The experimental protocol for the study was approved by the Volcani Center Animal Care Committee (approval number IL 797/18), and it was performed in accordance with the relevant guidelines and regulations. The experiment was conducted at the Volcani Center experimental farm in Rishon Lezion, Israel. A detailed description of this experiment was provided in [30]. Briefly, 42 multiparous Holstein dairy cows, with a mean parity of 3.8 ± 1.4 (mean ± standard deviation), participated in the study during the winter season (November 2018–March 2019). The average temperature-humidity index (THI) was 61 ± 8, indicating conditions well within the thermoneutral zone for dairy cows [36]. The cows were group-housed from d 257 of pregnancy to d 60 postpartum in a shaded loose pen that was equipped with a real-time electronic individual feeding system. Each feeding station included an individual identification system (ID tag; SAE Afikim, Kibbutz Afikim, IL) that allowed each cow to enter a specific feeding station only and automatically recorded each meal. The cows were stratified according to milk yields during the first 60 d of the previous lactation, body weight at drying off, and parity. The prepartum dietary treatments began at d 257 of pregnancy as follows: (i) control group (CTL; n = 14), fed a basal dry cow diet supplemented with encapsulated saturated fatty acid (SFA) at 240 g/d/cow; (ii) FLX (n = 14), fed a basal diet supplemented at 300 g/d/cow, with encapsulated flaxseed oil providing ALA at 56.1 g/d; and (iii) FO (n = 14), fed a basal diet supplemented at 300 g/d/cow with encapsulated fish oil providing EPA at 5.8 g/d and DHA at 4.3 g/d. The total fat content of the CTL supplement was 99%, compared to 80% in the FLX and FO supplements; therefore, the amounts of the supplements differed between the groups in order to maintain similar fat contents in all diets. The fat supplements were specially prepared and supplied by SILA (Venice, Italy). The ingredients and the chemical compositions of the rations, as well as the profile of the main FAs in the supplements were previously presented in [30, 37]. The neonate calves were separated from their mothers at calving, and were provided individually with their mothers’ colostrum immediately after blood samples were collected post-calving. A diagram describing the experimental design, the treatments, biological samples collected, and laboratory analyses are presented (Additional file 1: Fig. S1).

Placental tissues were collected immediately after the expulsion of the placenta postpartum. From each placenta, we dissected 4 or 5 samples (50 mg) of cotyledons that were stored immediately at −80 ºC for further RNA and protein extraction. In total, we collected 13 placenta samples during the experiment (3 CTL, 5 FLX, and 5 FO). Since we did not obtain sufficient placenta samples from the control cows, we collected 2 additional placenta samples from cows that were fed the basal dry cow diet (without SFA supplementation) during the winter season of the following year.

We analyzed the gene expression in 5 placenta samples from each dietary treatment. Total RNA extraction of placental tissue was performed using an Animal tissue RNA purification kit (#25700, NORGEN BioTek Corp, Ontario, Canada). The RNA purity was assessed using a NanoDrop One Microvolume UV-Vis Spectrophotometer (Thermo Scientific, Shoham, IL) with a 260/280 ratio of above 1.85. First-strand cDNA was generated using the RevertAid First Strand cDNA Synthesis Kit (#K1622, Thermo Fisher Scientific, Vilnius, Lithuania). Quantitative detection of specific mRNA transcripts was carried out by real-time PCR using a CFX Duet Real-Time PCR System (Bio-Rad Laboratories, Inc., Rishon LeZion, IL) with the SYBR Green PowerTrack™ Master Mix (#A46109, Applied BioSystems, MA, USA) and analyzed using Bio-Rad CFX Maestro Software. In placenta tissues we examined the transcription levels of genes related to lipid metabolism, inflammation, oxidative response, and ECS. The list of primers is presented in Table 1. The reference genes β-Actin (ACTB), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and ubiquitously expressed prefoldin like chaperone (UTX) were examined. NormFinder software suggested ACTB and UTX as the most stable housekeeping genes for the placenta samples. The relative quantity of each gene was normalized to the average transcription levels of the reference genes (2^−∆∆Ct method) according to Livak et al. [38].

The placental samples (15 mg) were ground in a BeadBug homogenizer (D1030-E, Benchmark Scientific, Sayreville, NJ, USA) with two 0.5-mm glass beads (#11079105, BioSpec, Bartlesville, OK, USA) and 1 mL lysis buffer [5% (w/v) sodium dodecyl sulfate (SDS, #L3771-100) in 100 mmol/L Tris-HCl buffer containing 1% (v/v) phenylmethylsulfonyl fluoride (PMSF, #P7626), 1% phosphatase inhibitor (#P5726), and 1% protease inhibitor (#P8340), all from Sigma-Aldrich, St. Louis, MO, USA]. After centrifugation at 20,000 × g for 15 min at 4 °C, the soluble fractions were collected and the protein concentrations were measured using a bicinchoninic acid (BCA) standard assay (9470BCAstand, Cyanogen, Bologna, Italy), then snap-frozen and stored at −80 °C.

For proteomic analysis, we analyzed 3 placenta samples from each dietary treatment. From the controls, we selected the 3 samples that were available from the original experimental cohort, and in FLX and FO we randomly selected 3 samples from each treatment; overall, 9 placenta samples were subjected to mass spectrometry-based proteomic analysis. The samples were lysed with 5% SDS and digested with trypsin (#V528B, Promega, Madison, USA) using the S-trap method [39]. Samples were stored at −20 °C until further use.

The FABPs link lipid metabolism and the ECS, since they serve as transporters of endocannabinoids within cells [42]; therefore, we aimed to determine whether bovine heart-type fatty acid binding protein (FABP3), epidermal-type fatty acid-binding protein (FABP5), and brain-type fatty acid binding protein (FABP7) bind to endocannabinoids by in silico modeling. To this end, FABP3 (Accession number: P10790), FABP5 (Accession number: P55052), and FABP7 (Accession number: Q09139) sequences were collected from the UNIPROT database using the Bos taurus model for modeling the protein structures. The endocannabinoids 2-AG, AEA and docosahexaenoyl ethanolamide (DHEA) were docked to target proteins using GOLD 3.0.1 software, a genetic algorithm that uses a strategy that covers three genetic operators such as migrations, mutations, and crossovers [43]. The compounds that docked into the active site of the target proteins were thoroughly studied by molecular mechanics calculations. The most energetically favorable conformation of each compound was identified and selected after docking. Each compound’s individual binding poses were studied, and interactions with the protein were calculated. A detailed description of the FABPs’ structure and docking modeling is shown in Additional file 3.

Blood samples were collected from calves (n = 9 per treatment) immediately after calving, before colostrum offering. They were collected from the jugular vein into vacuum tubes containing lithium heparin (#BD367526, Becton Dickinson Systems, Cowley, UK). Plasma was separated following centrifugation at 1,500 × g for 20 min at 4 °C, and stored at −80 °C, pending analysis. The FA composition in the placenta and in the plasma of neonate calves was determined as described previously [44]. Briefly, the samples were saponified in a mixture of 60% potassium hydroxide (KOH; #UN1813, Merck, Darmstadt, DE) and ethanol (#UN1170, Bio-Lab Ltd., Jerusalem, IL), extracted with petroleum ether (#UN1268, Bio-Lab Ltd.), and methylated with 5% (v/v) sulfuric acid (#UN1830, Bio-Lab Ltd.) in methanol (#UN1230, Bio-Lab Ltd.). FA methyl esters were analyzed using a 7890N gas chromatograph (Agilent Technologies, Santa Clara, CA) equipped with a DB-23 capillary column (60 m × 0.25 mm × 0.25 μm; Agilent Technologies) and a flame ionization detector. The initial temperature of the column was set at 130 °C, which was increased by 6.5 °C/min to 170 °C, and then by 2.75 °C/min to 215 °C, and held at 215 °C for 18 min. Then, the temperature was increased to 230 °C at 40 °C/min for the remainder of the analysis. The carrier gas was hydrogen, flowing at a linear velocity of 1.6 mL/min; the injection volume was 1 µL.

Plasma interleukin 2 (IL-2) and interleukin 6 (IL-6) concentrations were determined using Bovine Duoset ELISA kits (#DY2465 and #DY8190, respectively; R&D Systems, Inc., Minneapolis, MN, USA). Plasma haptoglobin (HP) concentrations were examined using a haptoglobin bovine ELISA kit (#E-10HPT, ICL, Portland, OR, USA).

The placenta gene expression levels and the plasma variables from calves (IL-2, IL-6, and HP), as well as the FA composition from placenta or plasma were analyzed using a generalized linear model (GLM), using the following model: Y_ijk = µ + T_i + C(T)_ij + E_ijk, where Y_ijk = dependent variable, µ = overall mean, T_i = treatment effect (i = CTL, FLX, or FO), C(T)_ij = cow or calf j nested in treatment i, and E_ijk = random residual. The data were analyzed using 9.4 Statistical Analysis System software after verifying the normality via the Shapiro-Wilk PROC single variable residual. The data were shown as the mean ± standard error of the mean (SEM) and declared significant when it reached P ≤ 0.05, and a tendency at 0.05 < P < 0.10 by Tukey’s test. The percentages of long-chain n-3 FAs EPA (C20:5n-3), DHA (C22:6n-3) and their intermediate docosapentanoic acid (DPA, C22:5n-3) from the calves’ plasma were not normally distributed; therefore, we analyzed the frequency of the appearance of these FAs by PROC FREQ and they were declared significant when they reached P ≤ 0.05 by Fisher’s test.

Several genes involved in lipid metabolism were examined; maternal FO supplementation increased the expression of the FABP4 gene, which is involved in FA transfer in the placenta, by 5-fold more than CTL (P = 0.05; Fig. 1A), with no difference observed between CTL and FLX. Additionally, there was a tendency towards higher gene expression of peroxisome proliferator-activated-receptor gamma (PPARG; P = 0.06; Fig. 1A) in the FO placenta, compared with CTL. On the other hand, the average relative expressions of sterol regulatory element binding transcription factor 1 (SREBP1; P = 0.51) and fatty acid synthase (FASN; P = 0.31) were similar between treatments (data not shown).

In the analysis of the inflammation response genes, as shown in Fig. 1B, the FO placenta had higher gene expression of interleukin 10 (IL-10; P = 0.05) and a tendency for an increased expression of IL-6 (P = 0.06) compared with CTL, with no difference observed between FLX and CTL. No significant differences between treatments were observed regarding the placental gene expressions of interleukin 1b (IL-1b; P = 0.28), interleukin-6 receptor (IL6-R; P = 0.32), the toll-like receptor 4 (TLR4; P = 0.44), and tumor necrosis factor alpha (TNFα; P = 0.86; data not shown).

Among the oxidative stress-related genes, the expression of catalase (CAT; P = 0.04) was significantly higher in FO than in CTL (Fig. 1C). There were no differences in glutathione peroxidase 3 (GPX3; P = 0.17) and superoxide dismutase 1 (SOD1; P = 0.45) expressions between treatments (data not shown).

Furthermore, maternal FO supplementation led to a significant increase in the placental expression of the ECS-related genes cannabinoid receptor 1 (CNR1; P = 0.03) and cannabinoid receptor 2 (CNR2; P = 0.04) compared with CTL (Fig. 1D). Both n-3 FA treatments (FLX and FO) significantly decreased the placental gene expression of the ECS enzyme FAAH (P = 0.002) compared with CTL (Fig. 1D). No difference was observed between treatments regarding the gene expressions of the ECS enzymes MGLL (P = 0.20) and N-acyl phosphatidylethanolamine phospholipase D (NAPEPLD; P = 0.57; data not shown).

Overall, the proteomic analysis of the bovine placenta identified 3,974 proteins (Additional file 2). Of these, 51 proteins were differentially abundant (P ≤ 0.05 and |FC| ≥ 1.5) in FLX vs. CTL, whereas 59 were differential in FO vs. CTL, and 51 were differential in FO vs. FLX. Volcano plots illustrated that compared with CTL, the FLX treatment up-regulated 66.7% and down-regulated 33.3% of DAPs (Fig. 2D), whereas the FO treatment up-regulated 44.1% and down-regulated 55.9% of DAPs (Fig. 2E). Comparing DAPs in FO vs. FLX, 52.9% were up-regulated and 47.1% were down-regulated (Fig. 2F).

Of the DAPs, 2 proteins of the FABP family, FABP3 and FABP7, were significantly increased in FLX and FO compared with CTL. Interestingly, the proteomic data also revealed a higher abundance of the FABP5 (FC = 1.64, P = 0.04) in FO placenta vs. FLX. An additional protein implicated in the lipid metabolism was Caspase 2 (CASP2; FC = 5.04, P = 0.03), which was more abundant in FLX than in CTL. Furthermore, the n-3 FAs dietary treatments affected immune related proteins: FLX increased the abundance of CD44 antigen (CD44; FC = 2.04, P = 0.04) and secreted phosphoprotein 1 (SPP1; FC = 5.41, P = 0.02), whereas the prostaglandin-endoperoxide synthase 2 (PTGS2; FC = −1.97, P = 0.01) decreased in FO. The antioxidative protein NAD(P)H quinone dehydrogenase 1 (NQO1) increased in FLX, compared with CTL (FC = 1.51, P = 0.003), whereas GPX3 increased in FO, compared with FLX (FC = 1.67, P = 0.02). Another interesting protein that differed regarding FO vs. FLX was the ECS component CNR1-cannabinoid receptor-interacting protein 1 (CNRIP1), which was reduced in FO vs. FLX placenta (FC = −1.72, P = 0.02).

The DAPs (proteins with |FC| ≥ 1.5 and P ≤ 0.05) were analyzed using Qiagen’s Ingenuity^® Pathway Analysis to identify the most relevant pathways, functions, and networks affected by FLX or FO dietary treatments. The main functions identified for DAP are presented in Additional file 1: Fig. S2. The most prominent biological function that changed between the treatments was regarding the levels of proteins with enzymatic functions: approximately 40% of DAPs in each comparison. Interestingly, FO supplementation resulted in a marked increase in the proportion of kinases and phosphatases among the differentially abundant enzymes, compared with FLX. Another important functional category identified in all the analyses was that of the proteins involved in transcription and translation regulation (Additional file 1: Fig. S2).

In-silico docking studies performed on bovine FABPs confirmed that FABP3 binds the n-6 series endocannabinoids AEA (4.16 kcal/mol, Fig. 5A) and 2-AG (6.35 kcal/mol, Additional file 3: Fig. S4) with good fitness scores. 2-AG and AEA also bind to FABP5 (5.90 kcal/mol and 3.55 kcal/mol, respectively) with good fitness scores (Additional file 3: Fig. S5). On the other hand, 2-AG and AEA had low fitness scores when binding to FABP7 (0.06 kcal/mol and 0.15 kcal/mol, respectively; Additional file 3: Fig. S6); however, as shown in Fig. 5B, FABP7 had a good binding score with the n-3 series endocannabinoid DHEA (10.43 kcal/mol).

First, we assessed the effects of maternal n-3 FA supplementation on the FA profile in the placenta tissues and the plasma of neonate calves. As shown in Table 3, in the placenta tissues, the average percentages of the n-3 FAs and the n-6/n-3 FA ratio remained similar across the treatments. On the other hand, in neonate calves, the plasma FA profile was affected by the maternal supplementation of the n-3 FAs; FO calves had a higher average percentage of total n-3 FAs (P = 0.02) compared with CTL (Table 4). Furthermore, a larger number of FO calves displayed increased percentages of plasma C20:5n-3 (EPA, P = 0.001), C22:6n-3 (DHA, P = 0.003), and their intermediate C22:5n-3 (DPA, P = 0.003), compared with both the CTL and FLX calves (Table 4).

Next, we evaluated the potential impact of maternal dietary n-3 FAs supplementation on the inflammatory status of neonatal calves by quantifying the plasma concentrations of several inflammatory markers on the day of calving before colostrum offering. A significant reduction of 14% and 22% in the average plasma concentrations of IL-6 was observed in the FLX and FO calves, respectively, compared with the CTL calves (P = 0.001; Table 5). On the other hand, there was no significant difference in IL-2 concentrations in both n-3 FAs treatments compared to CTL (Table 5); however, it was 22.8% higher in FLX than in the FO calves (P = 0.02). No differences were observed regarding the average concentrations of haptoglobin among the treatments (Table 5).

The fetus is dependent on the maternal supply of LCPUFAs; thus, the maternal FA composition is crucial for fetal growth and development, especially during late pregnancy, when DHA is particularly important for fetal brain development [46]. Placental cells are known to transfer the FAs selectively, with a preference for LCPUFAs, such as EPA and DHA. The selective uptake may involve intracellular metabolic channeling and a selective supply to the fetal circulation [2, 47, 48]. Several protein families were implicated in directional FA transport in the placenta: the membranal FATPs, FA translocase (FAT/CD36), and the cytoplasmic FABPs [2, 48, 49]. The uptake and accumulation of FAs are regulated by FABPs; FABP3, FABP4, FABP5, and FABP7 are expressed in cells with high FA requirements, such as the placenta [48, 50, 51]. In this study, we found that several FABPs were upregulated in FLX and/or the FO placenta; the abundances of FABP3 and FABP7 increased in both n-3 FAs groups, and these increases were related to the enriched pathway of triglyceride metabolism, whereas FABP4 gene expression was augmented by FO supplementation. Furthermore, FABP5 was higher in FO than in FLX placenta. FABPs have differential affinities to specific FAs: FABP5 binds to both DHA and AA, whereas FABP7 binds specifically to DHA [47, 52–54], and both FABP4 and FABP5 specifically increase the uptake of DHA [55–58]. Furthermore, although FABP3 was shown to bind preferentially to AA (n-6 FA) [54], it regulates both n-3 and n-6 FA transport in mouse trophoblasts [59]. Studies in bovine ovarian granulosa [60] and human placental cells [61] showed that the expression of FABP3, which may regulate the accumulation of triglycerides and the capacity of lipid transfer, can be influenced by the type of LCPUFA and the maternal health condition. In the present study, the plasma FA profile of neonate FO calves showed an increase in total n-3 FAs compared with CTL. We have previously demonstrated that the maternal plasma of FO cows was enriched with DHA and EPA [30]. Taken together, along with the increase in specific placental FABPs, we suggest that the increased maternal n-3 FAs may lead to increased FABPs in the placenta, which favors the trafficking of n-3 FAs to the fetus. The apparent lack of the effect of the dietary FA supplementation on the FA profile in placental cotyledons may stem from a highly efficient maternal-fetal transfer and/or rapid utilization of the FAs for placental functions.

Our proteomic data showed an increased abundance of CASP2 in the FLX placenta, compared with CTL. CASP2 is a protein in the apoptotic cascade, which was recently identified as a positive regulator of cholesterol and triacylglycerol homeostasis in human cells and mice [62, 63]. According to our bioinformatic analysis, CASP2 is involved in lipid concentration and triglyceride metabolism; it was demonstrated that the human CASP2 gene contains multiple sites that can be recognized by transcriptional regulators of the sterol regulatory element-binding protein (SREBP) family [62], which is involved in regulating pathways for cholesterol, triacylglycerol, and phospholipid synthesis [64]. In the placenta, the FAs are esterified into triglycerides and stored in droplets within the cells, after which they are transferred to the developing fetus [65]; thus, variations in the abundance of CASP2 may lead to modification of the cellular lipid levels, and specifically triacylglycerol [62, 63]. Therefore, our results suggest that maternal n-3 FA supplementation may modulate the metabolic pathway of triacylglycerol/lipid in the placenta by affecting the abundance of CASP2 protein.

Taken together, our results suggest that FO had a greater effect on the transfer of n-3 LCPUFAs to the fetus than did FLX. This transfer may be specifically mediated by modulating the FABPs in the placenta.

Changing the n-3 to n-6 ratio in the diet is a well-known strategy to modulate the ECS [66]. We previously reported that in dairy cows, peripartum n-3 FA supplementation reduces the abundance of certain ECS components in the adipose tissue, liver, and white blood cells [24, 31]. However, in this study, in the placental tissue, FLX supplementation did not affect the gene expression of most of the examined ECS components, except for the FAAH gene. Intriguingly, the placenta from FO cows exhibited a higher gene expression of the primary membranal ECS receptors (CNR1 and CNR2), and a tendency toward increased gene expression of a secondary ECS nuclear receptor PPARG. Both the CNR1 and CNR2 receptors have important functions in reproductive organs, including the placenta [67, 68], since activation of the ECS receptors can affect both pregnancy progression and labor [29, 67] by controlling the action of endocannabinoids, cytokine release, and by modulating the mitochondrial activity [69, 70]. Within the cells, n-3 FA DHA and EPA can be converted to the endocannabinoids DHEA and eicosapentaenoyl ethanolamide (EPEA), which have a chemical structure similar to the endocannabinoid AEA; it was reported that DHEA and EPEA exhibit binding affinity and agonist activity on both CNR1 and CNR2 [71, 72]. Thus, the effects of FO on the ECS can be mediated by these n-3 FA-derived endocannabinoids. However, in the present study we did not measure the levels of endocannabinoids in placenta; therefore, further studies are required to investigate this possibility.

Interestingly, FABP3, FABP5, and FAB7 are also involved in the uptake, intracellular transport, and hydrolysis of endocannabinoids [51, 55, 73–76]. Since our data showed increased abundances of several FABPs in n-3 FA supplemented placentas, together with modulations in the ECS components, we investigated, for the first time in bovine, the in-silico binding of these Bos taurus FABPs with endocannabinoids. We found that FABP3 and FABP5 showed good fitness scores in the docking with both 2-AG and AEA, whereas bovine FABP7 had a good fitness score with DHEA. These findings may indicate an association between lipid metabolism and ECS within the bovine placenta. Furthermore, we observed a lower abundance of CNRIP1 protein in FO compared with FLX. CNRIP1 is involved in modulating part of the CNR1, thereby regulating cell signaling initiated by modulation of CNR1 [77]. However, the function of CNRIP1 in placental cells is not yet fully understood. To the best of our knowledge, this is the first study to identify CNRIP1 in the synepitheliochorial placenta, and there is a suggested link between n-3 FA supplementation and this ECS component.

Modulating the ECS can also affect inflammatory processes. In myoblast cell cultures, EPEA and DHEA decreased the gene expression of the inflammatory marker IL-6 [78]. However, in our study, the upregulation of the CNR in FO placenta coincided with a tendency to increase IL-6 expression. On the other hand, both FO and FLX placentas had reduced gene expression of FAAH compared with CTL. FAAH inhibition is considered to have anti-inflammatory effects [79, 80]. Taken together, we propose that maternal FO supplementation can stimulate placental ECS, which may be associated with activating the inflammatory and lipid metabolism pathways.

Within the placenta, pro-inflammatory cytokines are produced during parturition, most likely contributing to uterine contractions and the expulsion of the fetus [81, 82]. Note that we specifically sampled the cotyledons of the expelled placentas, namely, the maternal-fetal interface tissues post-detachment. When we examined the expression of inflammatory genes in the placental cotyledons, we found an increase in the expression of anti-inflammatory IL-10 and a tendency toward a higher expression of pro-inflammatory IL-6 in FO compared with CTL. In addition, an increase in the protein abundance of CD44 and SPP1, both associated with the activation of immune cells during inflammation [83], was found in FLX compared to CTL, and SPP1 tended to be increased in FO compared with CTL. Moreover, our analysis revealed an enrichment of the MHC class II antigen presentation and the leukocyte extravasation signaling pathways, as well as in the connection to inflammatory molecules, such as p38 MAPK and NFkB in the network, which are pro-inflammatory, in FLX vs. CTL [84]. In addition, in FO vs. CTL we found an enrichment in the MIF regulation of innate immunity, NOD1/2 signaling, and the MAP kinase activation pathways, which are associated with activation and control of the inflammatory process [85–87]. These findings indicate that maternal supplementation of n-3 FAs mostly upregulates the inflammatory process in the expelled placenta, which contradicts the well-described anti-inflammatory effects of n-3 FAs in other tissues. However, our analyses were conducted specifically on placental cotyledons, revealing a specific localized pro-inflammatory response. This observation aligns with the findings of Peng et al. [23], who demonstrated that EPA supplementation can induce a distinctive local pro-inflammatory effect in placenta from mice. However, in contrast to the putative upregulation of the inflammatory processes in the placenta, the proteomic data in FO vs. CTL showed a decreased abundance of PTGS2, encoding the enzyme cyclooxygenase-2, which regulates the synthesis of pro-inflammatory prostaglandins [88, 89]. The lower abundance of PTGS2 may indicate the anti-inflammatory effects of FO, which is in line with the increased gene expression of IL-10, thus overall, suggesting that FO has both pro- and anti-inflammatory effects on the bovine placenta.

In the present study we also investigated the effects of maternal n-3 FA supplementation on inflammatory markers in the neonate calves. We found that both FLX and FO reduced the plasma concentrations of IL-6 in neonates, which is in accordance with the well-described anti-inflammatory effects of n-3 FAs. Although the placental synepteliochorial does not allow the transfer of large molecules, such as immunoglobulins and lipid-insoluble molecules from the maternal blood flow to the fetus [12, 90], based on our results, it is possible that during late pregnancy, the maternal diet can influence the inflammatory response of neonates before the ingestion of colostrum, by facilitating placental transfer of n-3 FAs that modulate the release of cytokines, thus, highlighting the role of the placenta in the immune status of neonates during their first hours of life.

The placental cells exhibit a high level of mitochondrial activity, which leads to the production of ROS [17] and increased oxidative stress [91]. Our proteomic data indicated that ROS synthesis was altered significantly in FLX placenta compared with CTL. In addition, we identified the up-regulation of CAT expression in FO placentas, whereas the expressions of GPX3 and SOD1 were similar across the treatments. In rats, n-3 FA supplementation increases the oxidative stress response in placental tissue, along with a positive correlation with the gestational quality [92]. Dietary n-3 PUFAs, and specifically DHA, were shown to act as activators of the Nrf2 antioxidant pathway [93–96]. Nrf2 is a transcriptional activator that regulates the expression of antioxidant genes including SOD1, NQO1, CAT, and GPX3 [97]. Indeed, we found that the abundance of NQO1 increased in the placental cotyledon cells of the FLX group. The NQO1 is a cytoplasmic protein that plays a key role in protecting against oxidative stress by stabilizing various proteins and preventing the reduction of a single electron that leads to the production of ROS [8]. We found that in FLX vs. CTL, NQO1 was most likely activated in biological functions related to ROS synthesis. Interestingly, although the analysis of the proteomic data did not reveal a significant impact on the ROS-related functions in the FO placenta, we did find an increased gene expression of the antioxidant enzyme CAT in FO compared with CTL. In a study with diabetic rats, supplementation with FO increased the gene expression of CAT, which is in agreement with our findings [98]. In the present study, there was a decrease in the gene expression of the ECS enzyme FAAH in both FO and FLX, compared with CTL. FAAH degrades AEA into ethanolamine and AA. AA, besides constituting a high percentage of the membrane phospholipids in cells, also serves as a precursor for the synthesis of prostaglandins-ethanolamids or ‘prostamides’, such as prostamide-E2 (PME2) [5, 89]. PME2 exhibits proapoptotic action and can lead to the production of ROS [68, 89]; therefore, reduction of FAAH may lessen oxidative stress.

In comparing the FO and FLX treatments, two functions associated with the production and synthesis of ROS were most likely inhibited: One of the proteins assigned to this function was GPX3, a member of the enzyme antioxidant family whose main biological role is to minimize oxidative damage by catalyzing the removal of lipid peroxides, which are products of ROS activity [99]. Indeed, GPX3 abundance was upregulated in FO compared with FLX. In sheep, Garrel et al. [100] suggested that GPX may be the major enzyme in defending against ROS in the fetus-placental unit. Overall, the increase in the oxidative stress response in the n-3 FA supplemented placental tissues is important to preserve gestational function [92]. An important consideration in interpreting our results is the specific highly- specialized tissue examined and the timing of the sample collection, which displays a physiological increase in the oxidative processes. Based on our results indicating that several oxidative stress components at the transcriptional and post-transcriptional levels were stimulated both in FLX and in FO, we suggest that maternal supplementation of n-3 FAs can have antioxidant effects on the bovine placenta.

The experimental protocol for this study was approved by the Volcani Center Animal Care Committee (approval number IL 797/18), and it was performed in accordance with the relevant guidelines and regulations.

Not applicable.

The authors declare that they have no competing interests.

The proteomic dataset generated and/or analyzed during the current study is available in the Supplementary files.