Animal experiment and sample collection

Twenty-four gilts (Landrace × Large White) were used in this study. These gilts were mated, housed individually. They had free access to drinking water and were fed a corn-and soybean meal-based diet (2 kg.d^-1) formulated to meet or exceed nutrient requirements according to the National Research Council (2012). At d 60, 90 and 110 of gestation, 8 gilts were randomly selected and exsanguinated after induction of anesthesia with halothane [10]. The maternal abdomen was opened and the pregnant uterus was extracted immediately from the body cavity. IUGR fetuses were identified as those whose weights were less than two standard deviation of the mean body weight for gestational age. One NBW fetus and one IUGR fetus, as well as their corresponding placenta and endometrium, were collected from each one of gilts. Placenta was easily separated from endometrium on d 60, 90 and 110 of gestation, and we did not observe any contamination between these two tissues. The tissues were rapidly transferred to liquid nitrogen and stored at -80°C.

Extraction of proteins from placental and endometrial samples

The placenta and endometrium from 18 IUGR and 18 NBW (6/stage) were used to extract protein, as we described previously [3]. Briefly, approximately 0.2 g frozen samples were crushed to powder in liquid nitrogen, then homogenized in a lysis buffer containing 7 M urea, 2 M thiourea, 4% 3-[3-(-cholamidopropyl)-dimethylammonio]-1-propanesulfonate, and 50 mM dithiothreitol with protease inhibitors (GE Healthcare, Piscataway, NJ). An ultrasonicater (Sonics Model VC 750, Sonics and Materials, Newtown, CT) was set at 20% power output and used to break down the mixture for 10 min at 0°C. After the addition of 1% (vol/vol) nuclease mix (GE Healthcare), the mixture solution was kept at room temperature for 1 h to completely solubilize proteins, followed by re-sonification for 10 min as described above to thoroughly break down cell membranes. The homogenate was centrifuged for 10 min at 13,000 g at 4°C to settle down the insoluble components. The supernatant fluid was obtained and its protein concentration was determined using the Brandford method. Portions of the homogenate (1 mg of protein for both placenta and endometrium) were stored at -80°C.

Comparative Proteome Analysis

Two-dimensional gel electrophoresis was run in triplicate for both placental and endometrial samples at each of 3 time-points (D60, 90, 110) to compare protein expression between IUGR and NBW fetuses. First, isoelectric focusing was performed using immobile DryStrip gels (pH 3 to 10, nonlinear, 24 cm long; GE Healthcare), as we described previously [11]. One mg protein was loaded onto immobile Drystrip gels strips using an in-gel sample rehydration technique. The first-dimensional isoelectric focusing was carried out at 20°C for 100,000 voltage h (IPGphor system, GE Healthcare), followed by equilibration in 4 mL of equilibration buffer 1 (6 M urea, 1% dithiothreitol, 30% glycerol, and 50 mM Tris-Cl pH 8.8) for 15 min on a shaker and then in 4 mL of equilibration buffer 2 (6 M urea, 2.5% iodoacetamide, 30% glycerol, 50 mM Tris-Cl pH 8.8) for 15 min as described above. The second-dimensional Gel was performed using 12.5% SDS-PAGE gels. The gels were run at 30 mA/gel for 30 min and then at 50 mA/gel until the Bromophenol Blue came out of the gels. Analytical gels were stained with Coomassie Brilliant Blue G-250 (Amresco, Solon, OH), and high-resolution gel images (400 dpi) were obtained using a scanner (ImageScanner, model PowerLook 2100XL, UMAX Technologies, Atlanta, GA). Images were analyzed using commercial software (Image-Master 2D Platinum Version 6.01, GE Healthcare) and differentially expressed protein spots that deviated more than 1.3-fold in their relative volume (% vol) were selected for in-gel digestion and protein identification.

In-Gel Digestion and MALDI-TOF/TOF MS Analysis

Protein spots of interest were cut from the gels to be destained with 100 μL of 50% (v/v) acetonitrile (ACN) in 25 mM ammonium bicarbonate for 1 h. Then an aliquot (3 μL) of trypsin solution (10 μg/mL; Amresco Inc., Solon, OH) in 25 mM ammonium bicarbonate was added to each gel piece after it was completely dried by vacuum centrifugation (Eppendorf Concentrator 5301, Eppendorf, Hamburg, Germany) for 30 min. The mixture was incubated in a trypsin solution at 4°C for rehydration for 1 h, followed by incubation at 37°C for 12 h. Thereafter, the gel pieces were vacuum-dried to evaporate the solvent and 8 μL of 5% trifluoroacetic acid (TFA) was added onto the dry gel pieces. After incubation at 37°C for 1 h, the solution was transferred into a microcentrifuge tube. The gel pieces were extracted twice separately with 8 μL of 2.5% TFA and 50% acetonitrile, and 5% TFA. The third extraction was performed using 8 μL of 100% acetonitrile and the solution was combined with the previous two extracts in the microcentrifuge tube. Finally, the combined solution was dried and resolubilized in 2 μL of 0.5% TFA for protein identification using MALDI-TOF/TOF MS (Matrix Assisted Laser De sorption Ionization-Time of Flight/Time of Flight MS).

Protein Identification

The peptide mass fingerprint analysis was performed using a search engine (MASCOT, Matrix Science, Lon- don, UK) [12]. Searching parameters included: (1) “trypsin” as the enzyme of protein digestion; (2) “monoisotopic” as mass value; (3) “unrestricted” as peptide mass; (4) “0.3 Da” as peptide mass tolerance; (5) “oxidation (M) and carbamidomethyl (C)” as variable modifications; and (6) “1” as maximum missed cleavages. Protein match with a score was considered significant (P < 0.05).

Western Blotting

Extracted protein samples were boiled for 5 min and separated by electrophoresis (Bio-Rad, Richmond, CA) in 12% SDS-PAGE gel before electroblotted (Bio-Rad) onto a polyvinylidene fluoride membrane (Millipore, Billerica, MA) [11]. After polyvinylidene fluoride membranes were blotted with Tris buffer containing 5% fat-free dry milk and 0.05% Tween-20 (TBST; 0.05% Tween 20, 100 mmol/L of Tris-HCl, and 150 mmol/L of NaCl, pH 7.5) for 1 h at 25°C, they were rinsed in TBST for 4 times and incubated overnight at 4°C with primary antibodies, which were mouse polyclonal GAPDH (Abcom, ab22556), rabbit polyclonal HIF-1α (Abcam, ab114977), mouse monoclonal VEGF (Abcam, ab1316) and rabbit polyclonal VEGFR1 (Abcam, ab2350) with 1:3000, 1:2000, 1:2000 and 1:2000 dilutions, respectively. The membranes were washed in the same manner as described above and incubated with the horseradish peroxidase-conjugated secondary antibody (1:5,000; Sigma) for 40 min. Then, the membranes were washed 6 times for 5 min in TBST and the blots were detected using Lumi-light Western Blotting substrates (Biofuture, Beijing, China) and analyzed with the Imaging Analysis Software of National Institutes of Health (Bethesda, MD).

Statistical analyses

Data are expressed as means ± SEM. Differences in fetal weights or western blot results in placenta or endometrium between IUGR and normal fetuses at a given day of gestation were statistically analyzed by the t-test (version 8.2; SAS Institute, Cary, NC, USA). Proteomic data for placenta or endometrium were analyzed by the t-test, a statistic program embedded in the commercial software (Image-Master 2D Platinum Version 6.01, GE Healthcare), to compare the difference between IUGR and normal fetuses at a given day of gestation. To avoid systemic error, we further calculated the ratio of protein levels between IUGR fetuses to those for normal weight fetuses, and only the proteins with a ratio higher than 1.3 or lower than -1.3 were chosen for further identification. P < 0.05 was taken to indicate statistical significance and data are expressed as means ± SEM.

Body weight of fetuses

Both IUGR and NBW fetuses grew between D60 and D110 of gestation (Table 1). Body weights of selected IUGR fetuses were 93.4, 460, and 858 g respectively and were significantly lower than that of NBW at the same gestational day, which were 143.3, 914.3, and 1528 g respectively (P < 0.01, Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Body weights of normal and IUGR fetuses at D 60, 90 and 110 of gestation.

https://doi.org/10.1371/journal.pone.0142396.t001

Comparative proteome analysis of placentae between IUGR and NBW fetuses

A total of 50 placental protein spots were differentially expressed in placentae between IUGR and NBW fetuses at D60, 90 and 110 of gestation. Their appearances in gel images were labeled in Fig 1A and biochemical information is summarized in Table 2 and Fig 1B. According to their biological function, these proteins were classified into the following groups: (1) energy metabolism; (2) nutrient transport; (3) stress response; (4) nutrient metabolism; (5) cell proliferation and apoptosis; (6) cell morphology and motility.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Abundance and classification of differentially expressed proteins between normal and IUGR placenta.

(A) Abundance of differentially expressed proteins in the placenta of normal and IUGR pig fetuses at D 60, 90 and 110 during gestation. (B) Functional classification of differentially expressed proteins between placentae from normal and IUGR pig fetuses.

https://doi.org/10.1371/journal.pone.0142396.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Biochemical information about differentially expressed protein in the placenta of IUGR fetuses during mid and late gestation.

https://doi.org/10.1371/journal.pone.0142396.t002

Comparative proteome analysis of endometria between normal and IUGR fetuses

A total of 52 protein spots were differentially expressed in the IUGR endometrium compared with NBW fetuses at D 60, 90 and 110 of gestation. Their appearances in gel images were labeled in Fig 2A and biochemical information is summarized in Table 3 and Fig 2B. According to the biological function, these proteins were involved in: (1) energy metabolism; (2) transportation; (3) stress response; (4) nutrient metabolism; (5) cell proliferation and apoptosis; (6) cell morphology and motility.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Abundance and classification of differentially expressed proteins between normal and IUGR endometrium.

(A) Abundance of differentially expressed proteins in the endometrium of normal and IUGR pig fetuses at D 60, 90 and 110 of gestation. (B) Functional classification of differentially expressed proteins between endometrium from normal and IUGR pig fetuses.

https://doi.org/10.1371/journal.pone.0142396.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Biochemical information about differentially expressed protein in the endometrium of IUGR fetuses during gestation.

https://doi.org/10.1371/journal.pone.0142396.t003

Western blotting Analysis

Fig 3 shows the western blotting analysis of GADPH, HIF-1α, VEGF and VREFR1. The expression of HIF-1α was higher in the IUGR placenta, compared with the NBW group. In contrast, GADPH, VEGF and VEGFR were reduced significantly (p < 0.01) in the IUGR placenta, as compared with the NBW fetuses.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Expression of GAPDH, HIF-α, VEGF and VEGFR1 in placenta of normal and IUGR fetuses.

(A) Western blot analysis of GAPDH. (B) Western blot analysis of HIF-α. (C) Western blot analysis of VEGF. (D) Western blot analysis of VEGFR1.

https://doi.org/10.1371/journal.pone.0142396.g003

Energy metabolism

Energy is required for nutrient and gas transport between mother and fetus in the uteroplacental unit, as well as for the synthesis of hormones and secretion of molecules within trophoblasts [14]. An interesting finding of this study is that the expression of several key enzymes involved in energy metabolism was reduced in IUGR fetuses. TPI1, which catalyzes the reversible inter-conversion of the triose phosphate isomers dihydroxyacetone phosphate and D-glyceraldehyde 3-phosphate, plays an important role in glycolysis, and is essential for efficient ATP production. The reduced expression of TPI1 may provide less energy produced through glycolysis in the IUGR placenta, which is consistent with a 28% decrease in lactate production by the IUGR placenta [15]. In order to further confirm the impairment of glycolysis in the IUGR placenta, we analyzed expression GAPDH, a key regulatory protein, using the western blot technique, and the result was consistent with our proteomic analysis. DLD is a component of the pyruvate dehydrogenase complex and responsible for converting pyruvate into acetyl-CoA, which is then fed into the citric acid cycle and contributes to ATP production. IDH1 and ACO1 are also important enzymes participating in the citric acid cycle, an important metabolic pathway to produce ATP. The decreased expression of TPI, DLD and IDH in the IUGE placenta suggests less ATP production through the citric acid cycle, thereby compromising energy-dependent nutrient transport from mother to fetus. In addition, BCK and V-ATPases, both of which are related to ATP production, were reduced in the placenta and endometrium of IUGR fetuses, further confirming the alteration of energy metabolism in uteroplacental tissues of IUGR pig fetuses.

Stress response

The uteroplacental tissues appear particularly vulnerable to oxidative stress due to extensive cell proliferation and high metabolic activity during pregnancy [16] and [17]. PARK7, which acts as a redox-sensitive chaperone and apparently protects cells against oxidative stress and apoptosis, is a sensor for oxidative stress. In this study, significant up-regulation of PARK7 expression provides evidence for the elevated levels of oxidative stress in the IUGR endometrium. Besides, abundances of several other proteins (PDIA3I1, PDIA3, TXN and PRDX6) related to cellular redox homeostasis were increased in the IUGR fetuses. PDIA3I1 and PDIA3 belong to the disulfide isomerase (PDI) family of proteins, which catalyze the reduction of mispaired thiol residues. Similarly, TXN1 (as an antioxidant by exchanging thiol-disulfide) can reduce placental oxidative stress and enhance fetal growth [18]. Furthermore, PRDX6 can protect the cell against oxidative stress or apoptosis by eliminating hydrogen peroxide (H₂O₂) and neutralizing other reactive oxygen species (ROS) [19]. The increased expression of PDIA3I, PDIA3, TXN1 and PRDX6 may suggest an adaptive protection against oxidative stress in IUGR utreoplacental tissues to compensate for fetal growth.

Hsps are molecular chaperones that control protein folding and protect cells and tissues from being damaged in adverse environmental conditions. Previous studies have indicated that Hsp70s and Hsp90s are strongly up-regulated by heat stress, oxidative stress, and toxic chemicals [20] and [21]. At the same time, expression of heat shock protein 70 was higher in placental tissues associated with increased apoptosis and decreased numbers of mitotic cells in cases of miscarriage [22]. In this study, expression of 5 members of heat shock 70 proteins family (GRP78, GRP75, Heat shock 71 kDa protein, HSPA8 and GRP94), HSP90A and CRYAB (a small heat shock protein) was all increased in the IUGR utero-placenta unit compared with the NBW counterpart, implying that uteroplacental tissues are under more intensive adverse stresses in IUGR fetuses, which may result in their poor development and organ dysfunction. Hypoxia is a major stress factor for the placenta during pregnancy. Therefore, we adopted the western blotting technique to confirm the increased expression of HIF-1α indicated from the proteomics analysis. Our results further support an augmented level of the HIF-1α protein in the IUGR placenta, as compared with the NBW group.

C3 plays a central role in activation of the complement system [23], contributing to innate immunity, and its deficiency can increase susceptibility to bacterial infection [24] and [25]. C3 is synthesized and secreted in the endometrium [26] and is expressed at a reduced level in patients with severe endometriosis [27]. Results of our study provide the first evidence that decreased abundance of C3 is associated with the IUGR endometrium, suggesting weaker immunity against infection in IUGR fetuses. This corroborates less abundance of IgG in the IUGR circulation.

Transport of nutrients

Efficiency of substrate transport in uteroplacental tissues is crucial for fetal intrauterine growth by supplying nutrients from mother to fetus and by eliminating fetal metabolic wastes [28]. Previous studies have shown that the expression and activity of transport systems for glucose, amino acids and fatty acids were altered in the IUGR placenta [29] and [30]. In this study, we did not find any differences in expression of transporters for glucose, amino acids and fatty acids in the IUGR uteroplacental tissue, probably because their abundances were too low to be detected in the 2D gel image. However, we did identify that TF, which is responsible for iron binding and transport, was down regulated in the IUGR fetus. Maternal iron deficiency during pregnancy directly impairs embryonic growth and development [31], and strongly increases the risk for low birth weight and preterm delivery, as well as abnormal development of affected offspring [32]. Similarly, we previously reported a continuous reduction in the expression of TF in the jejunal mucosa of postnatal IUGR piglets between D 1 and D 21 after birth [33], which can now be explained by less iron transfer from mother to fetus during pregnancy.

Another important finding of this study is that abundances of proteins related to substrate transport, including ALB, AFP and APOA1, were down regulated in IUGR fetuses. ALB acts as a carrier of various substances, including lipid soluble molecules and certain minerals (e.g., steroid hormones, bile salts, unconjugated bilirubin, free fatty acids, calcium, and iron), in the blood circulation. ALB present in placental and endometrial tissues is derived from fetal and maternal circulation, respectively. AFP, produced by the yolk sac and the liver during fetal development, is considered to be the fetal form of ALB. The decreased levels of ALB and AFB in uteroplacental tissues may reduce substance exchanges between maternal and fetal circulation, contributing to the impairment of nutrient uptake and waste elimination in IUGR fetuses. APOA1, a major component of high-density lipoproteins, mainly participates in reverse transport of cholesterol from tissues back to the liver for excretion and plays an important role in cholesterol efflux from tissues. The decreased expression of APOA1 probably can result in higher levels of cholesterol in the IUGR placenta, which can result in more extensive stress in IUGR fetuses.

Nutrient metabolism

Blood flow is a major determinant of uteoplacental function and fetal growth. Angiogenesis in the placental vasculature during pregnancy is necessary for increasing uteroplacental blood flow and, therefore, the supply of nutrients from maternal to fetal circulation [34]. Several previous studies reported that spermine plays an important role in angiogenesis, early mammalian embryogenesis, placental trophoblast growth, and embryonic development in the uterus [35] and [36]. In the current study, we found that spermine synthase was down regulated in the IUGR placenta, which could reduce the production of spermine and then placental vascular development. Consistent with this notion, results of the western blot analysis indicated that the expression of VEGF and VEGFR1 was decreased in the IUGR placenta, further contributing to compromised angiogenesis in the IUGR placenta.

The altered expression of several proteins related to amino acid metabolism in the IUGR endometrium could further aid in explanation of our previous findings of decreased concentrations of amino acids in the IUGR umbilical cord [13]. ASS1 is a key enzyme that catalyzes the synthesis of argininosuccinic acid (the immediate precursor of arginine) from citrulline and aspartic acid, and the loss of this enzyme is considered as a biomarker for susceptibility to arginine depletion [37]. Arginine is now known to be a nutritionally essential amino acid for mammalian fetal growth. RNPEP selectively catalyzes the cleavage of arginine and lysine from amino terminus of protein or peptide. The depressed expression of ASS1 and RNPEP may explain, in part, lower concentrations of arginine in the umbilical vein of IUGR fetuses. In addition, the expression of phosphoserine phosphatase, a key enzyme to produce serine, was up-regulated in the IUGR endometrium and could result in higher levels of serine in the fetal circulation, which is consistent with our previous results [13].

Proliferation and apoptosis

To fulfill the metabolic demands of the growing fetuses, placenta and endometrium undergo dramatic morphological and histological changes during pregnancy, including active cell proliferation and extensive cell differentiation [38]. Stresses induced by maternal nutrient restriction, hypoxia and bacterial infection could adversely affect the growth and morphology of the placenta, causing reduction in the body weight of the fetus [39]. In this study, we find that expression of several key regulators for cell proliferation, including EIF5A2, EIF6, AspRs, and ZFP 420 [40, 41], were decreased in the IUGR placenta, and eEF1A-1, HNRNPK, and LAMN were decreased in the IUGR endometrium. The reduced expression of these proliferation-related proteins suggest slower rates of cell proliferation, which cannot meet the requirement for rapid placental development and lead to a smaller size of the placenta in IUGR fetuses.

Programmed cell death is a fundamental process for maintaining homeostasis in multi-cellular organisms and involves highly regulated processes, including pro-apoptotic and anti-apoptotic signal pathways. CTSZ is a lysosomal cysteine protease to hydrolyze proteins and have a vital role in mammalian cellular apoptosis, while CTSB is thought to play a key role in protecting cells against leaking of proteases from lysosomes during apoptosis. Enhanced expression of proteinase CTSZ, may imply higher apoptosis levels in the IUGR placenta, whereas enhanced expression of protease inhibitor CSTB may be related to protective reactions against apoptosis. In the endometrium, we found that 3 members of serpins (ACT2, ACT3 and serpin A3-6), all of which are capable of inhibiting proteases to control proteolytic cascades and mediate apoptosis processes as caspase inhibitors by preventing apoptosis in mammalian cells [42, 43], were increased in the IUGR group, implying more intensive apoptosis stress in the IUGR placenta.

Cell morphology and motility

Fourteen proteins related to cell morphology and twelve proteins related to cell motility were differentially expressed in placentae and endometria between IUGR and NBW fetuses. Actin is a globular multi-functional protein that forms microfilaments of cytoskeleton and is essential for maintenance of cell shape and mobility. Several isoforms of actin were altered in IUGR, including ACTB (down-regulated in the IUGR placenta and endometrium); ACTN4, Macf1 (down-regulated in the IUGR endometrium); and ACTG (up-regulated in the IUGR endometrium). Besides, EZR, MRLC2, MLC-2A, MSN, GEL, and CCT2, which interact with actin to regulate cellular morphology, were differentially expressed between IUGR and NBW placentae. Similar results were obtained for ARP3 in the endometrium. In addition, both TUBA1B and TUBA1C, which are cytoskeleton microtubule elements, were differentially expressed in the placenta and endometrium. Furthermore, KRT7 and KRT8 were up regulated in the IUGR placenta, whereas CK19 was increased at D90, but deceased at D110 of gestation. Altered expression of these proteins related to actin, tublin and keratin could result in an abnormal change of cytoskeleton in the placenta and endometrium of IUGR fetuses.

Annexins provide a membrane scaffold and link to fibrinolysis, coagulation, inflammation and apoptosis. Reductions in two members of annexins (ANX4 and ANX8) may result in placental structural abnormality and dysfunction in IUGR fetuses. This novel finding provides another line of evidence for impaired utero-placental development in the IUGR conceptus.