Piglets with LBW2
Liver glycogen, mg/g	64.63	72.22	62.08	2.69	0.28
Longissimus dorsi glycogen, mg/g	56.60	44.02	50.55	3.00	0.24
Liver glycogen pools3, g	1.17a	1.95b	1.70ab	0.13	0.04
Serum triglyceride, mM	0.32	0.27	0.30	0.02	0.57
Serum FFA, mM	0.13	0.06	0.08	0.02	0.32
Serum glucose, mM	3.47	3.56	3.51	0.20	0.98
Piglets with NBW2
Liver glycogen, mg/g	98.01b	90.58b	62.18a	6.32	0.04
Longissimus dorsi glycogen, mg/g	52.49	34.47	44.56	2.88	0.48
Liver glycogen pools3, g	3.78	3.81	2.35	0.31	0.10
Serum triglyceride, mM	0.34	0.35	0.30	0.02	0.63
Serum FFA, mM	0.07	0.11	0.15	0.03	0.66
Serum glucose, mM	3.58	4.35	3.02	0.25	0.09
No. of sows	8	8	8
Day 109 of gestation
Serum glucose, mM	4.58ab	5.01b	4.27a	0.18	0.03
Serum triglyceride, mM	0.49	0.42	0.49	0.02	0.47
Serum FFA, mM	0.16a	0.16a	0.50b	0.05	<0.01

¹SO = 3.0% soybean oil diet group; FO = 3.0% fish oil diet group; CS = 32% corn starch diet group.

²LBW = low birth weight; NBW = normal birth weight.

³The glycogen pools in liver was calculated as glycogen pools in liver, g = glycogen content of liver, g/g × liver weight, g.

^a,bMean values with different superscript letters in the same row differ significantly (P < 0.05).

In addition, at day 109 of gestation, the serum glucose concentration was lower (P < 0.05) for the sows in the CS group than in the FO group, and the sows in the CS group had higher (P < 0.05) serum FFA concentration than those in the SO or FO group.

Expression of mRNA Related to Glucolipid Metabolism Genes

Figure 1 shows the mRNA abundances related to de novo lipogenesis, FA oxidation, glycogen synthase, and gluconeogenic genes in the liver of neonatal piglets under different dietary treatments. In the liver of LBW piglets, no differences (P > 0.05) were observed in the mRNA abundances of FA synthase (FAS), acetyl-CoA carboxylase (ACC), FA-binding protein 1 (FABP1), acyl-CoA oxidase (ACOX1), and carnitine palmitoyltransferase I (CPT-1) among the 3 dietary groups. Nonetheless, the mRNA abundances of ACC and ACOX1 in the liver of NBW piglets were downregulated (P < 0.05) in the CS group, with the mRNA abundances of CPT-1 showing a downward trend (P = 0.06) in the CS group vs. the SO group. In addition, the liver of NBW piglets also showed downregulation (P < 0.05) in the mRNA abundances of FAS and FABP1 in the CS and FO groups relative to the SO group. The mRNA abundances of glycogen synthase 2 (GYS2) and glucose 6-phosphate (G-6-P) were not affected in the CS and FO groups when compared with the SO group (P > 0.05), whereas the mRNA abundances of fructose-1,6-bisphosphatase 1 (FBP1) in the liver of NBW piglets showed a downward trend (P = 0.07) in the CS group compared with the SO or FO group.

Figure 1.

Effects of dietary supplementation with soybean oil, fish oil, and corn starch on relative mRNA abundances of de novo lipogenesis and fatty acid oxidation genes in the liver of LBW (A) and NBW (B) as well as glycogen synthesis and gluconeogenic genes in the liver of LBW (C) and NBW (D). The mRNA abundances of fatty acid synthase (FAS), acetyl-CoA carboxylase (ACC), fatty acid-binding protein 1 (FABP1), acyl-CoA oxidase (ACOX1), carnitine palmitoyltransferase I (CPT-1), glycogen synthase 2 (GYS2), fructose-1,6-bisphosphatase 1 (FBP1), and glucose 6-phosphate (G-6-P) were normalized using 18S ribosomal RNA (18S), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and beta-actin (β-actin) as reference genes. SO = 3.0% soybean oil diet group; FO = 3.0% fish oil diet group; CS = 32% corn starch diet group. LBW = low birth weight of neonatal piglets. NBW = normal birth weight of neonatal piglets. ^a,bMean values with different small letters were considered significantly different (P < 0.05). Data are expressed as means ± SEM; n = 6.

Oxidative and Antioxidative Indicators

Oxidative and antioxidative indicators measured in the plasma of sows and neonatal piglets are shown in Fig. 2. No differences (P > 0.05) were observed in the MDA concentration, 8-OHdG concentration, and FRAP value of plasma from the LBW piglets among the 3 dietary treatments. Nevertheless, when compared with the CS group, the NBW piglets in the FO group tended to have a higher plasma 8-OHdG concentrations (P = 0.09). Furthermore, a decreasing trend (P = 0.06) was observed in the FRAP values for the NBW piglets in the FO group vs. the CS group. The MDA concentrations of sow plasma at day 109 of gestation showed an increasing tendency (P = 0.07) in the FO group vs. the CS group.

Figure 2.

Diet effects on plasma levels of MDA (A), 8-OHdG (B), and FRAP (C) of G109-d sows and neonatal piglets. MDA = malondialdehyde; 8-OHdG = 8-hydroxy-deoxyguanosine; FRAP = ferric reducing ability of plasma. SO = 3.0% soybean oil diet group; FO = 3.0% fish oil diet group; CS = 32% corn starch diet group. LBW = low birth weight of neonatal piglets. NBW = normal birth weight of neonatal piglets. ^a,bMean values with different small letters were considered significantly different (P < 0.05). Data are shown as means ± SEM; n = 8 for G109-d sows and n = 6 for neonatal piglets.

Discussion

In this study, the CS group showed an increase in litter weight at birth and a decrease in the stillbirth rate when compared with the SO or FO group. Maternal dietary nutrients are primarily directed toward support of fetal tissue growth in response to energy-balance need (Wang et al., 2018, 2019). As main energy sources, starch relies on glycolysis and aerobic oxidation to generate energy, whereas fat produces energy by conversion of gluconeogenic precursors or FA beta oxidation, which is less efficient than the production of dietary glucose (Jones et al., 2002; Kim et al., 2017). Interestingly, the present results showed NBW piglets from CS group had a lower serum glucose concentration, which was in agreement with the studies by of Coffey et al. (1982) and Kim et al. (2017), implying that the neonatal piglets from CS-fed sows probably mainly used starch as an energy substrate for glycolysis of glucose. Besides, the CS diet had reduced the relative mRNA abundances of FAS, ACC, FABP1, ACOX1, and CPT-1 in the liver of NBW piglets, suggesting the inhibition of fat mobilization by feeding high starch as the maternal energy source, which was in accordance with the argument above. A possible explanation is that CS group ameliorated litter performance in weight and stillbirth rate not through the inefficient mobilization of fat, but through the consumption of starch to generate energy. However, the mechanisms of the effects of the high-starch diet on the sows glycometabolism in late gestation and lactation await further clarification.

These results of performance are consistent with previous studies (Bikker et al., 2007), but not in line with the studies by Seerley et al. (1981), Coffey et al. (1982), and Quiniou et al. (2008), who reported that high dietary starch levels had no or negative effects on weight and mortality of newborn piglets. Altering maternal nutrition in the third trimester of gestation affects fetal development, probably leading to long-lasting effects on the offspring (Almond et al., 2008). It has been reported that porcine fetal growth accelerates during the second half of pregnancy (McPherson et al., 2004). Therefore, the discrepancy can be attributed to differences in experimental timing and duration of starch supplementation. The experiment of Seerley et al. (1981) was performed from prefarrowing to day 21 of lactation, and that of Coffey et al. (1982) was initiated from day 100 of gestation to farrowing, whereas our experiment was performed from day 85 of gestation to day 21 of lactation. Another explanation for the difference of the results is possibly the dosage of diet starch, such as 11.3% in the study by Quiniou et al. (2008) in contrast to 32% in the present study.

In addition, the CS group also showed a decrease in farrowing duration, with the numerical regularity of farrowing duration being consistent with that of stillbirth rate in the 3 treatment groups. Stillbirth risk is higher in sows with prolonged farrowing because of hypoxia of fetus (Vanderhaeghe et al., 2013). Parturition is an energy-demanding process (Tokach et al., 2019). Arterial glucose concentration at 1 h after birth of the first piglet was negatively correlated with farrowing duration (Feyera et al., 2018), suggesting glucose is the key energy metabolite for oxidative metabolism of gravid uterus. In the present study, because of differences of dosage (fat: 3.0%/5.0% vs. starch: 32.0%/42.0%) and glucose production efficiency (fat produces glucose by conversion of gluconeogenic precursors, which is less efficient than the production of dietary starch by hydrolysis of glycosidic bonds; Jones et al., 2002), feeding a diet containing starch rather than fat a day prior to parturition could probably supply more readily absorbed glucose required by the uterus during parturition, which could have positive impact on uterine contractions and farrowing duration.

During pregnancy and lactation, the sow undergoes many physiological and metabolic changes, such as progressive and reversible insulin resistance, which corresponds to a decrease in the effectiveness of insulin in regulating blood glucose (Tan et al., 2016, 2018). Here, we found that the CS supplementation significantly decreased the serum glucose in the sows at day 109 of gestation and increased the ADFI in lactation. This implies that CS supplementation for sows could improve insulin sensitivity, which is consistent with the findings of Hoffman et al. (2003), Van der Peet-Schwering et al. (2004), and George et al. (2009). Nevertheless, this hypothesis needs to be further confirmed in future studies. The available evidence indicates that there is an obvious negative correlation between feed intake during lactation and insulin resistance of sows (Mosnier et al., 2010). Moreover, Louis-Sylvestre (1999) indicated that the role of glucose is dynamic in control of feed intake, which is a satiety factor and an initiation signal. Glucose receptors are present in the brain's center, and a decrease in the plasma glucose concentration of CS-fed sows possibly means the decline of neuronal signaling for glucose metabolism, thereby promoting the feed intake of sows.

Our results also showed that the preweaning mortality rate of piglets was decreased in response to maternal FO supplementation, which is in agreement with the results of Rooke et al. (2001) and Jin et al. (2017) who found dietary supplementation of salmon oil or fish oil could reduce preweaning mortality. Piglets are born deficient of energy; thus, neonatal piglets rely on glycogen, colostrum, and transient milk to overcome challenges to their survival. During the first 2 d post-birth, liver glycogen pools are fast oxidized, enabling LBW piglets to cover their energy requirements, leading to decrease in the number of dead LBW piglets in the postnatal stage (Theil, 2017). In the study, liver glycogen pools in LBW piglets were positively affected by the fish oil supplementation when compared with the SO group. The health status of LBW piglets directly affects the survival rate of weaned piglets. Therefore, this finding indicates that inclusion of fish oil in the sow diet possibly reduces preweaning mortality by increasing the liver glycogen pools of LBW piglets to boost competitiveness of weak piglets. Another possible explanation for this result was that fish oil can enhance concentrations of colostrum IgM or IgG after birth (Mitre et al., 2005; Jin et al., 2017), which is crucial for newborn piglets to acquire passive immunity from sows.

In the present study, fish oil was shown to increase oxidative stress, which can be supported by previous studies showing that ingestion of fish oil in maternal diet increased MDA concentrations, but decreased α-tocopherol levels and glutathione peroxidase in sow serum (Cools et al., 2011; Tanghe et al., 2015). A possible explanation for this phenomenon is that the unsaturated bonds of PUFA are prone to oxidation. According to the FA composition in the sow diet, the FO diet contains more readily oxidizable components than the CS diet. However, the SO diet is also rich in FA, but failed to show the same results of the FO diet. Total n-6 was calculated as the sum of C18:2n-6 + C18:3n-6 + C20:3n-6 + C20:4n-6, and total n-3 was calculated as the sum of C18:3n-3 + C20:3n-3 + C20:5n-3 + C22:6n-3. Based on the FA composition in the maternal diet, the ratio of n6:n3 in the SO, FO, and CS diets was 10.53, 1.57, and 14.26 (8.28, 0.88, and 12.04 for lactation), respectively. Theoretically, a lower n6:n3 ratio in the maternal diet means higher oxidative stress. A possible explanation for the results is that the SO diet is rich in n-6 PUFA with linoleic acid containing 2 double bonds in its structure, whereas the FO diet is rich in n-3 PUFA with eicosapentaenoic acid and docosahexaenoic acid containing 5 and 6 double bonds in their structure, respectively, thus providing more substrates for lipoperoxidation (Fer et al., 2008). As reported by Cools et al. (2011), Shen et al. (2015), and Vitali et al. (2019), a n6:n3 ratio below 3:1 tends to induce oxidative stress. Tanghe et al. (2015) showed that the n6:n3 ratio of 4:1 or 6:1 had no remarkable influence on the oxidative damages, whereas Leskovec et al. (2019) indicated that the n6:n3 ratio of 5:1 did not increase lipid oxidation. Taken together, the ratio of n6:n3 should be above 3:1 to avoid lipoperoxidation and further oxidative stress due to excessive n-3 PUFA.

In this study, the sow diet with high starch as an energy source in late gestation and lactation was demonstrated to ameliorate litter performance in litter weight and stillbirth rate, decease liver glycogen, and downregulate the mRNA abundances of genes involved in fat synthesis and FA oxidation in liver. Meanwhile, the sow diet supplemented with fish oil was found to decrease preweaning mortality rate and the diet with a low n6:n3 ratio increased the oxidative status of sows and their offspring. This research has provided a scientific and theoretical basis for sow diets in late gestation and lactation and facilitates the choice of a suitable composition of energy sources and n6:n3 ratio to improve the performance and oxidative status of sows and their offspring.