Effects of Chronic Exposure to Diets Containing Moldy Corn or Moldy Wheat Bran on Growth Performance, Ovarian Follicular Pool, and Oxidative Status of Gilts

Background: We investigated the effect of replacing normal corn (NC) or normal wheat bran (NW) with moldy corn (MC) or moldy wheat bran (MW) on growth, ovarian follicular reserves, and oxidative status. Methods: Sixty-three Landrace × Yorkshire gilts were assigned to seven diets formulated by using MC to replace 0% (control), 25% (25% MC), 50% (50% MC), 75% (75% MC), and 100% NC (100% MC), MW to replace 100% NW (100% MW), and MC and MW to replace 100% NC and 100% NW (100% MC + MW), from postnatal day 110 to day 19 of the second estrous cycle. Results: Feeding the gilts with MC or MW induced a lower average daily gain at days 29–56 of the experiment. Age at puberty remained unchanged, but MC inclusion resulted in a linear decrease in antral follicles with diameter >3.0 mm, and control gilts had a 12.7 more large antral follicles than gilts in the 100% MC + MW treatment. MC inclusion linearly decreased the numbers of primordial follicles, growing follicles, and corpora lutea, associated with a lower anti-Müllerian hormone level in serum and 17β-estradiol level in follicular fluid. MC inclusion decreased the serum concentrations of insulin-like growth factor 1 and its mRNA levels in the liver, combined with higher malondialdehyde concentration and lower total superoxide dismutase activities in serum and liver. Conclusion: Chronic exposure to MC-containing diets caused the loss of follicles, even if levels of deoxynivalenol, zearalenone, and aflatoxin B1 were below the levels allowed by China and Europe standards.


Introduction
The annual culling rate of sows in swine production reaches~50% worldwide, including in China [1], Europe [2], and the USA [3]. Thus, the breeding herd has to be replenished by young replacement gilts. Consequently, the successful rearing of replacement gilts is important for the overall productive performance of the breeding herds. The successful development of gilts requires not only suitable growth and accretion of body tissues but also the synchronization of sexual development (e.g., ovarian and uterine development). The recruitment of small antral follicles to form large antral follicles is essential for the attainment of puberty for gilts. Additionally, ovarian follicle reservation can determine the lifetime fertility of mammals [4] because there are no functional oogonial stem cells in ovaries to generate new oocytes in postnatal mammals [5,6]. A large loss of oocytes during the replacement phase as gilts may have both short-term and long-term negative effects on the reproductive capacity as sows [7].
In this study, we assessed the effects of mycotoxins produced by molds (fungi) on the development of gilts, as mycotoxins are estimated to contaminate over 25% of the world's crops [8,9]. Meanwhile, mycotoxins may induce substantial impairment of ovarian follicular development. Numerous studies have investigated the effects of mycotoxins, such as deoxynivalenol (DON), zearalenone (ZEA), aflatoxin B1 (AFB1), ochratoxin A (OA), fumonisins (FUM), and T-2 toxin (T-2), on the growth and health of pigs [10][11][12][13][14]. Awareness of the negative effect of mycotoxins on replacement gilts is growing due to the particular actions of these toxins on the reproductive tract. For example, ZEA and DON belong to the trichothecene group of mycotoxins [10]. In vivo and in vitro experiments indicated that ZEA and its metabolites exert estrogenic effects, resulting in functional and morphological alterations in ovarian follicles and oocytes, and ZEA is unequivocally implicated in reproductive disorders of swine and other domestic animals [15][16][17]. However, it is unclear whether mycotoxins have dose-dependent effects on the ovarian reserves of gilts.
Corn is the most widely available feed to provide dietary energy for domestic animals, but it is usually contaminated with mycotoxins that limit its use in swine production [18]. Wheat is another ingredient commonly contaminated by molds, and mycotoxin contamination is even greater in its byproduct wheat bran [19,20]. In grains, ZEA is often found with DON or other mycotoxins as co-contaminants because they are produced by the same fungal species [10]. Compared with a single mycotoxin, multiple mycotoxins contained in naturally moldy feed have more severe toxicity due to their synergism. However, most studies of the effects of mycotoxins on the growth and development of pigs added a purified mycotoxin to the diet [13,16,[21][22][23], and the hygiene standards in China and the EU for porcine diets were established according to results based on experiments conducted with purified mycotoxin. To date, the dose-dependent effects of including different levels of mycotoxin-contaminated cereal grains on the growth performance and ovarian follicle reserve in replacement gilts is poorly understood. Therefore, we conducted this study to investigate the effects of different levels of moldy corn (MC) or moldy wheat bran (MW) on the growth and ovarian follicular development in replacement gilts.

Animal Ethics
All experimental procedures were approved on 2 Mart 2015 by the Animal Care and Use Committee of Sichuan Agricultural University (Approval No. 20150122), and they were conducted in accordance with the National Research Council's Guide for the Care and Use of Laboratory Animals.

Preparation of Normal Corn (NC) and Wheat Bran (NW)
The corn and wheat bran were purchased from northeast China where the weather is dry and crisp to ensure that the grains had a low level of mycotoxin contamination according to our experience. The normal corn (NC) and wheat bran (NW) were placed in an environment-controlled room with temperature 28-32 • C and humidity 85-90% for 1 week to prepare MC and MW. In our preliminary study, this regimen induced the production of mycotoxins. Samples of NC and NW were left in a dry and cool place to serve as the control.

Animals, Diet, and Treatment
Sixty-three gilts (Landrace × Yorkshire, aged 110 ± 8 days, initial body weight (BW) 48.50 ± 3.61 kg) were used in this study. The gilts were from different dams to exclude any possible maternal effects. The control diet was formulated for two feeding phases (Table 1). In phase 1 from 1 to 28 days of the experiment, the feed contained 66.7% corn and 10.0% wheat bran to provide 3.22 Mcal/kg digestible energy (DE), 15.37% crude protein, and 0.70% standardized ileal digestible (SID) lysine. In phase 2 from 29 days to the end of the experiment, the diet contained 70.7% corn and 10.0% wheat bran to provide 3.16 Mcal/kg DE, 14.22% crude protein, and 0.63% SID lysine. Other nutrients were added to meet or exceed the requirements for growing replacement gilts recommended by NRC (2012). To create seven dietary treatments, the NC in the control diet was replaced with MC at 25% (25% MC), 50% (50% MC), 75% (75% MC), and 100% (100% MC), the NW in the control diet was 100% replaced with MW to form the 100% MW group, and NC and NW were 100% replaced with MC and MW to form the 100% MW + MC group. Gilts were randomly assigned to one of the seven dietary groups with three replicates per treatment and three gilts per replicate for days 1-28 of the experiment. The gilts then were individually housed from day 29 until the end of the experiment.

Analysis of Mycotoxin Contents
The formulated diets were stored at 4 • C until feeding, and feed samples were collected for storage at −20 • C until future analysis of mycotoxin contents, including DON, ZEA, AF, FUM, OA, and T-2, using high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). Briefly, feed samples were sequentially extracted, and then the extracts were pooled and cleaned using a multifunction cleaning column (MycoSpinTM400, R-Biopharm Rhone, Darmstadt, Germany) before being separated by a TC-C18 chromatographic column. The MS was conducted using electrospray ionization and multiple reaction monitoring models, according to the internal standard method. The detection limit for all specimens was 0.1 µg/kg at a sample weight of 5 g. Table 2 shows the mycotoxin content of NC, NW, MC, MW, and each diet. For days 1-28 and days 29-56 of the experiment, the gilts were provided with feed ad libitum. They then were provided with 2.5 kg/day feed from day 57 to the end of the experiment at the 19th day of the second estrous cycle. Half of each day's feed was provided equally twice daily at 8:00 a.m. and 4:00 p.m. Water was provided ad libitum throughout the entire experiment. The environment temperature and humidity were maintained at 20-24 • C and 50-60%, respectively. Artificial light was provided from 7:00 a.m. to 7:00 p.m.

Detection of Pubertal Onset and Growth Performance
All gilts were exposed (but separated by a fence) to mature boars to encourage pubertal onset at day 29 of the experiment. The naturally occurring estrous was carefully checked by one experienced stockperson on the basis of the behavioral and vulvar characteristics of the gilts [24]. Age at first observed estrous was recorded, and the day of standing heat was designed as day 1 of the first estrous cycle. The BW and backfat (BF) thickness were recorded at the beginning of the experiment, on days 28, 56, and 96, and at puberty. The BF thickness was measured at 65 mm on both sides of the dorsal midline at the last rib (P 2 ) using an ultrasound scanner (Renco LeanMeater). The average daily gain (ADG) for days 1-28, 29-56, and 57-96 of the experiment was calculated, and the feed intake for days 1-28, 29-56, and 57-96 was recorded to calculate the average daily feed intake (ADFI). The feed/gain (F/G) ratio for each respective period was also calculated. The ADFI, ADG, and F/G ratio for days 1-28 of the experiment are presented for each pen as a unit; the individual gilt was the experimental unit for all parameters analyzed from day 29 to the end of the experiment.

Sample Collection
Blood samples before the morning meal and at 60 and 120 min after the morning meal were collected on day 18 of the second estrous cycle to measure follicle-stimulating hormone (FSH) and luteinizing hormone (LH) levels. Blood samples were also collected at day 19 of the second estrous cycle at 8:00 a.m. before the morning meal. Blood samples were allowed to coagulate for 40 min before centrifugation (3500× g, 10 min, 4 • C).
After serum collection, seven gilts from the control group and six gilts per group from the remaining treatment groups were killed in compliance with euthanasia principles, and the caudolateral portion of the lateral lobe of the liver was removed, snap-frozen in liquid nitrogen, and stored in a freezer at −80 • C until analysis. Ovary, uterus, and oviduct samples were washed with ice-cooled phosphate-buffered saline and dried with sterile tissue paper, and the uterus weight and length in both directions (left and right) were measured. The numbers of corpora lutea and antral follicles with diameters between 1 and 3 mm and >3 mm were recorded. The cumulus oocyte complex (COC) and granulosa cells (GCs) were obtained in large antral follicles from ovaries on both sides with diameters > 3 mm as previously described [25], snap-frozen, and stored at −80 • C.
For analysis, the right ovary was cut into four sagittal slices that were then fixed in 4% paraformaldehyde (100 mmol/L phosphate buffer, pH 7.4), dehydrated, embedded in paraffin, and sectioned (5 µm thickness). One section in each part of the ovary was stained with hematoxylin and eosin (HE), and then examined under a microscope (Nikon 80i, Tokyo, Japan) for the presence of different classes of follicles, including primordial follicles, primary follicles, secondary follicles, and antral follicles; their numbers were quantified as previously described [25]. The primordial, primary, and secondary follicles with visible nuclei were counted, and those without nuclei were not counted. However, antral follicles with diameters <1 mm, which were not easy to find in ovarian sections, were counted when present in visible oocytes even if they did not have a visible nucleus ( Figure 1) [25].

Sample Collection
Blood samples before the morning meal and at 60 and 120 min after the morning meal were collected on day 18 of the second estrous cycle to measure follicle-stimulating hormone (FSH) and luteinizing hormone (LH) levels. Blood samples were also collected at day 19 of the second estrous cycle at 8:00 a.m. before the morning meal. Blood samples were allowed to coagulate for 40 min before centrifugation (3500× g, 10 min, 4 °C).
After serum collection, seven gilts from the control group and six gilts per group from the remaining treatment groups were killed in compliance with euthanasia principles, and the caudolateral portion of the lateral lobe of the liver was removed, snap-frozen in liquid nitrogen, and stored in a freezer at −80 °C until analysis. Ovary, uterus, and oviduct samples were washed with ice-cooled phosphate-buffered saline and dried with sterile tissue paper, and the uterus weight and length in both directions (left and right) were measured. The numbers of corpora lutea and antral follicles with diameters between 1 and 3 mm and >3 mm were recorded. The cumulus oocyte complex (COC) and granulosa cells (GCs) were obtained in large antral follicles from ovaries on both sides with diameters > 3 mm as previously described [25], snap-frozen, and stored at −80 °C.
For analysis, the right ovary was cut into four sagittal slices that were then fixed in 4% paraformaldehyde (100 mmol/L phosphate buffer, pH 7.4), dehydrated, embedded in paraffin, and sectioned (5 μm thickness). One section in each part of the ovary was stained with hematoxylin and eosin (HE), and then examined under a microscope (Nikon 80i, Tokyo, Japan) for the presence of different classes of follicles, including primordial follicles, primary follicles, secondary follicles, and antral follicles; their numbers were quantified as previously described [25]. The primordial, primary, and secondary follicles with visible nuclei were counted, and those without nuclei were not counted. However, antral follicles with diameters <1 mm, which were not easy to find in ovarian sections, were counted when present in visible oocytes even if they did not have a visible nucleus ( Figure 1) [25].
(a) (b) (c) Figure 1. Follicle classifications: * denotes a primordial follicle and ** denotes a primary follicle in (a); *** denotes a secondary follicle in (b); **** denotes an early antral follicle with diameter <1 mm in (c). For primordial, primary, and secondary follicles, only those with a visible nucleus were counted (); those without a nucleus were not counted (×). For the antral follicles that were not easy to find in ovarian sections, those with visible oocytes were counted () even they did not have a visible nucleus.
The HE-stained sections were placed on millimeter graph paper, and the area was estimated by calculating the number of paper blocks that were covered by the HE sections ( Figure 2). Blocks were not included if the border of the HE section did not occupy at least half of the block. The number of follicles at each stage was normalized by the area of HE-stained ovarian tissue in the sections, and results are presented as the number of follicles per cm 2 [25]. The numbers of follicles in different parts from each ovary were pooled to calculate an average. Follicular fluid from large antral follicles was collected as previously described [26] and stored at −20 °C until analysis. For primordial, primary, and secondary follicles, only those with a visible nucleus were counted (); those without a nucleus were not counted (×). For the antral follicles that were not easy to find in ovarian sections, those with visible oocytes were counted () even they did not have a visible nucleus.
The HE-stained sections were placed on millimeter graph paper, and the area was estimated by calculating the number of paper blocks that were covered by the HE sections ( Figure 2). Blocks were not included if the border of the HE section did not occupy at least half of the block. The number of follicles at each stage was normalized by the area of HE-stained ovarian tissue in the sections, and results are presented as the number of follicles per cm 2 [25]. The numbers of follicles in different parts from each ovary were pooled to calculate an average. Follicular fluid from large antral follicles was collected as previously described [26] and stored at −20 • C until analysis.

Analysis of Hormone Levels in Serum and Follicular Fluid
The concentrations of insulin-like growth factor-1 (IGF-1), 17β-estradiol (E 2 ), and anti-Müllerian hormone (AMH) in serum or follicular fluid were measured using enzyme-linked immunosorbent assay (ELISA) kits purchased from Alpco Diagnostics (Salem, NH, USA), R&D systems (Minneapolis, MN, USA), and CUSABIO Biotech (Wuhan, China), respectively, according to the manufacturer's instructions. The optical density (OD) value was read on an ELISA plate reader, and sample values were calculated from a standard curve according to the manufacturer's instructions. The follicular fluid was diluted at 1:1000 for the E 2 measurement. The detection limits of IGF-1, E 2 , and AMH were 0.09 ng/mL, 1.1 pg/mL, and 1.25 ng/mL, respectively.
Toxins 2022, 14, x FOR PEER REVIEW Figure 2. Determination of area of ovarian hematoxylin and eosin stained (HE) se HE-stained sections were placed on millimeter graph paper, and the area was estimate lating the number of blocks covered by the HE sections. The blocks were not counted if of the HE sections did not occupy at least half of the block, as indicated by the white a erwise, they were counted, as indicated by the black arrow.

Analysis of Hormone Levels in Serum and Follicular Fluid
The concentrations of insulin-like growth factor-1 (IGF-1), 17β-estradiol (E ti-Müllerian hormone (AMH) in serum or follicular fluid were measured zyme-linked immunosorbent assay (ELISA) kits purchased from Alpco Diagn lem, NH, USA), R&D systems (Minneapolis, MN, USA), and CUSABIO Biotech China), respectively, according to the manufacturer's instructions. The optic (OD) value was read on an ELISA plate reader, and sample values were calcula standard curve according to the manufacturer's instructions. The follicular flu luted at 1:1000 for the E2 measurement. The detection limits of IGF-1, E2, and A 0.09 ng/mL, 1.1 pg/mL, and 1.25 ng/mL, respectively.

Analysis of Oxidative Stress Biomarkers in Serum, Follicular Fluid, and Liver Sam
The thawed serum and follicular fluid samples were centrifuged at 2000× g 20 min before the assays were conducted. Liver samples were defrosted and hom on ice with 0.9% NaCl. The homogenate was centrifuged at 4000× g at 4 °C for obtain the supernatant for biochemical analysis. Malondialdehyde (MDA) co enzymatic activities of superoxide dismutase (SOD) were measured in serum liver homogenate, and follicular fluid samples using commercial assay kits Jiancheng Institute, Jiangsu, China). Briefly, MDA was quantified using the thio acid method. The SOD activities of serum, follicular fluid, and liver homogena were measured [27]. Protein concentration of liver tissues was determined [27].

Total RNA Extraction and Quantitative Real-Time PCR (qRT-PCR)
The total RNA of frozen liver samples and COCs was extracted using TR logue no. 15 596-026, Invitrogen, Waltham, MA, USA) according to the manufac structions. The cDNA was then synthesized using a reverse transcription kit ( Shiga, Japan) following the manufacturer's instructions. The qRT-PCR for IG lin-like growth factor acid-labile subunit (ALS), AMH, growth differentiatio (GDF9), bone morphogenetic protein (BMP15), phosphatase and tensin homolo on chromosome ten (PTEN), LIM homeobox gene (LHX8), cytochrome P450 subfamily A member 1 (CYP11A1), steroid acute regulatory protein (StAR), a was performed on a CFX96 Real-Time PCR Detection System (Bio-Rad, Her USA) with a commercial SYBR Green kit (TAKARA). The sequences of the pri as follows: IGF1, forward 5′-ATCTGAGGAGGCTGGAGATGTA-3′ and

Analysis of Oxidative Stress Biomarkers in Serum, Follicular Fluid, and Liver Samples
The thawed serum and follicular fluid samples were centrifuged at 2000× g at 4 • C for 20 min before the assays were conducted. Liver samples were defrosted and homogenized on ice with 0.9% NaCl. The homogenate was centrifuged at 4000× g at 4 • C for 20 min to obtain the supernatant for biochemical analysis. Malondialdehyde (MDA) content and enzymatic activities of superoxide dismutase (SOD) were measured in serum samples, liver homogenate, and follicular fluid samples using commercial assay kits (Nanjing Jiancheng Institute, Jiangsu, China). Briefly, MDA was quantified using the thiobarbituric acid method. The SOD activities of serum, follicular fluid, and liver homogenate samples were measured [27]. Protein concentration of liver tissues was determined [27].

Statistical Analysis
One gilt in the control, two in 25% MC, two in 50% MC, two in 75% MC, three in 100% MC, three in 100% MW, and two in 100% MW + MC did not show estrous until day 260 of age, and their data were excluded, except for their growth performance results. The growth data for days 1-28 of the experiment were analyzed using pen (replicate) as the experimental unit. The individual gilt was the experimental unit for all parameters analyzed from day 29 to the end of the experiment. The current data were checked their homogeneity of variances and normal distribution of the residuals before using parametric analyses. The data were tested for normal distribution using the Kolmogorov-Smirnov test. The data were analyzed using the general linear model procedure of SAS 9.4 (SAS Institute, Inc., Cary, NC, USA) according to the following model: where Y i is the response variable, µ is the overall mean, α i is the fixed effect of dietary treatment, and ε i is the residual error. We used orthogonal linear contrast analysis to test the linear and quadratic effect of MC inclusion levels on parameters. Furthermore, the following statistical model was used to analyze the inclusion of MC or MW on the parameters within the control, 100% MC, 100% MW, and 100% MC + MW groups: where Y ijk is the response variable, µ is the overall mean, α i and β j are the fixed effects of 100% MC inclusion level and 100% MW inclusion level, respectively, αβ ij is the interaction among fixed effects, and e ij is the residual error. The results are expressed as the mean ± pooled standard error of the mean. A p-value < 0.05 was considered to be statistically significant, whereas a p-value < 0.10 was considered to indicate a tendency. Table 3 summarizes the effects of the diets containing various proportions of MC or MW on growth performance of gilts. The BWs at days 1, 28, 56, and 96 of the experiment were not affected by dietary treatment, with only a tendency for 100% MC inclusion to decrease the BW at days 56 (p = 0.067) and 96 (p = 0.099) of experiment. A tendency for the ADFI of gilts to be affected by dietary treatment was detected at days 29-56 of the experiment (p = 0.098). The ADG of gilts at days 1-28 was affected by the inclusion of 100% MW (p < 0.05), and it showed linear (p = 0.073) and quadratic effects (p < 0.05) for MC. The F/G ratio at days 29-56 of the experiment was also affected by dietary treatment. A quadratic effect, but not a linear effect, of MC inclusion level on the F/G was observed at days 29-56 (p < 0.05). The F/G ratio at days 29-56 was also affected by inclusion of 100% MC (p < 0.05) or 100% MW (p = 0.072).

Effects of MC or MW Inclusion on Pubertal Onset of Gilts
Age, BW, and BF at puberty and estrous cycle length were not affected by dietary treatment (p > 0.05), with the exception of a tendency for BF at puberty to be affected by dietary treatments (p = 0.089) ( Table 4). The age at puberty was affected by the quadratic effects of MC inclusion level (p < 0.05).  Table 5 lists the organ weight at slaughter on day 19 of the second estrous cycle. The weight and relative weight of the liver, spleen, and uterus were not affected by dietary treatment (p > 0.05). The ovary weight was affected by dietary treatment (p < 0.05), and a linear effect of MC inclusion level was observed (p < 0.05). The ovary weight of control gilts was greater than those of 100% MC and 100% MW + MC gilts (p < 0.05). The weight (p < 0.01) and relative weight (p = 0.074) of the ovary were decreased by 100% MC but not by 100% MW or the interaction between 100% MC and 100% MW (p > 0.05).

Effects of MC or MW Inclusion on Organ and Ovarian Follicular Development of Gilts
The number of small antral follicles with diameters of 1-3 mm on the ovarian surface was not affected by dietary treatment (p > 0.05). In contrast, the number of large follicles (>3 mm) was significantly affected by dietary treatment (p < 0.01) and decreased linearly with increasing MC inclusion level (p < 0.01). Inclusion of 100% MC or 100% MW resulted in a decrease in large antral follicles; the number decreased from 20.2 in control samples to 7.5 in 100% MW + MC gilts (p < 0.05). The numbers of primordial follicles (p < 0.01), growing follicles (p = 0.073), and total primordial and growing follicles (p < 0.01) per cm 2 of ovarian sections were affected by dietary treatment, and a linear effect of MC inclusion level was detected (p < 0.01). The primordial follicles (p < 0.01), growing follicles (p < 0.05), and total primordial and growing follicles were affected by 100% MC inclusion (p < 0.05) Toxins 2022, 14, 413 9 of 16 but not by 100% MW inclusion or their interaction (p > 0.05). The number of corpora lutea was affected by dietary treatment, with a linear effect of MC inclusion level. The control gilts had 6.6 and 7.1 more corpora lutea than the 100% MC and 100% MW + MC groups, respectively (p < 0.05).

Effects of MC or MW Inclusion on the Oxidative Status of Serum, Liver, and Follicular Fluid
The antioxidant status-related parameters MDA and SOD in serum, liver, and follicular fluid were significantly affected by dietary treatment (p < 0.05 or p < 0.01), with the exception of MDA in follicular fluid (Table 6). MDA contents linearly increased, and SOD contents linearly decreased with increasing MC inclusion level in serum, liver, and follicular fluid. SOD content in liver samples showed a quadratic effect of MC inclusion level (p < 0.05). A higher MDA content and lower SOD content in serum, liver, and follicular fluid were detected in the 100% MC group (p < 0.05 or p = 0.055) but not in the 100% MW or the 100% MC + MW groups (p > 0.05).

Effects of MC or MW Inclusion on the Hormone Levels in Serum and Follicular Fluid
The serum levels of FSH and LH on day 18 of the second estrous cycle were not affected by dietary treatment before feeding (0 min) or at 60 and 120 min after feeding (p > 0.05) ( Table 7). Serum IGF-1 concentration on day 19 of the second estrous cycle was significantly affected by dietary treatment (p < 0.01), and effects of 100% MC, 100% MW, and their interaction were detected (p < 0.01). The control gilts had a higher level of serum IGF-1 than the 75% MC and 100% MW groups (p < 0.05), while 100% MW + MC gilts had a lower serum IGF-1 level than the gilts in the other groups (p < 0.05). Serum E 2 concentration was not affected by dietary treatment (p > 0.05), but serum AMH concentration was significantly affected by dietary treatment, with a linear effect of MC inclusion level (p < 0.01). Serum AMH level was also affected by 100% MC (p < 0.01) but not by 100% MW or the interaction between 100% MC and 100% MW (p > 0.05). The E 2 concentration in follicular fluid decreased linearly with increasing MC inclusion level, and it was affected by 100% MC, 100% MW, and their interaction (p < 0.01). The highest level of follicular fluid E 2 was detected in control gilts, and the lowest level was found in 100% MW + MC gilts. Table 7. Hormone secretions of gilts with different levels of moldy corn or wheat bran. Values are presented as the mean ± SEM. Within a row, values without the same lowercase superscript letters differ significantly (p < 0.05); p 1 , significance of linear and quadratic effects; p 2 , significance of effects of 100% MC, 100% MW, and their interaction. **, p < 0.01; n = 6-7 per group.

Effects of MC or MW Inclusion on the mRNA Expression
The mRNA expression levels of IGF1 and ALS in the liver, GDF9, BMP15, PTEN, and LHX8 in COCs, and StAR, CYP11A1, and AMH in GCs were affected by the dietary treatment (p < 0.05) ( Figure 3A-I). There were linear effects of MC inclusion level on the mRNA expression levels of liver IGF1 and ALS, COC GDF9 and BMP15, and GC StAR, CYP11A1, and AMH (p < 0.05). Comparison of the inclusion of 100% MC or 100% MW in a 2 × 2 set showed that the mRNA expression levels of liver IGF1 and ALS, COC GDF9, BMP15, and PTEN, and GC StAR, CYP11A1, and AMH were affected by 100% MC (p < 0.01 or p < 0.05). The mRNA expression levels of PTEN in COCs and CYP11A1 and StAR in GCs were affected by 100% MW (p < 0.05). The CYP11A1 mRNA level in GCs was affected by the interaction between 100% MC and 100% MW (p < 0.05).

Discussion
Mycotoxins, which are fungal secondary metabolites, cause serious problems for the health of domestic animals. Pigs are known to be particularly impacted by mycotoxins [28]. However, despite the hundreds of existing mycotoxins, only a small number of them are relevant to the swine production industry because they are commonly detected in cereal grains such as corn and wheat, and they are highly resistant to processing. These toxins include AF (produced by Aspergillus flavus) and DON and ZEA (produced by Fusarium graminearum) [10,28]. Growth and feed intake reduction are often the initial and most prominent symptoms observed as a result of mycotoxin consumption. Swamy et al. (2003) reported that feed intake and cumulative weight gain of starter pigs decreased linearly with the inclusion of grains naturally contaminated with Fusarium mycotoxins in the diet. In another pig feeding study, diets containing both AF and DON at levels >60 and 300 μg/kg, respectively, reduced growth and decreased feed intake [29]. Although the high doses of mycotoxins that induced feed refusal of pigs led to the early premise that weight gain reduction is exclusively caused by reduced food consumption, we found that the ADFI at days 1-28 and 29-56 of the experiment was not different among the different dietary groups. However, the ADG for days 29-56 of the experiment was significantly greater in the control gilts than in gilts in the 25% MC, 50% MC, 100% MC, and 100% MC + MW groups. Amuzie and colleagues offered new insight into the mechanisms contributing to growth inhibition by DON in an animal model [30,31]. They showed that DON suppressed growth in mice by reducing growth hormone (GH)

Discussion
Mycotoxins, which are fungal secondary metabolites, cause serious problems for the health of domestic animals. Pigs are known to be particularly impacted by mycotoxins [28]. However, despite the hundreds of existing mycotoxins, only a small number of them are relevant to the swine production industry because they are commonly detected in cereal grains such as corn and wheat, and they are highly resistant to processing. These toxins include AF (produced by Aspergillus flavus) and DON and ZEA (produced by Fusarium graminearum) [10,28]. Growth and feed intake reduction are often the initial and most prominent symptoms observed as a result of mycotoxin consumption. Swamy et al. (2003) reported that feed intake and cumulative weight gain of starter pigs decreased linearly with the inclusion of grains naturally contaminated with Fusarium mycotoxins in the diet. In another pig feeding study, diets containing both AF and DON at levels >60 and 300 µg/kg, respectively, reduced growth and decreased feed intake [29]. Although the high doses of mycotoxins that induced feed refusal of pigs led to the early premise that weight gain reduction is exclusively caused by reduced food consumption, we found that the ADFI at days 1-28 and 29-56 of the experiment was not different among the different dietary groups. However, the ADG for days 29-56 of the experiment was significantly greater in the control gilts than in gilts in the 25% MC, 50% MC, 100% MC, and 100% MC + MW groups. Amuzie and colleagues offered new insight into the mechanisms contributing to growth inhibition by DON in an animal model [30,31]. They showed that DON suppressed growth in mice by reducing growth hormone (GH) signaling through mechanisms mediated by IGF-1 and ALS. Regulation of growth is complex, and the interrelated roles of ALS, IGF-1, insulin-like growth factor-binding protein 3 (IGFBP3), and GH have been thoroughly reviewed elsewhere [32]. In our study, the liver expression levels of IGF-1 and ALS and the serum levels of IGF-1 were downregulated with increasing level of mycotoxins induced by the inclusion of MC or MW, which might explain the observed decreased growth performance in gilts.
However, despite the greater mycotoxin content, the 75% MC and 100% MC diets did not result in a decreased ADG at days 29-56 compared with the control gilts. A possible explanation for this result was that gilts in these groups had a relatively lower ADG (664.6 and 674.3 g/day, respectively) at 1-28 days, but they developed adaptive physiological changes to the mycotoxins that allowed catch-up growth during days 29-56 of the experiment. Similarly, pigs were able to develop adaptation to the effects of DON-contaminated diets during an extended exposure period [11,33]. This might explain why the ADG and F/G ratio were similar between different groups at days 57-96 of the experiment.
Puberty onset is an important signature of female maturation that allows gilts to enter the reproductive cycle. It was reported that prepubescent exposure to dietary ZEA at the level of 1.5-2.0 ppm disturbed the hypothalamic-hypophysial function and stimulated an early attainment of puberty onset in gilts [8][9][10]. In the current study, we did not observe an obvious effect of MC or MW inclusion on the age at puberty, BW at puberty, or BF thickness at puberty, and this observation was consistent with early studies in which prepubescent gilts were fed with 3.61 ppm or 4.33 ppm SEN [34]. The onset of puberty is a quite complex physiological process that requires the coordination of the hypothalamus, pituitary, and ovary, and one of the most important features is the pulsatile release of FSH and LH. In the present study, we measured the serum concentrations of FSH and LH at 0, 60, and 120 min after feeding on day 18 of the second estrous cycle. We did not find any difference among groups, which suggested that mycotoxins did not directly influence the hypothalamus.
After sacrificing the gilts on day 19 of the second estrous cycle, we found that liver, kidney, spleen, and uterus development was similar among groups. Recent studies of growing pigs that were fed different levels of DON [13] and growing gilts fed with different levels of ZEA [21] also found no effects on liver development. However, the organ health (such as liver) of pigs was impaired by mycotoxins [11,13,21,22]. In the present study, we measured the antioxidant capacity in the serum, liver, and follicular fluid of gilts and found that inclusion of either MC or MW resulted in a significant elevation of MDA and suppressed secretions of SOD, suggesting that oxidative stress was induced by a high level of mycotoxins. Oxidative stress is a phenomenon that occurs in a cell when the concentration of reactive oxygen species exceeds the antioxidant capacity, and it usually impairs the reproductive performance of sows [23]. Luderer et al. suggested that oxidative stress was involved in reproductive toxicity caused by various stimuli such as a dietary toxicity. For example, T-2 toxin induced apoptosis in rat GCs through oxidative stress [35]. Researchers reported that the early oocyte environment in the follicular fluid was rich in antioxidant enzymes, which helped protect oocytes from oxidative damage [36,37]. However, the concentration of MDA and activity of SOD in follicular fluid did not differ significantly among the treatment groups in our study. However, the SOD activity in follicular fluid decreased linearly with increasing inclusion of MC or MW. SOD is an important intracellular antioxidant enzyme that detoxifies superoxide anion and provides significant protection against the oxidative effect of mycotoxin-contaminated maize [38]. In the present study, the increased oxidative stress would have consequently resulted in negative effects on ovarian development.
We observed a dose-dependent decline in large antral ovarian follicles with increasing level of mycotoxins in the diet. The number of large antral follicles decreased from 20.2 in the control group to 7.5 in the 100% MW + MC group, but the number of small antral follicles with diameter 1-3 mm remained unchanged. The decrease in large antral follicles might be due to the failure of small antral follicles to be recruited by FSH and LH. Furthermore, we found that the numbers of primary follicles and growing follicles were significantly decreased by the inclusion of MC or MW. The primary follicles and growing follicles serve as the main source of the oocyte pool, and the observed decline in the ovarian pool suggested low reproductive potential in the long-term. Recently, Zhuo et al. (2019) reported that increased follicle death during the early period of life would decrease lifetime fertility in mice. AMH is a glycoprotein produced by the GCs in growing ovarian follicles, and it serves as a biomarker of ovarian reservation [39,40]. The AMH level in gilts can also be a predictor of their subsequent fertility as sows [41]. In line with the greater size of the ovarian pool, the serum levels of AMH decreased with increasing mycotoxin levels in the diet, suggesting that greater mycotoxin content in the diet impaired ovarian function.
Ovarian follicle health and survival are regulated by signaling pathways that govern primordial follicle recruitment and growth. Reddy et al. (2008) showed that PTEN, a major negative regulator of phosphatidylinositol 3-kinase, is a critical regulator controlling oocyte growth and survival. In the present study, the mRNA expression levels of PTEN were elevated with increasing mycotoxin content in the diet, and there is evidence that mycotoxins such as FUM B1 [42], OA [43], and DON [44] can regulate PTEN signaling. Results of these studies suggest that mycotoxins might exert their negative effects on follicle survival via PTEN signaling.
We found that mycotoxins not only reduced the number of oocytes in the ovary, but also had negative effects on the oocyte developmental potential. The expression levels of BMP15 and GDF9, two well-known markers of oocyte developmental competence in mammals [45], were significantly downregulated by the inclusion of mycotoxins in the diet. Consistent with this, ZEA exposure in vitro impaired organelle function during porcine oocyte meiotic maturation [17]. Secretion of E 2 by GCs is also important for porcine oocyte developmental competence [46]. In the present study, the concentrations of E 2 in the follicular fluid, as well as expression of CYP11A1 and StAR, two genes encoding enzymes responsible for the de novo synthesis of E 2 , were decreased by the inclusion of MC or MW in the diet. This result suggested that mycotoxins might affect E 2 signaling, thereby impairing porcine oocyte quality.
The current study did not provide solutions to alleviate the negative effect of mycotoxins on the ovarian follicular development. However, dietary baicalin [47][48][49] or berberine [50] supplementation could alleviate the DON challenge-induced gut impairment by decreasing oxidative stress and immune stress in pigs. It remains uncertain whether the negative effect of mycotoxins on the ovarian follicular development could be attenuated by an oxidative stress modulator.

Conclusions
In conclusion, we found that chronic exposure to diets containing mycotoxins by inclusion of MC or MW reduced the ovarian follicle pool size and oocyte quality, and these changes were associated with increased oxidative stress and suppressed liver IGF-1 and ALS secretion. According to hygienic standards in China and EU, mycotoxin levels (DON, ZEA, and AFB1) in the diet of the 75% MC group were below the maximum levels allowed for porcine diets, but this diet resulted in a significant decrease in the ovarian follicle pool and oocyte quality. Our results suggest that mycotoxin levels allowed for the production of replacement gilts should be lower than the current standards. Funding: This work was supported the Major Scientific and Technological Special Project of Sichuan Province (No. 2021ZDZX0009). The funding sources played no role in study design or the collection, analysis, and interpretation of data, writing the report, or in the decision to submit the paper for publication.

Institutional Review Board Statement:
The present trial was conducted at research center of Sichuan Agricultural University. All experimental procedures were approved on 2 Mart 2015 by the Animal Care and Use Committee of Sichuan Agricultural University (Approval No. 20150122), and followed national laws and National Research Council (NRC) guidelines for the care and use of laboratory animals.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.