Dietary fiber supplementation during the last 50 days of gestation improves the farrowing performance of gilts by modulating insulin sensitivity, gut microbiota, and placental function

Abstract

Our previous study found dietary konjac flour (KF) supplementation could improve insulin sensitivity and reproductive performance of sows, but its high price limits its application in actual production. This study aimed to investigate the effects of supplementation of a cheaper combined dietary fiber (CDF, using bamboo shoots fiber and alginate fiber to partially replace KF) from the last 50 days of gestation to parturition on farrowing performance, insulin sensitivity, gut microbiota, and placental function of gilts. Specifically, a total of 135 pregnant gilts with a similar farrowing time were blocked by backfat thickness and body weight on day 65 of gestation (G65d) and assigned to 1 of the 3 dietary treatment groups (n = 45 per group): basal diet (CON), basal diet supplemented with 2% KF or 2% CDF (CDF containing 15% KF, 60% bamboo shoots fiber, and 25% alginate fiber), respectively. The litter performance, insulin sensitivity and glucose tolerance parameters, placental vessel density, and short-chain fatty acids (SCFAs) levels in feces were assessed. The gut microbiota population in gilts during gestation was also assessed by 16S rDNA gene sequencing. Compared with CON, both KF and CDF treatments not only increased the piglet birth weight (P < 0.05) and piglet vitality (P < 0.01) but also decreased the proportion of piglets with birth weight ≤ 1.2 kg (P < 0.01) and increased the proportion of piglets with birth weight ≥ 1.5 kg (P < 0.01). In addition, KF or CDF supplementation reduced fasting blood insulin level (P < 0.05), homeostasis model assessment-insulin resistance (P < 0.05), serum hemoglobin A1c (P < 0.05), and the level of advanced glycation end products (P < 0.05) at G110d, and increased the placental vascular density (P < 0.05) at farrowing. Meanwhile, KF or CDF supplementation increased microbial diversity (P < 0.05) and SCFAs levels (P < 0.05) in feces at G110d. Notably, the production cost per live-born piglet was lower in CDF group (¥ 36.1) than KF group (¥ 41.3). Overall, KF or CDF supplementation from G65d to farrowing could improve the farrowing performance of gilts possibly by improving insulin sensitivity, regulating gut microbiota and metabolites, and increasing placental vascular density, with higher economic benefits and a similar effect for CDF vs. KF, suggesting the potential of CDF as a cheaper alternative to KF in actual production.

This work examined the impact of 2% konjac flour (KF) and 2% combined dietary fiber (CDF, containing 15% KF, 60% bamboo shoots fiber, and 25% alginate fiber) supplementation during the last 50 days of gestation on farrowing performance, insulin sensitivity, gut microbiota, and placental function of gilts. CDF dietary supplementation was found to have higher economic benefits than and a similar effect to KF on improving the reproductive performance of sows, thus a cheaper alternative to KF in actual production.

Introduction

Accumulated evidence supports progressive insulin resistance in the perinatal period, thus reducing the gilts’ reproductive performance (Père and Etienne, 2007; Miller et al., 2019). In our previous studies, konjac flour (KF), a dietary fiber with excellent hydration properties and fermentation characteristics, was shown to favor reproductive performance and insulin sensitivity of sows (Tan et al., 2016, 2018), but its high price limits its wide application. In addition, there is still a significant portion of cheap non-conventional dietary fiber sources, such as bamboo shoot fiber and alginate fiber. Their effect on improving glucose metabolism disorders in mice has been demonstrated (Li et al., 2016), but their intervention effects on reproductive performance and insulin resistance of gilts are unclear.

Dietary fiber was reported to positively influence insulin sensitivity by regulating gut microbiota (Torres-Fuentes et al., 2017; Huang et al., 2020). For example, adding dietary fiber to the pregnancy diet was reported to increase the number of Ruminococcaceae.UCG.005 and Ruminococcus to improve insulin sensitivity (Tan et al., 2016; Jarrett and Ashworth, 2018). In addition, dietary fiber can be fermented by gut microorganisms to produce short-chain fatty acids (SCFAs), thus activating G-protein-coupled receptors and triggering the secretion of gut hormones to improve insulin sensitivity (Canfora et al., 2015). However, the mechanism underlying the correlation of dietary fiber with reproductive performance and insulin sensitivity remains unclear.

Uteroplacenta, the only link between mother and fetus, plays an important role in fetal development, and enhanced placental function was reported to improve the sow’s farrowing performance by promoting fetal survival and growth (Hu et al., 2020). Studies in mice have shown that dietary fiber supplementation during gestation could improve placental nutrient transport capacity (Yan et al., 2011), while the dietary fiber effect on the placenta in sows has rarely been reported. In recent studies, decreased insulin sensitivity was found to be accompanied by trophoblasts’ impaired invasion and vascularization (Martino et al., 2016), leading to decreased placental efficiency and increased stillbirth rate (Tanaka et al., 2018).

Here, we hypothesize that a combination of several dietary fibers beneficial to insulin homeostasis in gilt diets may be a more feasible and economical strategy to improve their reproductive performance. This study tried to replace purified KF with a combined dietary fiber (CDF), a mixture of KF, bamboo shoots fiber, and alginate fiber at the ratio of 15%: 60%: 25%, with the purpose to investigate the impact of the supplementation of CDF and KF from the last 50 days of gestation to parturition on farrowing performance, insulin sensitivity, gut microbiota, and placental function of gilts.

Materials and Methods

Animal ethics

All procedures involving animals are carried out in accordance with the protocol approved by the Animal Ethics Committee of South China Agricultural University (2022f239).

Animals, diets, and management

This study was conducted in Jiangxi WanNianXinXing Agri-animal Co., Ltd., China (Jiangxi, China). On day 65 of gestation (G65d), a total of 135 pregnant gilts (Duroc-Landrace-Yorkshire) with a similar farrowing time were blocked by backfat thickness and body weight, and assigned to 1 of the 3 dietary treatment groups (n = 45 per group): basal diet (CON), basal diet supplemented with 2% KF or 2% CDF, respectively. KF and CDF were added at the expense of corn. Notably, CDF is a mixture of purified KF, bamboo shoots fiber, and alginate fiber at the ratio of 15%: 60%: 25%. The three diets were formulated based on the same nitrogen and energy level (Table 1) to meet or exceed the nutrient requirements for gilts as recommended by National Research Council (Council, 2012).

Table 1.

Ingredients and nutrient composition of the three experimental diets (as-fed basis)¹

Item	CON	KF	CDF
Ingredient, %
Corn	34.8	32.8	32.8
Rice bran meal	10.0	10.0	10.0
Broken rice	16.7	16.7	16.7
Soybean meal	14.5	14.5	14.5
Triticale	20.0	20.0	20.0
Konjac flour		2.0
Combined fiber			2.0
Mountain flour	1.1	1.1	1.1
Dicalcium phosphate	1.0	1.0	1.0
Sodium sulfate	0.4	0.4	0.4
Lysine	0.1	0.1	0.1
Choline chloride	0.1	0.1	0.1
Threonine	0.1	0.1	0.1
Mildewcide	0.1	0.1	0.1
Antiseptic	0.1	0.1	0.1
Premix2	1.0	1.0	1.0
Calculated composition3
Digestible energy, Mcal/kg	3.2	3.2	3.2
Crude protein, %	13.9	13.9	13.9
Ether extract, %	2.3	2.3	2.3
Crude fiber, %	3.2	3.6	3.7
Acid detergent fiber, %	4.0	4.4	4.6
Neutral detergent fiber, %	11.2	12.4	12.1
Ca, %	0.7	0.7	0.7
P, %	0.6	0.6	0.6
Lysine, %	0.8	0.8	0.8
Methionine, %	0.2	0.2	0.2
Threonine, %	0.6	0.6	0.6
Tryptophan, %	0.2	0.2	0.2
Analyzed composition
Crude protein, %	14.7	14.4	14.6
Crude fiber, %	3.5	3.7	3.6
Neutral detergent fiber, %	13.0	13.4	13.4
Soluble dietary fiber	0.9	2.3	1.3
Insoluble dietary fiber	20.7	20.5	21.3
Total dietary fiber	21.5	22.8	22.6

¹ CON = basal diet group, KF = konjac flour diet group; CDF = combined dietary fiber diet group (KF: bamboo shoots fiber: alginate fiber, 15%: 60%: 25%); KF ingredient contains 84.1% total dietary fiber, 74.2% soluble dietary fiber, and 9.8% insoluble dietary fiber; CDF ingredient contains 69.5% total dietary fiber, 19.2% soluble dietary fiber, and 50.3% insoluble dietary fiber.

² Provided the following per kilogram of diet: VA, 12,000 IU; VD₃, 4,800 IU; VE, 205 mg; VC, 200 mg; VK, 3.6 mg; VB₁, 3.6 mg; VB₂, 12 mg; VB₆, 7.2 mg; VB_12, 0.048 mg; pantothenic acid, 30.0 mg; nicotinic acid, 48.0 mg; folic acid, 8.6 mg; biotin, 0.6 mg; Cu, 10.0 mg; Fe, 130 mg; Zn, 60 mg; Mn, 45 mg; I, 0.3 mg; Co, 0.1 mg. Premix is provided by Xinxing Agriculture and Animal Husbandry Co., Ltd., Wannian County, Shangrao City (Jiangxi, China).

³ Calculated chemical concentrations using values for feed ingredients from NRC (2012). Amino acid levels in diets are expressed as totals.

Gilts were housed in individual stalls and fed once daily (06:00). Gestation diets were produced in pellet form, and the gilts were fed 2.0 kg/day from G65d to G100d, and 2.5 kg/day from G100d to parturition (Supplementary Table S1). Gilts were given free access to water throughout the experiment.

Intravenous glucose tolerance test (IVGTT)

On G100d, IVGTT was performed as previously reported (Bowe et al., 2014; Yang et al., 2019). Briefly, a total of 24 gilts with similar backfat, weight, and farrowing time were subjected to IVGTT (n = 8 per group). After an overnight fast, a catheter (Optiva 2, Johnson and Johnson, Ascot, Berkshire, UK) was placed into the central ear vein or the lateral ear vein of both ears, one of which was used for repeated blood sampling and the other for the infusion of glucose. Then, the fasting blood samples were collected before intravenous administration of 0.5 g glucose· kg body weight⁻¹ as 50% glucose solution (Sigma, USA). Blood was sampled from the ear vein at 0, 15, 30, 60, and 120 min after the onset of IVGTT, and glucose was measured immediately using an automatic blood glucose analyzer (Sinocare Inc., Changsha, China). For each IVGTT, the following parameters were calculated: area under the curve (AUC) for glucose (calculated by linear interpolation of glucose concentrations between the measurements, using the fasting glucose concentration as the baseline).

Measurements of reproductive performance and sample collection

On G100d, fasting blood samples (n = 8 per group) were collected from the gilts of each group via the ear vein using a venous blood collection needle (KINDLY GROUP, China) before feeding. Then, plasma was isolated by centrifugation at 1,500 × g and 4°C for 10 min and stored at −20°C until chemical analysis. Additionally, fresh fecal samples (n = 6 per group) were collected directly by massaging the rectum from G100 gilts, followed by immediate storage at −80°C until further analysis.

Changes in body condition and litter performance of gilts were recorded (Table 2 and Supplementary Table S2). Briefly, individual body weight and backfat thickness at the last rib were recorded for gilts on G65d and at 24 h post-farrowing. Immediately after birth, the piglets’ vitality was visually assessed based on a score described by Konig et al. (2021). Briefly, two independent observers blind to the experimental conditions assessed the vitality of the newborn piglets according to the following criteria: 0, no movement, no breathing after 15 s; 1, no movement after 15 s, but breathing or attempting to breathe; 2, piglet shows some movement within 15 s, breathing or attempting to breathe; 3, good movement, good breathing, and attempting to stand within 15 s. Then, the birth weight of each piglet was recorded. After parturition, total litter size, live-born piglets, stillborn piglets, mummified piglets, and litter weight of each gilt were recorded. The mean within-litter coefficient of variation of birth weight was calculated as the mean of the SD obtained for each litter (Wei et al., 2021).

Table 2.

Effects of KF and CDF supplementation during the last 50 days of gestation on the reproductive performance of gilts¹

Item	CON	KF	CDF	SEM	P-value
No. of litters	44	44	44
Litter size	10.8	10.4	9.6	0.24	0.16
No. of live-born piglets/litter	10.0	10.0	9.3	0.22	0.28
No. of mummified fetus/litter	0.1	0.1	0.1	0.24	0.86
The rate of stillbirth2, % (no. stillbirth in group)	5.4 (26)a	2.6 (12) ab	1.5 (7)b	0.19	<0.01
Birth weight of piglet, kg	1.4a	1.5b	1.5b	0.02	0.04
Birth weight of litter, kg	13.7	14.5	13.6	0.31	0.43
Placental weight, kg	3.3	3.0	3.0	0.05	0.22
Placental efficiency, %	4.2a	4.9b	4.5ab	0.08	<0.01
Piglet vitality3	2.4a	2.8b	2.6b	0.38	<0.01
Coefficient of variation in litter, %	14.7b	14.1ab	12.0a	0.46	0.04
Birth weight distribution of piglets2, %
<1.07, kg	12.0a	8.0b	6.7b	0.11	0.02
1.07–1.2, kg	13.6a	9.1b	9.9b	0.12	<0.01
1.2–1.5, kg	45.6	40.5	39.7	0.22	0.12
≥1.5, kg	28.8a	42.5b	43.8b	0.25	<0.01

¹CON = basal diet group, KF = konjac flour diet group; CDF = combined dietary fiber diet group (KF: bamboo shoots fiber: alginate fiber, 15%: 60%: 25%).

² Stillbirth rates and birth weight distribution of piglets were calculated using chi-square.

³The number of gilts in each treatment was 19, 20, and 20, respectively.

^a,b Indicates that the data of the same row with different small letters indicate a significant difference at P < 0.05.

The placenta samples were collected using a modified technique based on a previous study (Wilson et al., 1998). Briefly, during farrowing of gilts, umbilical cords were tied with a non-silk line, and each piglet was marked with a numbered tag to match the individual piglets with their placentae. After placental expulsion, the weight of the placenta and the weight of the corresponding piglet were recorded. Then, two placentas per litter were selected as samples from two piglets of a birth weight close to the average birth weight of the litter, with 12 placentas from 6 gilts in each group. Then, a 5-cm²-sized piece of tissue was cut in each placenta (3–4 cm from the cord insertion point), and a portion of it was placed in liquid nitrogen for rapid freezing, while the remaining tissue was immediately fixed in 4% paraformaldehyde (Hu et al., 2020). Moreover, the placental weights in all gilts were recorded and the placental efficiency was calculated by dividing piglet weight by placental weight.

Placental vascular density

The density of vessels was determined through image analysis by estimating the average value of six slices of one placenta (Hu et al., 2021; Konig et al., 2021). Briefly, fresh placental tissues fixed in 4 % paraformaldehyde were embedded in paraffin and sectioned at 5 μm thickness, followed by staining with hematoxylin and eosin. The area occupied by placental tissues and the placental vessels in these areas were traced using a projecting microscope (Olympus CX41, Japan). For each of the 5 μm sections, the total number of vessels in the placental stroma areas was determined and corrected with the total placental stroma areas measured (per unit area as mm²). Notably, the vascular density of two placentas from the same gilt was averaged and included in the follow-up statistics.

Chemical analyses

Fibrous ingredients (KF and CDF) and three diets were dried at 60 °C for 4 d by placing them in a heat-drying room and keeping the moisture at less than 5%, followed by grinding them separately through a 1-mm screen, and analyzing them in terms of crude protein (CP, AOAC method 990.03), crude fiber (CF, ISO 6865-2000), and neutral detergent fiber (NDF, ISO 16472-2006). Total dietary fiber, insoluble dietary fiber, and soluble dietary fiber were obtained by the AOAC method 991.43 (Nguyen et al., 2019). Briefly, three enzymes were used to hydrolyze samples under different conditions (a heat-stable α-amylase; a protease; and an amyloglucosidase) (Aladdin, China). After this enzymatic digestion, the dietary fiber fractions were obtained by removing the non-dietary fiber component from the indigestible residues. Then, the insoluble dietary fiber was recovered by filtration of the digested samples, and soluble dietary fiber was precipitated from the filtrate by adding 96% ethanol. The residual ash and protein contents were determined from the fiber residues for corresponding data correction. Total dietary fiber is defined as the sum of soluble and insoluble dietary fiber. All procedures were performed in duplicate.

The fasting plasma glucose (FPG) concentration was determined using a glucose dehydrogenase activity colorimetric assay kit as instructed by the manufacturer (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). The fasting plasma insulin (FPI) level using an ultrasensitive pig insulin ELISA kit (Wuhan MskbioBiotechnology Institute, Wuhan, China). The hemoglobin A1c (HbA1C) and advanced glycation end products (AGEs) were measured by using the kit (Jiangsu Meimian Industry Co., Ltd, Jiangsu, China). Insulin resistance and sensitivity were evaluated through homeostasis model assessment (HOMA): HOMA − insulin resistance (HOMA − IR) = [fasting insulin (mIU/L) × fasting glucose (mmol/L))/22.5] (the coefficient 22.5 is a correction factor, which is defined as 5 μU/mL plasma insulin corresponding to a blood glucose level of 4.5 mmol/L in a normal desirable individual) (Hare et al., 2022). All procedures were performed in duplicate.

Fecal short-chain fatty acids (SCFAs) analysis

Fecal SCFAs concentrations were measured as previously reported (Liu et al., 2021). Briefly, fecal samples (n = 6 per group) were thawed on wet ice, followed by placing 0.5 g of fecal sample in a centrifuge tube with 1 mL of distilled water, and homogenizing the mixture by vortexing for 1 min. After heating in an ultrasonic bath for 30 min, the mixture was centrifuged at 5,000 × g for 10 min to obtain the supernatant, followed by pouring all the supernatant into a new 2 mL centrifuge tube, adding 20 μL of 25% metaphosphoric acid and 0.25 g anhydrous sodium sulfate, and homogenizing the mixture thoroughly by vortexing for 1 min. After adding 1 mL methyl tert-butyl ether (operated in the fume hood), the mixture was thoroughly homogenized by vortexing for 5 min, followed by centrifugation at 5,000 × g for 5 min and collecting the upper methyl tert-butyl ether extract (operated in fume hood). After filtration through a 0.22 μm microporous membrane, the filtrate was placed in a sample bottle with a lined tube for fecal SCFAs analysis.

Fecal 16S rDNA gene sequencing and bioinformatics analysis

Total genome DNA of fecal samples (n = 6 per group) was extracted using the CTAB/SDS method. Briefly, DNA concentration and purity were monitored by 1% agarose gel electrophoresis. According to the concentration, DNA was diluted to 1 ng/μL using sterile water, followed by amplifying 16S rRNA/18SrRNA/ITS genes of distinct regions (16S V4/16S V3/16S V3-V4/16S V4-V5,18S V4/18S V9, ITS1/ITS2, Arc V4) using a specific primer with the barcode (e.g. 16S V4: 515F-806R, 18S V4: 528F-706R, 18S V9: 1380F-1510R, etc.). All PCR reactions were carried out with 15 μL of Phusion High-Fidelity PCR Master Mix (New England Biolabs), 2 μM of forward and reverse primers, and ~ 10 ng template DNA. After mixing the same volume of 1× loading buffer (containing SYB green) with PCR products, electrophoresis was performed on 2% agarose gel for detection. After mixing PCR products at an iso-density ratio, the mixture was purified with Qiagen Gel Extraction Kit (Qiagen, Germany). The sequencing libraries were generated following the manufacturer’s protocol of TruSeq DNA PCR-Free Sample Preparation Kit (Illumina, USA) and index codes were added. The library quality was assessed on the Qubit@ 2.0 Fluorometer (Thermo Scientific) and AgilentBioanalyzer 2100 system. The number of valid tags for two samples was less than 30,000 and the quality is unqualified, thus unable to meet the requirements for inclusion in the follow-up analysis. Finally, the library was sequenced on an Illumina NovaSeq platform and 250 bp paired-end reads were generated.

Bioinformatics analysis was performed as previously described (Wang et al., 2018). Briefly, the Shannon index, Chao1 index, and observed species were calculated using QIIME. Unweighted Unifrac principal coordinates analysis (PCoA) was performed with QIIME using the unweighted UniFrac distance matrix between the samples. Linear discriminant analysis (LDA) effect size (LEfSe) was used to elucidate the differences of bacterial taxa, with an LDA score ≥3 as an important contributor to the model.

Western blotting

Total proteins were extracted from the placental samples following the manufacturer’s guide for the protein extraction kit (Beyotime, Beijing, China). Briefly, an amount of 10 μg protein was loaded and separated by SDS–PAGE gel electrophoresis, followed by transferring the proteins onto the polyvinylidene fluoride membranes (Merck Millipore). After blocking with TBS/T buffer containing 5% milk, the membranes were incubated with the primary antibodies against vascular endothelial growth factor A (VEGF-A) (19003-1-AP, Proteintech, USA, 1:1,000), platelet endothelial cell adhesion molecule (CD31) (A0378, Abclonal, China, 1:1,000), and β-actin (4970, CST, USA, 1:1,000). Subsequently, the membranes were incubated with appropriate HRP-conjugated anti-rabbit IgG secondary antibody (AS014, Abclonal, China, 1:5,000). Images were captured using the ChemiDoc MP system (Bio-Rad, Hercules, CA, USA), and band densities were quantified using Image Lab software (Bio-Rad, Hercules, CA, USA) and then normalized to β-actin content.

Economic returns of feeding dietary fiber

In this study, economic returns were approximated by a partial budget analysis of marginal costs. Dietary supplementation with KF or CDF increased gilt feed costs (Table 3). We only calculated the sow feed cost per gilt producing each live-born piglet and gilt culling cost for each diet group. This information was obtained through direct interviews with pig farm managers and feed sellers. Notably, fixed costs such as treatment cost, labor cost, depreciation of fixed assets, and maintenance cost did not change between treatment groups in this experiment, so they were not included in the cost analysis.

Table 3.

Economic benefit analysis of KF and CDF supplementation during the last 50 days of gestation¹

Item	KF	CDF	Unit price, ¥/kg
Fibrous ingredient, %
KF	2.0	0.0	55.0
CDF	0.0	2.0	12.0
Number of gilts culled2	1	1
Gilt culling costs, ¥	1,100	1,100
Live-born piglets/litter, n	10.0	9.3
Feed cost
Feed price, ¥/1,000 kg	3830.9	2990.9
Feed consumption during test per gilt, kg	107.8	107.8
Feed cost per gilt, ¥	413.0	322.4
Production cost per live-born piglet, ¥	41.3	34.7

¹ KF = konjac flour diet group; CDF = combined dietary fiber (KF: bamboo shoots fiber: alginate fiber, 15%: 60%: 25%) diet group.

² Gilts were excluded before farrowing because of serious lameness and death.

Statistical analysis

A total of 132 gilts (n = 44 per group) completed the experiment (three gilts were excluded before farrowing because of serious lameness and death) and were used for reproductive performance data analysis. Chemical analyses and fecal SCFAs analyses were performed in duplicate and then averaged for statistics. Before analysis, all data were tested for normality and homogeneity of variance using the Kolmogorov–Smirnov and Levene tests (with the significance level set at 5%) in SPSS 24.0 (SPSS, Inc, Chicago, USA). Data were presented as mean or mean ± SE of the mean (SEM) and statistically analyzed using one-way analysis of variance (ANOVA) and Duncan’s multiple range test in SPSS 24.0. Tamhane’s T2 test was used to assess variance heterogeneity. The stillbirth rate and birth weight distribution of piglets were analyzed using the Chi-square test. Differences were considered significant at P < 0.05, and a tendency was considered at 0.05 ≤ P < 0.1. Spearman’s correlational analysis was also used to examine potential effects of the relative abundance of selected bacterial genera and SCFAs levels in feces on HOME-IR, litter performance, and placental vessel density. The results with R > 0.6 and P < 0.05 were considered to have a strong correlation.

Results

Dietary fiber supplementation improves farrowing performance of gilts

The three groups were shown to have no differences (P > 0.05; Table 2) in the number of live-born piglets, placental weight, and birth weight of litters (P > 0.05). However, compared with CON group, KF and CDF groups exhibited greater birth weight of piglets (P < 0.05), placental efficiency (P < 0.01), and the piglet vitality (P < 0.01), but less stillbirth rate (P < 0.05). Meanwhile, KF and CDF treatments showed a smaller (P < 0.05) proportion of piglets with birth weight ≤ 1.2 kg, while a greater (P < 0.05) proportion of piglets with birth weight ≥ 1.5 kg.

Economic benefit analysis

Gilt culling costs did not differ between the three groups, with only one sow culled in each group. However, compared to KF group, the CDF group had lower feed cost per gilt (CDF vs. KF, ¥ 322.4 vs. ¥ 413.0) and production cost per live-born piglet (CDF vs. KF, ¥ 34.7 vs. ¥ 41.3) (Table 3).

Dietary fiber supplementation improves the insulin sensitivity and glucose tolerance of gilts

The three groups showed a similar AUC of glucose from 0 to 120 min (Figure 1A; P > 0.05) after intravenous administration of glucose solution, with no differences in FPG (P > 0.05; Figure 1B). However, KF and CDF groups were lower than the CON group in FPI and HOMA-IR values (P < 0.05; Figure 1C and D). Moreover, KF treatment had better performance than CDF treatment in enhancing insulin sensitivity. Furthermore, when compared with the CON group, both KF and CDF groups had lower concentrations of HbA1C and AGEs (P < 0.05; Figure 1E and F).

Figure 1.

Effects of konjac flour (KF) and combined dietary fiber (CDF) supplementation during the last 50 days of gestation on insulin sensitivity and glucose tolerance status of gilts. (A) Intravenous glucose tolerance test (IVGTT) from 0 to 120 min and area under the curve. (B) Fasting plasma glucose (FPG) level. (C) Fasting plasma insulin (FPI) level. (D) Homeostasis model assessment of insulin resistance (HOMA-IR). (E) Serum hemoglobin A1c (HbA1C). (F) Advanced glycation end products (AGEs) levels. ^a,bMean values with different small letters indicate a significant difference at P < 0.05.

Dietary fiber supplementation increases fecal acetate and total SCFAs of gilts

Compared with CON treatment, KF and CDF treatments had higher concentrations of acetate and total SCFAs in feces of gilts (P < 0.05; Table 4), but showed no significant changes in the concentrations of propionate, butyrate, isobutyrate, pentanoate, and isovalerate (P > 0.05).

Table 4.

Effects of KF and CDF supplementation during the last 50 days of gestation on fecal SCFAs of gilts¹

Item	CON	KF	CDF	SEM	P-value
No. of gilts	6	6	6
Acetate, μg/g	1272.9a	1876.6b	1795.7b	78.5	<0.01
Propionate, μg/g	788.8	899.9	927.8	30.3	0.14
Butyrate, μg/g	423.7	453.5	437.7	19.6	0.84
Isobutyrate, μg/g	107.5	113.2	128.5	6.0	0.35
Valerate, μg/g	144.6	154.9	168.3	8.1	0.52
Isovalerate, μg/g	156.6	173.0	186.7	9.1	0.42
Total SCFAs, μg/g	2894.2a	3671.0b	3644.8b	129.0	0.01

¹CON = basal diet group, KF = konjac flour diet group; CDF = combined dietary fiber diet group (KF: bamboo shoots fiber: alginate fiber, 15%: 60%: 25%).

^a,bMean values with different small letters indicate a significant difference at P < 0.05.

Dietary fiber supplementation changes the fecal microbiota of gilts

The dilution curve and hierarchical clustering curve based on the whole samples indicated that with the increase of sequencing number, the curve tended to be flat and finally remained at a certain level, indicating that the sequencing amount was saturated, and the species richness and uniformity of the measured samples were relatively high, which was asymptotically reasonable (Supplementary Figure S1). In the PCoA analysis chart (Figure 2A), all the three treatments showed obvious separation, with significant differences among them. Community composition at the phylum level indicated that for all the three treatments (Figure 2B), the microbiota was dominated by Firmicutes (53.06–58.89%), Euryarchaeota (15.61–26.90%), Bacteroidetes (7.30–10.34%), Proteoidetes (3.50–5.28%), and Spirochaetes (0.56–4.00%). Moreover, compared with CON group, both KF and CDF groups showed higher Chao 1 index (P < 0.05; Figure 2C), Shannon index (P < 0.05; Figure 2D), and observed species (P < 0.05; Figure 2E).

Figure 2.

Effects of konjac flour (KF) and combined dietary fiber (CDF) supplementation during the last 50 days of gestation on microbiota alpha-diversity in feces of gilts. (A) In the unweighted Unifrac distance PCoA analysis, with abscissa for one principal component, ordinate for another principal component, and percentage for the contribution value of the component to sample difference. (B) Phylum horizontal relative abundance histogram, with abscissa for the sample name, ordinate for the relative abundance, and others for the sum of the relative abundance of all phyla except for the 10 phyla in the figure. (C) Shannon index. (D) Chao1 index. (E) Observed species. Each gilt was regarded as an experimental unit (n = 5, 6, and 5, respectively). ^a,bMean values with different small letters indicate a significant difference at P < 0.05.

LEfSe analysis revealed the 12 core genera among the three groups (Figure 3A). Abundance comparison of predominant genera showed the enrichment of Clostridium in CON group (Figure 3B), Lachnospiraceae_NK4A136_group, Sarcina, Ruminococcus, Christensenellaceae_R_7_group, Romboutsia, Clostridium_sensu_stricto_1, and Terrisporobacter, and Treponema in KF group (Figure 3C-J), while Cetobacterium, Pseudomonas, and Escherichia in CDF group (Figure 3K-M).

Figure 3.

Effects of konjac flour (KF) and combined dietary fiber (CDF) supplementation during the last 50 days of gestation on core genera in feces of gilts. (A) The most differentially abundant taxa between three groups identified by linear discriminant analysis (LDA) effect size. Only taxa with an LDA threshold ≥ 3.5 are shown. (B–M) Relative abundances of the 12 core genera. Each gilt was regarded as an experimental unit (n = 5, 6, and 5, respectively). ^a,bMean values with different small letters indicate a significant difference at P < 0.05.

Dietary fiber supplementation enhances placental angiogenesis of gilts

The effects of dietary fiber supplementation on placental function were further explored by analyzing the placental vascular density and the expression of angiogenesis-related genes. Compared with KF and CDF groups, CON group also showed lower placental vessel density (Figure 4A and B; P < 0.05) and lower protein abundances of VEGF-A and CD31 (Figure 4C–E; P < 0.05).

Figure 4.

Effects of konjac flour (KF) and combined dietary fiber (CDF) supplementation during the last 50 days of gestation on placental angiogenesis of gilts. (A) The hematoxylin and eosin methods were used to examine blood vessel density in the CON, KF, and CDF placental tissues, with black arrows for placental blood vessels (bar = 50 μm). (B) The percentage of blood vessels in the placental tissues. Each gilt was regarded as an experimental unit (n = 6 per group). (C) Western blot analysis of angiogenesis-related factors and protein expression in different placentae. (D) Protein expression levels of CD31 and VEGF-A. Each placenta was regarded as an experimental unit (n = 12 per group). ^a,bMean values with different small letters indicate a significant difference at P < 0.05.

Gut microbiota and SCFAs are related to HOMA-IR and placental function

Ruminococcus and acetate had a significant positive correlation with placental vascular density (P < 0.01; Figure 5) and piglet vitality (P < 0.05), while a significant negative correlation with HOMA-IR (P < 0.01). Clostridium was negatively correlated with placental vascular density (P < 0.05).

Figure 5.

Spearman’s correlation coefficient analysis of the impact of relative abundance of selected bacterial genera and short chain fatty acids (SCFAs) in feces on homeostasis model assessment of insulin resistance (HOME-IR), litter performance, and placental vessel density. *P < 0.05 and **P < 0.01.

Discussion

Our previous study demonstrated that KF could improve the reproductive performance of sows, but the high price made it difficult to be widely applied (Sun et al., 2015; Tan et al., 2015, 2016). In the present study, we tried to replace purified KF with a combined dietary fiber (CDF), a mixture of KF, bamboo shoots fiber, and alginate fiber at the ratio of 15%: 60%: 25%. We selected bamboo shoots and alginate fiber based on their multiple beneficial aspects: 1) cheap and accessible non-conventional fibrous materials in China; 2) similar physicochemical properties to KF, i.e. high fermentability and suitable hydration properties as detected in our previous study (Huang et al., 2022); 3) improved glycemic control and dyslipidemia in high-fat diet fed mice (Li et al., 2016). Here, we investigated the impact of 2% CDF and 2% KF supplementation during the last 50 days of gestation on farrowing performance, insulin sensitivity, placental function, and gut microbiota of gilts. We noted that the two fibrous ingredients differed in that the KF diet had a higher composition of soluble dietary fiber, while the CDF diet had a higher percentage of insoluble dietary fiber. Previous studies have reported that both soluble dietary fiber and insoluble dietary fiber could improve insulin sensitivity and reproductive performance in sows through intestinal health (Li et al., 2019). The related mechanisms were shown to include the ready fermentation of soluble dietary fiber by microbiota to produce SCFAs, which are beneficial to host health, as well as the binding of insoluble dietary fiber with water, thereby increasing fecal volume and promoting normal defecation, contributing to reduce endotoxin production and absorption and inhibit pathogen colonization in the intestine (Li et al., 2019). In this study, we observed that KF or CDF supplementation from G65d to farrowing had a similar effect in reducing stillbirth rates and increasing piglet birth weight of gilts, possibly by regulating gut microbiota and metabolites, improving insulin sensitivity, and increasing placental vascular density. Compared with KF, CDF had higher economic benefits (lower feed cost per gilt and production cost per live-born piglet) and a similar effect to KF, suggesting its potential as a cheaper alternative to KF in actual production.

Uteroplacenta, the key interface between mother and fetus, plays an important role in fetal development (Hu et al., 2020), suggesting that placental function is a key factor affecting reproductive performance, such as placental efficiency and piglet birth weight. In the present study, KF and CDF supplementation were observed to increase placental efficiency and piglet birth weight, which is in line with a previous report (Mahany et al., 2018). Meanwhile, the placental function is influenced by placental vascular growth and placental vascularization, which are regulated by angiogenic factors, such as VEGF-A (Zhang et al., 2014; Hu et al., 2019; Huang et al., 2021b). In this study, we verified that the protein abundance of CD31, a marker of endothelial cells, was increased in the placentae of KF and CDF groups, suggesting that dietary fiber could promote placental angiogenesis. A richer vascular system allows the efficient delivery of oxygen and nutrients from mother to fetus, thereby increasing piglet vitality (Huang et al., 2021a; Konig et al., 2021). We also found that KF and CDF treatments could decrease stillbirth rate. A possible explanation is the increased placental angiogenesis. A previous study on pregnant women showed a significant correlation between a reduced placental growth factor/anti-angiogenic factor ratio and the incidence of stillbirth (Chaiworapongsa et al., 2013). In addition, the stillbirth rate is strongly influenced by the size of the uterine volume. A previous study has indicated that at moderate intrauterine crowding, fetal size reached a peak and competition for intrauterine space between embryos brought by further crowding beyond this point may reduce the number of viable embryos/fetuses (Vallet et al., 2014). Moreover, we were unable to determine whether KF and CDF reduced the duration of labor or increased the intensity of uterus contractions in this study, which may also have a significant effect on stillbirth rate (Theil et al., 2022). These possibilities need to be further clarified in future studies.

Glucose intolerance or insulin resistance during gestation was also considered as an adverse effect on fetal outcomes such as high stillbirth rate and low birth weight (Cheng et al., 2020). During gestation, the gilt undergoes the physiological and metabolic processes such as progressive and reversible glucose intolerance and insulin resistance (Yuan et al., 2020; Yang et al., 2021b). To investigate whether the supplementation with KF or CDF during the last 50 days of gestation could have similar effects on improving maternal insulin sensitivity, we examined parameters related to glucose metabolism in the different treatment groups. Our results showed that KF and CDF supplementation could significantly reduce the values of HOMA-IR, FPI, serum HbA1C, and AGEs of G110d gilts, indicating the improvement of insulin sensitivity. Notably, KF and CDF improved insulin sensitivity, which is consistent with the changing pattern in placental efficiency and placental vascular density, also agreeing with previous reports (Martino et al., 2016; Tanaka et al., 2018). Therefore, we hypothesize placenta as a central target for the response and regulation of maternal insulin sensitivity. In a previous study, researchers attributed changes in insulin sensitivity during late gestation to the increased production of placental tumor necrosis factor-α, which negatively regulates placental insulin signaling, thus reducing maternal insulin sensitivity (Kirwan et al., 2002). Meanwhile, impaired insulin signaling in placenta may in turn negatively affects the development of vascular system. Our unpublished data demonstrated that the secretions of angiogenesis-related signaling pathways and regulatory factors were dramatically negatively affected in an in vitro high-glucose-induced insulin resistance model in ­porcine placental endothelial cells. However, the logical relationship between maternal and placental insulin sensitivity with placental angiogenesis needs to be further clarified.

Our and other previous studies demonstrate that dietary fiber supplementation can affect insulin sensitivity by modulating gut microbiota (Tan et al., 2016; Cheng et al., 2018). Therefore, 16S rRNA sequencing was used to determine whether KF and CDF supplementation could affect the microbiome diversity and composition of fecal samples in gilts. We found that KF and CDF could increase Chao 1 index, Shannon index, and observed species, indicating an improved gut microbiota structure, agreeing well with the results of our previous study (Cheng et al., 2018). LEfSe analysis further showed that KF promoted the enrichment of Lachnospiraceae_NK4A136_group, Sarcina, Ruminococcus, Christensenellaceae_R_7_group, Romboutsia, Clostridium_sensu_stricto_1, Terrisporobacter, and Treponema, while CDF promoted the enrichment of Cetobacterium, Pseudomonas, and Escherichia in CDF group, suggesting that the response of gilts to different dietary fibers during pregnancy is inconsistent. This difference can be largely attributed to the source, type, and physicochemical properties of the dietary fiber. Specifically, dietary fibers with low hydration properties, such as bamboo shoots, may affect the colonization of beneficial bacteria by flushing the mucous layer, while highly fermentable dietary fibers, such as KF, can increase their degradation rate by microorganisms and promote the proliferation of microorganisms such as Lactobacillus and Bifidobacterium (Huang et al., 2022). Moreover, we note that some dominant genera are the previously reported producers of SCFAs, such as Lachnospiraceae_NK4A136_group, Ruminococcus, and Christensenellaceae_R_7_group (Li et al., 2021). These data prompted us to assess changes in SCFAs in feces of gilts, and we found that dietary fiber-fed gilts produced more acetate and total SCFAs. SCFAs have been proposed to regulate insulin sensitivity (Zaky et al., 2021) as signal molecules, which is the most abundant metabolite in the gut and acetate has been reported to improve the insulin-resistant state by regulating inflammation and oxidative stress in peripheral tissues, ultimately improving insulin sensitivity (Hernández et al., 2019; Yuan et al., 2020; Jang and Lee, 2021). In the present study, Ruminococcus and acetate had significant negative correlation with HOMA-IR, suggesting that increasing acetate-producing bacteria and acetate production may be one of the ways for dietary fiber supplementation during pregnancy to improve the gut health of gilts.

Several studies have reported the role of the gut-placenta axis in metabolic disorders during pregnancy and its related potential mechanisms (Yang et al., 2021a). For instance, the gut microbiota of preeclamptic patients could significantly increase the blood pressure in pregnant mice and gut barrier damage, allowing intestinal bacteria to colonize the placenta, leading to metabolic abnormalities during pregnancy (Chen et al., 2020). Moreover, propionate and SCFAs promoted autophagy and M2 polarization of placental bed macrophages and trophoblast invasion, thereby inhibiting placental inflammation, and improving spiral artery remodeling, contributing jointly to resist the onset of preeclampsia (Jin et al., 2022). In this study, Ruminococcus and acetate had significant positive correlation with placental vascular density and piglet vitality, and our findings enrich the gut–placenta axis theory and contribute to the development of microecological products for improving insulin sensitivity in gilts. However, more comprehensive studies need to be performed on the regulation of placental function and insulin sensitivity in gilts by Ruminococcus and acetate.

In addition, we note that fiber fractions (CF and NDF) differ significantly in the analyzed and calculated values, with higher analyzed values. There are two possible explanations for this difference: 1) before chemical analysis, feed samples were dried to constant weight and their moisture was preserved at ~5%, which is lower than the default moisture content at the time of calculation (~10%); 2) nutritional parameters varied in different batches of raw materials when feed manufacturers configured the diets.

Conclusion

Dietary KF or CDF supplementation from G65d to farrowing showed a similar effect in reducing stillbirth rate and increasing piglet birth weight of gilts, possibly by regulating gut microbiota and metabolites, improving insulin sensitivity, and increasing placental vascular density of gilts. Moreover, CDF has higher economic benefits than and a similar effect to KF, suggesting its potential as a cheaper alternative to KF in actual production.

Glossary

Abbreviations

AGEs

advanced glycation end products

ANOVA

analysis of variance

AUC

area under the curve

CD31

platelet endothelial cell adhesion molecule

CDF

combined dietary fiber

CON

control

DF

dietary fiber

FPG

fasting plasma glucose

FPI

fasting plasma insulin

HbA1C

hemoglobin A1c

HOMA

homeostasis model assessment

HOMA-IR

homeostasis model assessment-insulin resistance

IVGTT

intravenous glucose tolerance test

KF

konjac flour

LDA

linear discriminant analysis

LEfSe

linear discriminant analysis effect size

NRC

national research council

PCoA

principal coordinates analysis

SCFA

short-chain fatty acid

VEGF-A

vascular endothelial growth factor A

Conflict of Interest Statement

We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, and there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the content of this paper.