Advertisement

Metabolic and antioxidant status during transition is associated with changes in the granulosa cell transcriptome in the preovulatory follicle in high-producing dairy cows at the time of breeding

Open AccessPublished:July 12, 2022DOI:https://doi.org/10.3168/jds.2022-21928

      ABSTRACT

      In this study, we hypothesized that early postpartum (pp) metabolic and oxidative stress conditions in dairy cows (particularly those with severe negative energy balance, NEB) are associated with long-term changes in granulosa cell (GC) functions in the preovulatory follicle at the time of breeding. Blood samples were collected at wk 2 and wk 8 pp from 47 healthy multiparous cows. Follicular fluid (FF) and GC were collected from the preovulatory follicle after estrous synchronization at wk 8. Several metabolic and antioxidant parameters were measured in blood and FF, and their correlations were studied. Subsequently, 27 representative GC samples were selected for RNA sequencing analysis. The GC gene expression data of LH-responsive genes and the estradiol:progesterone ratio in FF were used to identify pre- and post-LH surge cohorts. We compared the transcriptomic profile of subgroups of cows within the highest and lowest quartiles (Q4 vs. Q1) of each parameter, focusing on the pre-LH surge cohort (n = 16, at least 3 in each subgroup). Differentially expressed genes (DEG: adjusted P-value < 0.05, 5% false discovery rate) were determined using DESeq2 analysis and were functionally annotated. Blood and FF β-carotene and vitamin E concentrations at wk 2, but not at wk 8, were associated with the most pronounced transcriptomic differences in the GC, with up to 341 DEG indicative for lower catabolism, increased oxidoreductase activity and signaling cascades that are known to enhance oocyte developmental competence, increased responsiveness to LH, and a higher steroidogenic activity. In contrast, elevated blood NEFA concentrations at wk 2 (and not at wk 8) were associated with a long-term carryover effect detectable in the GC transcriptome at wk 8 (64 DEG). These genes are related to response to lipids and ketones, oxidative stress, and immune responses, which suggests persistent cellular stress and oxidative damage. This effect was more pronounced in cows with antioxidant deficiencies at wk 8 (up to 148 DEG), with more genes involved in oxidative stress-dependent responses, apoptosis, autophagy and catabolic processes, and mitochondrial damage. Interestingly, within the severe NEB cows (high blood NEFA at wk 2), blood antioxidant concentrations (high vs. low) at wk 8 were associated with up to 194 DEG involved in activation of meiosis and other signaling pathways, indicating a better oocyte supportive capacity. This suggests that the cow antioxidant profile at the time of breeding might alleviate, at least in part, the effect of NEB on GC functions. In conclusion, these results provide further evidence that the metabolic and oxidative stress in dairy cows early postpartum can have long-term effects on GC functions in preovulatory follicles at the time of breeding. The interplay between the effects of antioxidants and NEFA illustrated here might be useful to develop intervention strategies to minimize the effect of severe NEB on fertility.

      Key words

      INTRODUCTION

      Many high-yielding dairy cows suffer from negative energy balance (NEB) during the early postpartum period due to a mismatch between DMI and the high energy demand for milk production (
      • Leroy J.L.
      • Opsomer G.
      • Van Soom A.
      • Goovaerts I.G.
      • Bols P.E.
      Reduced fertility in high-yielding dairy cows: Are the oocyte and embryo in danger? Part I. The importance of negative energy balance and altered corpus luteum function to the reduction of oocyte and embryo quality in high-yielding dairy cows.
      ). Cows may suffer from excessive fat mobilization and high rate of BCS loss, which may lead to metabolic and oxidative stress conditions. These changes have been described in detail and are strongly correlated with a negative effect on fertility (
      • Sakaguchi M.
      Practical aspects of the fertility of dairy cattle.
      ).
      Reduced fertility in modern high-yielding dairy cows is one of the major challenges in today's dairy industry (
      • Roche J.R.
      • Burke C.R.
      • Crookenden M.A.
      • Heiser A.
      • Loor J.L.
      • Meier S.
      • Mitchell M.D.
      • Phyn C.V.C.
      • Turner S.A.
      Fertility and the transition dairy cow.
      ) and is responsible for more than 25% of culling on dairy farms (
      • Chiumia D.
      • Chagunda M.G.
      • Macrae A.I.
      • Roberts D.J.
      Predisposing factors for involuntary culling in Holstein-Friesian dairy cows.
      ). The majority of these cows are culled during their first or second lactation, which creates a burden on farm economics and affects livestock sustainability (
      • Brickell J.S.
      • Wathes D.C.
      A descriptive study of the survival of Holstein-Friesian heifers through to third calving on English dairy farms.
      ).
      Meeting fertility targets depends on the success rate and time at which a cow becomes pregnant again after calving. This in turn depends on uterine health and the speed of uterine recovery after calving, together with an early resumption of ovarian cyclicity (
      • Leroy J.L.
      • Vanholder T.
      • Van Knegsel A.T.
      • Garcia-Ispierto I.
      • Bols P.E.
      Nutrient prioritization in dairy cows early postpartum: Mismatch between metabolism and fertility?.
      ). Subsequently, normal follicular growth with the formation of a healthy follicular microenvironment to support the development of a good quality oocyte at the time of breeding is a key factor that determines the cow's reproductive performance.
      An important factor that influences dairy cow fertility during the early postpartum period is the increase in blood nonesterified fatty acid (NEFA) concentrations due to excessive fat mobilization (
      • Shin E.K.
      • Jeong J.K.
      • Choi I.S.
      • Kang H.G.
      • Hur T.Y.
      • Jung Y.H.
      • Kim I.H.
      Relationships among ketosis, serum metabolites, body condition, and reproductive outcomes in dairy cows.
      ;
      • Miqueo E.
      • Chiarle A.
      • Giuliodori M.J.
      • Relling A.E.
      Association between prepartum metabolic status and resumption of postpartum ovulation in dairy cows.
      ). Elevated blood NEFA during the first 2 wk postpartum (pp) depends on the degree of NEB and is also reflected in the follicular fluid (FF;
      • Leroy J.L.
      • Vanholder T.
      • Mateusen B.
      • Christophe A.
      • Opsomer G.
      • de Kruif A.
      • Genicot G.
      • Van Soom A.
      Non-esterified fatty acids in follicular fluid of dairy cows and their effect on developmental capacity of bovine oocytes in vitro.
      ). Accumulation of NEFA in the peripheral tissues causes lipotoxic effects on different cell functions (
      • Yoon H.
      • Shaw J.L.
      • Haigis M.C.
      • Greka A.
      Lipid metabolism in sickness and in health: Emerging regulators of lipotoxicity.
      ). Previous studies have shown that injection of high concentrations of NEFA in the ovarian follicle in vivo or exposure of follicles or granulosa cells (GC) to high concentrations of NEFA in vitro has been shown to cause apoptosis, altered steroidogenesis, and reduced cell proliferation and follicular growth (
      • Mu Y.M.
      • Yanase T.
      • Nishi Y.
      • Tanaka A.
      • Saito M.
      • Jin C.H.
      • Mukasa C.
      • Okabe T.
      • Nomura M.
      • Goto K.
      • Nawata H.
      Saturated FFAs, palmitic acid and stearic acid, induce apoptosis in human granulosa cells.
      ;
      • Valckx S.D.
      • Van Hoeck V.
      • Arias-Alvarez M.
      • Maillo V.
      • Lopez-Cardona A.P.
      • Gutierrez-Adan A.
      • Berth M.
      • Cortvrindt R.
      • Bols P.E.
      • Leroy J.L.
      Elevated non-esterified fatty acid concentrations during in vitro murine follicle growth alter follicular physiology and reduce oocyte developmental competence.
      ;
      • Baddela V.S.
      • Sharma A.
      • Vanselow J.
      Non-esterified fatty acids in the ovary: Friends or foes?.
      ;
      • Ferst J.G.
      • Missio D.
      • Bertolin K.
      • Gasperin B.G.
      • Leivas F.G.
      • Bordignon V.
      • Goncalves P.B.
      • Ferreira R.
      Intrafollicular injection of nonesterified fatty acids impaired dominant follicle growth in cattle.
      ).
      • Aardema H.
      • Gadella B.M.
      • van de Lest C.H.
      • Brouwers J.F.
      • Stout T.A.
      • Roelen B.A.
      • Vos P.L.
      Free fatty acid levels in fluid of dominant follicles at the preferred insemination time in dairy cows are not affected by early postpartum fatty acid stress.
      demonstrated that cows with high blood NEFA concentrations at wk 1 to 2 pp exhibited delayed first ovulation (>35 d pp; i.e., delayed postpartum ovarian cyclicity). Although the postpartum voluntary waiting period varies from one farm to another, the preferred time of first insemination is most commonly after 50 d. By that time, NEFA concentrations have returned to basal physiological levels (
      • Leroy J.L.
      • Vanholder T.
      • Mateusen B.
      • Christophe A.
      • Opsomer G.
      • de Kruif A.
      • Genicot G.
      • Van Soom A.
      Non-esterified fatty acids in follicular fluid of dairy cows and their effect on developmental capacity of bovine oocytes in vitro.
      ;
      • Aardema H.
      • Gadella B.M.
      • van de Lest C.H.
      • Brouwers J.F.
      • Stout T.A.
      • Roelen B.A.
      • Vos P.L.
      Free fatty acid levels in fluid of dominant follicles at the preferred insemination time in dairy cows are not affected by early postpartum fatty acid stress.
      ). Nevertheless, cows with the highest BCS change during the early postpartum period (i.e., with a history of severe NEB and highest NEFA concentration at wk 1–2) produced higher proportions of low quality, fragmented embryos after superovulation and AI at 8 to 9 wk and had lower pregnancy rates compared with other cows with less BCS changes early postpartum (
      • Carvalho P.D.
      • Souza A.H.
      • Amundson M.C.
      • Hackbart K.S.
      • Fuenzalida M.J.
      • Herlihy M.M.
      • Ayres H.
      • Dresch A.R.
      • Vieira L.M.
      • Guenther J.N.
      • Grummer R.R.
      • Fricke P.M.
      • Shaver R.D.
      • Wiltbank M.C.
      Relationships between fertility and postpartum changes in body condition and body weight in lactating dairy cows.
      ). Loss of BCS during the first 3 wk pp is also strongly linked with reduced pregnancy or AI rates in cows bred after estrus synchronization (
      • Wiltbank M.C.
      • Carvalho P.D.
      • Barletta R.V.
      • Fricke P.M.
      • Shaver R.D.
      Relationships among changes in body condition score and reproductive efficiency in lactating dairy cows.
      ). This suggests that the degree of fat mobilization during the early postpartum period might have a long-lasting effect on follicular physiology and oocyte quality up to several weeks after the end of the NEB period. This notion was postulated 30 yr ago and is better known as the Britt hypothesis (
      • Britt J.H.
      Impacts of early postpartum metabolism on follicular development and fertility.
      ). However, only a few studies have provided data to support this hypothesis, and the mechanisms mediating such a carryover effect have not been elucidated. In addition, the uncoupled growth hormone–IGF-1 axis during transition leads to low IGF-1 concentrations. This affects reproductive functions because IGF-1 is known to enhance follicular growth, estradiol synthesis, and LH secretion (
      • Webb R.
      • Garnsworthy P.C.
      • Gong J.G.
      • Armstrong D.G.
      Control of follicular growth: Local interactions and nutritional influences.
      ). However, the carryover effect on follicle physiology at the moment of breeding is not known.
      In addition to energy shortage–related metabolic stress, elevated oxidative stress (OS) levels play an important role in a dairy cow's vulnerability to disease. In a large study across Europe involving >2,500 cows on 127 farms, it was found that more than 97% and 78% of the cows were deficient in vitamin E (Vit. E) and β-carotene (βC), respectively, early after calving, indicating higher metabolic needs and reduced dietary uptake during transition (
      • Mary A.E.P.
      • Artavia Mora J.I.
      • Ronda Borzone P.A.
      • Richards S.E.
      • Kies A.K.
      Vitamin E and beta-carotene status of dairy cows: A survey of plasma levels and supplementation practices.
      ). Other studies confirmed higher rates of OS early postpartum, as measured by evaluating both pro-oxidant and antioxidant levels (oxidative stress index, OSi;
      • Abuelo A.
      • Hernández J.
      • Benedito J.L.
      • Castillo C.
      Oxidative stress index (OSi) as a new tool to assess redox status in dairy cattle during the transition period.
      ). Elevated OS is known to disrupt cellular functions due to protein and lipid peroxidation, altered gene expression, and cell signaling, which can contribute to, or exacerbate, the effect of metabolic stress on reproductive functions. Similar to those of NEFA, antioxidant concentrations return to optimal levels by the time of breeding, 2 to 3 mo after calving (
      • De Bie J.
      • Proost K.
      • Van Loo H.
      • Callens J.
      • Bols P.E.J.
      • Fransen E.
      • Leroy J.L.M.R.
      β-carotene and vitamin E in the dairy industry: Blood levels and influencing factors—A case study in Flanders.
      ;
      • Mary A.E.P.
      • Artavia Mora J.I.
      • Ronda Borzone P.A.
      • Richards S.E.
      • Kies A.K.
      Vitamin E and beta-carotene status of dairy cows: A survey of plasma levels and supplementation practices.
      ). Nevertheless, it is not known whether deficiencies in antioxidants during the early postpartum period might also contribute to the long-lasting negative effect on follicle physiology at the time of breeding.
      Exposure of follicular cells to metabolic (e.g., lipotoxic) stress or other suboptimal conditions (e.g., antioxidant deficiencies), together with suboptimal concentrations and bioavailability of growth factors such as IGF-1, can have serious consequences on follicular cell proliferation, steroidogenic capacity, circulating estradiol concentrations, and response to gonadotropic hormones, and cause alterations in the timing of the LH surge and ovulation (for review, see
      • Roche J.R.
      • Burke C.R.
      • Crookenden M.A.
      • Heiser A.
      • Loor J.L.
      • Meier S.
      • Mitchell M.D.
      • Phyn C.V.C.
      • Turner S.A.
      Fertility and the transition dairy cow.
      ). Granulosa cells also deliver cargos rich in energy substrates, signaling molecules, antioxidative factors, and RNA transcripts (
      • Macaulay A.D.
      • Gilbert I.
      • Caballero J.
      • Barreto R.
      • Fournier E.
      • Tossou P.
      • Sirard M.A.
      • Clarke H.J.
      • Khandjian E.W.
      • Richard F.J.
      • Hyttel P.
      • Robert C.
      The gametic synapse: RNA transfer to the bovine oocyte.
      ) to the oocyte through transzonal projections and gap junctions. These compounds are crucial for oocyte cytoplasmic maturation and postfertilization development (
      • Da Broi M.G.
      • Giorgi V.S.I.
      • Wang F.
      • Keefe D.L.
      • Albertini D.
      • Navarro P.A.
      Influence of follicular fluid and cumulus cells on oocyte quality: Clinical implications.
      ). Granulosa cell gene expression patterns have been successfully linked to the oocyte's developmental capacity (
      • Uyar A.
      • Torrealday S.
      • Seli E.
      Cumulus and granulosa cell markers of oocyte and embryo quality.
      ;
      • Liu C.
      • Su K.
      • Chen L.
      • Zhao Z.
      • Wang X.
      • Yuan C.
      • Liang Y.
      • Ji H.
      • Li C.
      • Zhou X.
      Prediction of oocyte quality using mRNA transcripts screened by RNA sequencing of human granulosa cells.
      ). Effects on GC functions may also alter postovulation events; for example, due to formation of a deficient corpus luteum and reduced progesterone concentrations. This may affect oviducal functions, uterine receptivity, early embryo survival (particularly in the first 7 d), and implantation (
      • Sartori R.
      • Bastos M.R.
      • Wiltbank M.C.
      Factors affecting fertilisation and early embryo quality in single- and superovulated dairy cattle.
      ).
      In this study, we hypothesize that early postpartum metabolic and OS conditions in dairy cows are associated with changes in the GC transcriptomic profile in the preovulatory follicle at the time of breeding. The aim of this study was to examine whether metabolic stress and antioxidant deficiencies during the early postpartum period (wk 2 pp) can influence the FF microenvironment and are linked with functional alterations in the physiology of the GC (studied here at the transcriptomic level) in the preovulatory follicle at the time of breeding (wk 8 pp). We also investigated whether an optimal antioxidant profile could attenuate the effect of NEB on GC physiology.

      MATERIALS AND METHODS

      Animals and Experimental Design

      All animal interventions and procedures were ethically approved by the Ethical Committee for Animal Testing, University of Antwerp (dossier number 2018–09). Forty-seven multiparous Holstein Friesian cows kept under the same housing and livestock management conditions were randomly selected, including cows with various BCS changes during the transition period (mean BCS = 2.9, range = 2.5–3.6; mean BCS loss = 0.4, range = 0.0–0.7) and various daily milk yields in the previous lactation (mean = 31.55, range = 20.1–41.7 kg). Cows were fed grass and maize silage supplemented with beet pulp, concentrates, and a standard mineral supplement during the dry period until calving. Cows received calcium, phosphorus, magnesium, and sodium as extra supplements during lactation. Cows were milked and monitored daily using an automated milking robot (VMS, De Laval). Only healthy animals that did not suffer from any disease (e.g., lameness, mastitis, and other respiratory and digestive system diseases) during the last 3 mo before the start of the experiment were included in the study. Blood samples were collected at 14 and 60 DIM via the jugular vein. These time points were selected to coincide with the potential metabolic stress due to NEB at wk 2 pp and the time at which cows are routinely prepared for breeding (wk 8 pp). Ultrasound-guided transvaginal follicular aspiration of the preovulatory follicle was performed at wk 8 following estrous synchronization for the collection of FF and GC (Figure 1).
      Figure thumbnail gr1
      Figure 1Experimental design and sample collection. Blood was collected from 47 multiparous cows at wk 2 and wk 8 postpartum (pp). Follicular fluid was collected at wk 8 following estrous synchronization and transvaginal ultrasound-guided needle aspiration. Granulosa cells were separated from the follicular fluid by centrifugation at 971 × g for 5 min at room temperature (18–22°C). PRID = progesterone-releasing intravaginal device; OPU = ovum pick-up.

      Synchronization Protocol

      At wk 7 pp ± 3 d, estrus synchronization of the cows began by inserting a progesterone-releasing intravaginal device (PRID Delta; 1.55 g of progesterone; CEVA). The presence of the PRID in the vagina was checked twice daily thereafter. Six days after PRID insertion, cows received a single intramuscular injection of Dinolytic (Dinoprostum, 5 mg/mL, 5 mL per cow; Pfizer) at 1800 h. The PRID was removed 24 h later. On d 9, exactly 38 h after PRID removal, ultrasound-guided transvaginal follicular aspiration of the dominant follicle was performed, as described previously (
      • Bols P.E.J.
      • Vandenheede J.M.M.
      • Van Soom A.
      • de Kruif A.
      Transvaginal ovum pick-up (OPU) in the cow: A new disposable needle guidance system.
      ;
      • Leroy J.L.
      • Vanholder T.
      • Mateusen B.
      • Christophe A.
      • Opsomer G.
      • de Kruif A.
      • Genicot G.
      • Van Soom A.
      Non-esterified fatty acids in follicular fluid of dairy cows and their effect on developmental capacity of bovine oocytes in vitro.
      ).

      Sampling of Blood and Collection of Serum

      Blood samples were collected at wk 2 and 8 using a Vacuette system (Greiner Bio-One) in EDTA, sodium fluoride (NaF), and serum clot-activating tubes. Blood tubes were kept on ice during transport to the laboratory, except the serum tubes, which were transported at room temperature (RT; 18–22°C). Blood was centrifuged at the laboratory within 3 h after collection (serum tubes: 30 min, 1,400 × g at RT; EDTA tubes: 10 min, 1,000 × g at 4°C; NaF tubes: 10 min, 1,250 × g at 4°C). Serum and plasma samples were then aliquoted in cryotubes and stored at −80°C until analysis. Plasma samples from the EDTA tubes were used to measure the concentrations of βC, Vit. E, and vitamin A (Vit. A). Plasma collected from the NaF tubes was used to measure blood glucose. Serum samples were used to measure the concentrations of NEFA, IGF-1, total antioxidant status (TAS), and derivatives of reactive oxygen metabolites (dROM) to assess OS. One EDTA tube was not centrifuged; whole-blood samples were aliquoted and stored at −80°C and used to measure glutathione peroxidase (GPx) and reactive oxygen species (ROS) as described below.

      Aspiration of FF and Samples of GC

      Ovaries were palpated transrectally, and the cow was gently positioned for transvaginal ultrasound visualization (DP-50Vet, PS-3B0015EA Ultrasonic Diagnostic Imaging System, Mindray, 65C15EA Transducer). Both ovaries were scanned and the preovulatory follicle (the follicle with the largest diameter >0.8 cm) was identified and its diameter was recorded. The preovulatory follicle was subsequently punctured by an adapted oocyte pick-up (OPU) technique as described by
      • Leroy J.L.
      • Vanholder T.
      • Mateusen B.
      • Christophe A.
      • Opsomer G.
      • de Kruif A.
      • Genicot G.
      • Van Soom A.
      Non-esterified fatty acids in follicular fluid of dairy cows and their effect on developmental capacity of bovine oocytes in vitro.
      . Follicular fluid was aspirated and collected in separate microcentrifuge tubes without any flushing. With the aspiration needle still in place, the collapsed follicle was flushed twice with 1 to 2 mL of warm physiological saline (37°C) containing 100 IU/mL heparin and 4 mg/mL BSA to collect the GC, which were also transferred to microcentrifuge tubes. The FF samples were centrifuged for 5 min at 971 × g at RT, and the FF supernatant was transferred to a cryotube and transported on ice to the laboratory where it was stored at −80°C. The flushing medium with GC was transferred to a searching dish and checked under a stereomicroscope to remove cumulus–oocyte complexes. The flushing medium was then centrifuged for 5 min at 971 × g at RT, and the supernatant was removed. The GC pellets were treated with RBC lysis buffer (10% 3.7 g/L EDTA, 80.2 g/L NH4Cl, 8.4 g/L NaHCO3 and 90% sterile H2O) followed by cold PBS to remove any red blood cell contamination. The GC samples were centrifuged again for 5 min at 971 × g, and the cell pellets were snap-frozen in liquid nitrogen. All procedures described here were performed on the farm immediately after sample collection. The GC samples were snap-frozen within 30 min after OPU, transported to the laboratory on dry ice, and stored at −80°C until further analysis. Follicular fluid samples were used to analyze NEFA, 17β-estradiol (E2), progesterone (P4), βC, Vit. E, Vit. A, TAS, and dROM. The GC were used for total RNA extraction for RNA sequencing (RNaseq) analysis, which was the primary outcome measure in this study.

      Blood and FF Analyses

      Blood and FF analyses were performed in a commercial accredited laboratory (Algemeen Medisch Labo, AML, Antwerp, Belgium) using automated or semi-automated facilities. Plasma and FF βC were photometrically analyzed with a laboratory spectrophotometric method [with UV detection at 450 nm, coefficient of variance (CV) = 4.5%, DR3900, Hach Lange].
      Plasma and FF Vit. E were analyzed by liquid-liquid extraction and HPLC (with UV detection at 292 nm, CV = 9.5%; 1260 Infinity, Agilent Technologies). Plasma and FF Vit A were analyzed by liquid-liquid extraction and HPLC (with UV detection at 325 nm, CV = 3.5%; UltiMate 3000, Thermo Fisher Scientific).
      Serum and FF NEFA were colorimetrically and enzymatically determined (Randox Laboratories) with a Gallery Plus Automated Photometric Analyzer (with detection at 550 nm and 340 nm, respectively, CV = 5%; Thermo Fisher Scientific).
      Plasma GPx was routinely analyzed according to
      • Paglia D.E.
      • Valentine W.N.
      Studies on the quantitative and qualitative characterization of erythrocyte glutathione peroxidase.
      using a commercially available GPx kit (Randox Laboratories) and spectrophotometrically detected (Cobas 8000; within and between CV = 1.8% and 6.8%, respectively). A radioimmunoassay kit was used to measure serum IGF-1 concentrations (Diasource; within and between CV = 9.8% and 10.4%, respectively).
      Estradiol (within and between CV = 2.1% and 2.2%, respectively) and P4 (within and between CV = 2.0% and 3.8%, respectively) were analyzed using a chemiluminescent particle immunoassay (i2000 analyzers, Abbott Diagnostics).
      In addition, OSi was calculated as ROS, determined calorimetrically using a dROM test, within and between CV = 6.27% and 2.46%, respectively), divided by the serum antioxidant capacity (TAS, determined colorimetrically using an OXY-Adsorbent test; within and between CV = 2.88% and 6.47%) as described by
      • Abuelo A.
      • Hernández J.
      • Benedito J.L.
      • Castillo C.
      Oxidative stress index (OSi) as a new tool to assess redox status in dairy cattle during the transition period.
      .

      GC Total RNA Extraction and cDNA Synthesis

      Isolation of total RNA was carried out from the GC samples using PicoPure RNA Isolation Kit (Thermo Fisher) according to the manufacturer's instructions. An on-column DNase treatment (RNase-Free DNase Set; Qiagen) was implemented to eliminate genomic DNA contamination. The isolated RNA samples were stored at −80°C until RNA sequencing. An aliquot was used to determine total RNA concentration and integrity using an RNA 6000 Nano kit (Agilent Genomics) on an Agilent 2100 bioanalyzer (Agilent Genomics). Samples with an RNA integrity value ≥7 and an RNA concentration of ≥2 ng/µL were considered qualified. A selection of 27 qualified RNA samples (individual samples without pooling) representative of different metabolic profiles (based on blood concentration of NEFA and antioxidants at wk 2 pp, as explained below) were selected and submitted to RNaseq analysis. Therefore, each cow served as an experimental unit.

      RNA Sequencing

      Granulosa cell transcriptomics were analyzed by BGI Tech Global, BGI Genomics Co. Ltd. Briefly, poly-A-containing mRNA molecules were purified from total RNA using oligo(dT)-attached magnetic beads and fragmented into small pieces using divalent cations under elevated temperature. First-strand cDNA was synthesized using random hexamer-primed reverse transcription, followed by a second-strand cDNA synthesis using DNA polymerase I and RNase H. The synthesized cDNA was subjected to end-repair and 3′ adenylation. Adapters were ligated to the ends of 3′-adenylated cDNA fragments. cDNA fragments with adapters from previous steps were amplified by PCR. The resulting PCR products were purified with Agencourt AMPure XP Beads (Beckman Coulter) and dissolved in buffer EB. Double-stranded PCR products were heat-denatured and circularized by the splint oligo sequence. Single-stranded circular DNAs (ssCir DNA) were used for PE100 strand-specific library construction and validation on the Bioanalyzer 2100 (Agilent Genomics). The library was amplified with phi29, and DNA nanoballs were generated with ssCir DNA by rolling circle replication to intensify the fluorescent signals during the sequencing process. The DNA nanoballs were loaded into the patterned nanoarray, and pair-end reads of 100 bp were read on the BGISEQ-500 (BGI Genomics Institute) platform for subsequent data analysis.

      RNaseq Data Analysis

      The RNaseq data analysis was performed in Galaxy (https://usegalaxy.org). Quality control was performed by FastQC (version 0.71;
      • Andrews S.
      FastQC: A Quality Control Tool for High Throughput Sequence Data.
      ). Paired reads were mapped to the Bos taurus reference genome (UMD 3.1, NCBI project 32899, GenBank GCA_000003055.3) using HISAT2 (version 2.1.0;
      • Kim D.
      • Paggi J.M.
      • Park C.
      • Bennett C.
      • Salzberg S.L.
      Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype.
      ), after which the counts per gene model was performed with FeatureCounts (version 1.6.2;
      • Liao Y.
      • Smyth G.K.
      • Shi W.
      FeatureCounts: An efficient general purpose program for assigning sequence reads to genomic features.
      ). Expression values were normalized and global and pairwise statistics were performed using DESeq2 (version 1.18.1;
      • Love M.I.
      • Huber W.
      • Anders S.
      Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2.
      ) to determine the differentially expressed genes (DEG) using 5% false discovery rate–corrected P-values. These pairwise comparisons were performed between subgroups of cows separated based on cut-off values for each blood and FF parameter to study the association between each parameter and the GC transcriptomic profile (see further details in the Analytical Approach section below). The number of DEG reflects both the transcriptomic homogeneity within a subgroup of cows and the distinct differences compared with the other subgroup in the same comparison. Canonical pathways of the DEG were annotated using the HyperGTest in Bioconductor in R statistics (https://www.r-project.org/). A cut-off value for the DEG with fold-changes >1.5 and adjusted P-values ≤0.05 (5% FDR) was applied. Heatmaps were generated with MeV software (https://mev.tm4.org/), applying average linkage and the Spearman rank correlation method for clustering and distance measurements.

      Analytical Approach

      In this study, the correlations between blood (wk 2 and wk 8) and FF (wk 8) metabolic and antioxidant parameters were first calculated (Pearson correlation coefficient with a 2-tailed test of significance) to understand possible underlying systemic factors that may influence the follicular microenvironment and GC functions. This was followed by an intermediate analysis in which the associations between blood and FF factors with the GC transcriptomic profile were examined. For that, median and quartile values of each blood and FF parameter were calculated (Supplemental Table S1; https://data.mendeley.com/datasets/s3c28xj492/2;
      • Marei W.
      JDS 21928 supplemental data. Version 2.
      ). Cows were then split into 2 groups based on the median value (high vs. low) to include more cows in each comparison or were split based on quartile values (highest vs. lowest; Q4 vs. Q1) to include fewer cows (at least 3 per subgroup) with more distinct metabolic and antioxidant differences. We compared the transcriptomic profile of the GC samples collected from these subgroups of cows using DESeq2, and the number of DEG was recorded in each comparison. In all studied parameters, we found that the Q4 vs. Q1 comparison gave the maximum contrast and yielded more DEG than the corresponding median split, showing a greater influence of the distinct levels of the parameter being assessed compared with the influence of the number of cows included in each comparison. Therefore, mainly the Q4 vs. Q1 comparisons are reported here, unless otherwise stated. These intermediate comparisons revealed that differences due to variation in E2:P4 ratios had a strong effect on GC gene expression patterns, which could mask or interact the effect of other metabolic and antioxidant factors (Supplemental Index S1; https://data.mendeley.com/datasets/s3c28xj492/2;
      • Marei W.
      JDS 21928 supplemental data. Version 2.
      ). These results suggested that the wide variation in E2:P4 ratio was due to the occurrence of an LH surge in some cows before sample collection. Based on the E2:P4 ratio and the expression pattern of specific LH-surge responsive genes (as described in
      • Gilbert I.
      • Robert C.
      • Dieleman S.
      • Blondin P.
      • Sirard M.A.
      Transcriptional effect of the LH surge in bovine granulosa cells during the peri-ovulation period.
      , we categorized the samples into pre-LH surge (n = 16) and post-LH surge (n = 11) cohorts. For more details on sample categorization and on the metabolic and antioxidant profile differences between the cows in both cohorts, please see Supplemental Index S2 (https://data.mendeley.com/datasets/s3c28xj492/2;
      • Marei W.
      JDS 21928 supplemental data. Version 2.
      ). The DESeq2 comparisons were then repeated after excluding samples in the post-LH surge cohort; that is, focusing only on the potential effect of the metabolic and antioxidant factors on the GC transcriptomic patterns within the pre-LH surge cohort (where the GC transcriptome is not yet influenced by the LH surge). The E2:P4 effects on the GC transcriptome within the pre-LH cohort were very small, and the predominant effects were attributed to differences in βC and Vit. E, as well as NEFA. In all comparisons listed so far, the effect of each factor was studied independently of other factors. To make these comparisons more clinically relevant, we also studied the potential effect of NEFA on the GC transcriptome only in cows with low antioxidant status (i.e., only in cows with low serum concentrations of βC or Vit. E at wk 2). We also examined the effects of different antioxidants (high vs. low βC and Vit. E) in severe NEB cows (i.e., only in cows with high serum concentrations of NEFA at wk 2). Representative functional annotation analyses were performed to provide a functional biological insight of the DEG lists.

      RESULTS AND DISCUSSION

      Correlation Between FF Parameters (at wk 8) and Blood Parameters (at wk 2 and 8)

      All correlations are depicted in Figure 2. Follicular fluid βC concentrations were positively correlated with blood βC at the time of OPU in wk 8 (P = 0.001). This is in line with the observations of
      • De Bie J.
      • Langbeen A.
      • Verlaet A.A.J.
      • Florizoone F.
      • Immig I.
      • Hermans N.
      • Fransen E.
      • Bols P.E.J.
      • Leroy J.
      The effect of a negative energy balance status on beta-carotene availability in serum and follicular fluid of nonlactating dairy cows.
      . Moreover, we found that both blood and FF βC concentrations at wk 8 were positively correlated with blood βC at wk 2 (P < 0.05). The FF Vit. A concentrations were positively correlated with blood Vit. A at wk 8 (P = 0.006) but not at wk 2 (P = 0.607). In contrast, Vit. E concentrations in FF and blood were not correlated at either time point (P > 0.05).
      Figure thumbnail gr2
      Figure 2Correlation between follicular fluid (FF) collected at wk 8 (w8) postpartum and blood parameters (serum) at wk 8 and wk 2 (w2). Pearson correlation values are indicated by a color range, where red is positive and blue is negative. Correlations labeled with an asterisk (*) are significant at P < 0.05, and those labeled with $ tend to be significant at P > 0.05 and < 0.1. bC = β-carotene; VitE = vitamin E; VitA = vitamin A; NEFA = nonesterified fatty acids; E2 = estradiol; P4 = progesterone; DF = dominant follicle; GPx = glutathione peroxidase; dROM = derivatives of reactive oxygen metabolites; OXYads = OXY adsorbent as a measure of total antioxidant capacity; OSi = oxidative stress index; THI = temperature-humidity index.
      It is worth noting that βC and Vit. A concentrations were not correlated in blood and FF despite the fact that βC is the precursor for Vit. A. Cattle can convert βC to Vit. A in the small intestine, which is then mostly stored in the liver (70–90%) and, to a lesser extent, in fat or distributed to other organs (
      • Green A.S.
      • Fascetti A.J.
      Meeting the vitamin A requirement: The efficacy and importance of β-carotene in animal species.
      ). Therefore, circulating concentrations of βC absorbed from the diet that escape conversion to Vit. A might not necessarily correlate with circulating Vit. A concentrations. Additionally, GC can actively take up and convert βC into Vit. A (
      • Brown J.A.
      • Eberhardt D.M.
      • Schrick F.N.
      • Roberts M.P.
      • Godkin J.D.
      Expression of retinol-binding protein and cellular retinol-binding protein in the bovine ovary.
      ).
      In contrast, wk 8 FF and blood NEFA concentrations were not correlated with wk 2 blood NEFA concentrations (P > 0.8). This indicates that the association between wk 2 NEFA and the GC transcriptomic profile described later are not mediated by persistent elevation of NEFA in these cows. In addition, FF NEFA concentrations were not correlated with blood NEFA concentration at the time of sample collection at wk 8. The concentration of NEFA in the FF might be influenced by selective uptake of saturated free fatty acids into growing follicles as well as local cellular metabolic activities (
      • Argov N.
      • Moallem U.
      • Sklan D.
      Lipid transport in the developing bovine follicle: messenger RNA expression increases for selective uptake receptors and decreases for endocytosis receptors.
      ;
      • Wonnacott K.E.
      • Kwong W.Y.
      • Hughes J.
      • Salter A.M.
      • Lea R.G.
      • Garnsworthy P.C.
      • Sinclair K.D.
      Dietary omega-3 and -6 polyunsaturated fatty acids affect the composition and development of sheep granulosa cells, oocytes and embryos.
      ), which may result in reducing NEFA concentration in the FF. Follicular cells have the capacity to internalize NEFA and store them in the form of lipid droplets, which can then be used as a source of energy via fatty acid β-oxidation (
      • Aardema H.
      • van Tol H.T.A.
      • Vos P.
      An overview on how cumulus cells interact with the oocyte in a condition with elevated NEFA levels in dairy cows.
      ). This is dependent on cell viability and metabolic activity (
      • Yang R.
      • Le G.
      • Li A.
      • Zheng J.
      • Shi Y.
      Effect of antioxidant capacity on blood lipid metabolism and lipoprotein lipase activity of rats fed a high-fat diet.
      ;
      • Chen L.
      • Wang C.
      • Huang S.
      • Gong B.
      • Yu J.
      • Shi Q.
      • Chen G.
      Effects of individual and multiple fatty acids (palmitate, oleate and docosahaexenoic acid) on cell viability and lipid metabolism in LO2 human liver cells.
      ), which may in turn be affected by the antioxidant status within the FF. Indeed, we found that NEFA concentrations in the FF (but not in blood) tended to be negatively correlated with blood βC (P = 0.08) and Vit. E (P = 0.067) at both wk 2 and wk 8. This indicates that high levels of antioxidants during the postpartum period might not only increase the antioxidant capacity within the ovarian follicle but can also be associated with reduced local metabolic stress by lowering FF NEFA concentrations. Apparently, this effect takes place particularly at the ovarian follicle level because blood NEFA were not correlated with blood βC and Vit. E at either wk 8 or wk 2. Follicular fluid NEFA also had a strong negative correlation with FF Vit. A (P = 0.017), which could reflect the (enhanced) GC viability and ability to utilize the circulating βC.
      Follicular fluid E2 concentrations tended to be negatively correlated with wk 8 blood NEFA (P = 0.09). This may suggest an altered steroidogenic activity of follicular cells in cows with higher blood NEFA levels at the time of sample collection. However, in contrast, we noticed that both E2 (P = 0.053) and E2:P4 ratio (P = 0.049) in FF were positively correlated with blood NEFA at wk 2. As shown by further analysis of the data, this might be mediated through a long-term effect on follicular growth dynamics (as discussed in Supplemental Index S2).
      It is interesting to note that wk 8 blood glucose, IGF-1, and GPx did not correlate with any of the measured FF parameters. Other correlations were noted between some wk 2 and wk 8 blood parameters (Figure 2), which are beyond the focus of the current study or will be mentioned later in the discussion.

      Association Between the Metabolic and Antioxidant Profiles with the GC Transcriptomic Profile

      We compared subgroups of cows in the highest and lowest (Q4, Q1) quartiles of each metabolic and antioxidant parameter measured, using DESeq2 and focusing only on the pre-LH surge cohort. The cut-off values used for grouping are shown in Supplemental Table S1, and the average concentration of each parameter and number of cows included in each subgroup are shown in Supplemental Table S2 (https://data.mendeley.com/datasets/s3c28xj492/2;
      • Marei W.
      JDS 21928 supplemental data. Version 2.
      ). We found that the antioxidant and metabolic blood profile at wk 2 had a relatively greater effect on the GC transcriptomic profile based on the number of DEG compared with wk 8 blood and FF comparisons (Table 1). Blood wk 2 βC and Vit. E comparisons were associated with 192 and 157 DEG, respectively. The highest number of DEG was obtained when comparing cows in Q4 vs. Q1 of both βC and Vit. E (βC + Vit. E) at wk 2 (341 DEG). Blood wk 2 NEFA concentrations were associated with 64 DEG. A heatmap illustrating the variation in GC gene expression corresponding to different levels of βC, Vit. E, and NEFA is shown in Supplemental Figure S1 (https://data.mendeley.com/datasets/s3c28xj492/2;
      • Marei W.
      JDS 21928 supplemental data. Version 2.
      ). Other studied factors (Vit. A, glucose, IGF-1, GPx) had very subtle or no effects (≤10 DEG). Oxidative stress index, which is calculated as a ratio between pro-oxidant and antioxidant concentrations, has been shown to give a good indication of the OS of the cow compared with using separate measurements (
      • Abuelo A.
      • Hernández J.
      • Benedito J.L.
      • Castillo C.
      Oxidative stress index (OSi) as a new tool to assess redox status in dairy cattle during the transition period.
      ). However, using OSi as a splitting factor in the present study could not be linked with any transcriptomic effects on the GC (≤3 DEG).
      Table 1The number of differentially expressed genes
      Data are shown as number of differentially expressed genes (DEG), where ↑ = upregulated and ↓ = downregulated.
      in granuloma cells of pre-LH surge preovulatory follicles (n = 16) that are associated with blood and follicular fluid (FF) metabolic and antioxidant parameters measured at wk 2 and wk 8 postpartum
      Q4 vs. Q1
      Q4 vs. Q1 = highest and lowest quartiles; βC = β-carotene; Vit. E = vitamin E; Vit. A = vitamin A; NEFA = nonesterified fatty acids; GPx = glutathione peroxidase; OSi = oxidative stress index; E2 = estradiol; P4 = progesterone.
      Blood (wk 2)Blood (wk 8)FF (wk 8)
      No. of DEGNo. up- or downregulatedNo. of DEGNo. up- or downregulatedNo. of DEGNo. up- or downregulated
      βC1929 ↑, 183 ↓17014 ↑, 156 ↓18975 ↑, 114 ↓
      Vit. E157118 ↑, 39 ↓00
      βC + Vit. E341162 ↑, 179 ↓
      The functional annotation of blood wk 2 βC + Vit. E and NEFA-associated DEG is described hereafter as a representation of the functional impact of antioxidant and metabolic stress (Figure 4A and 4B).
      100 ↑, 10 ↓0
      Vit. A000
      NEFA6416 ↑, 48 ↓
      The functional annotation of blood wk 2 βC + Vit. E and NEFA-associated DEG is described hereafter as a representation of the functional impact of antioxidant and metabolic stress (Figure 4A and 4B).
      00
      Glucose061 ↑, 5 ↓
      IGF-1101 ↑, 9 ↓11 ↑, 0 ↓
      GPx022 ↑, 0 ↓
      OSi032 ↑, 1 ↓
      E2:P4 ratio5856 ↑, 2 ↓
      1 Data are shown as number of differentially expressed genes (DEG), where ↑ = upregulated and ↓ = downregulated.
      2 Q4 vs. Q1 = highest and lowest quartiles; βC = β-carotene; Vit. E = vitamin E; Vit. A = vitamin A; NEFA = nonesterified fatty acids; GPx = glutathione peroxidase; OSi = oxidative stress index; E2 = estradiol; P4 = progesterone.
      3 The functional annotation of blood wk 2 βC + Vit. E and NEFA-associated DEG is described hereafter as a representation of the functional impact of antioxidant and metabolic stress (Figure 4A and 4B).
      In wk 8 comparisons, βC concentration in blood and FF was the only parameter that could be associated with notable GC transcriptomic changes (170 and 189 DEG respectively; Table 1). We found several common DEG between different comparisons, particularly between blood wk 2 βC + Vit. E, wk 2 βC, wk 8 βC, and FF wk 8 βC (Figure 3). Interestingly, only one of these genes was in common with the DEG of the wk 2 NEFA comparison. This suggests that the effect of NEFA on GC gene expression patterns is completely distinct from that of antioxidant status (Figure 3). For the full list of DEGs of each comparison, see Supplemental Data File S1: GC_DEGs.xlsx (https://data.mendeley.com/datasets/s3c28xj492/2;
      • Marei W.
      JDS 21928 supplemental data. Version 2.
      ).
      Figure thumbnail gr3
      Figure 3Common differentially expressed genes (DEG) between the comparisons shown in that yielded >10 DEGs. The numbers of common genes are shown in circles on the connecting lines. w2 = wk 2; w8 = wk 8; FF = follicular fluid; bl. = blood; βC = β-carotene; Vit. E = vitamin E; NEFA = nonesterified fatty acids.
      Functional annotation of the DEG strongly suggested that the pre-LH surge follicles of cows with high wk 2 βC + Vit. E (Q4) were of better quality than those with low wk 2 βC + Vit. E (Q1; Figure 4A); these follicles exhibited less ubiquitin-dependent protein catabolism, higher RNA biosynthesis and splicing, and increased expression of genes involved in response to LH and estrogen (Figure 4A). Phospholipase C activating G-protein coupled receptor signaling was also upregulated in the GC of these cows, a feature that has been associated with a higher steroidogenic activity and lower apoptosis in cultured granulosa cells (
      • Chen H.
      • Yang Y.
      • Wang Y.
      • He Y.
      • Duan J.
      • Cheng J.
      • Li Q.
      The effects of phospholipase C on oestradiol and progesterone secretion in porcine granulosa cells cultured in vitro.
      ,
      • Chen H.
      • Yang Y.
      • Wang Y.
      • Li Y.
      • He Y.
      • Duan J.
      • Xu D.
      • Pei Y.
      • Cheng J.
      • Yang L.
      • Hua R.
      • Li X.
      • Wang J.
      • Jiang X.
      • He H.
      • Wu L.
      • Liu D.
      • Li Q.
      Phospholipase C inhibits apoptosis of porcine primary granulosa cells cultured in vitro.
      ). The increased oxidoreductase activity, MAPK cascade, and pathways related to meiosis activation in oocyte also suggest that wk 8 preovulatory follicles in the cows with high wk 2 βC + Vit. E might have a higher capacity to support oocyte quality and enhance developmental competence (
      • Liang C.-G.
      • Su Y.-Q.
      • Fan H.-Y.
      • Schatten H.
      • Sun Q.-Y.
      Mechanisms regulating oocyte meiotic resumption: Roles of mitogen-activated protein kinase.
      ). This is further confirmed by the fact that the following pathways were all downregulated: genes related to acute inflammation, negative regulation of nuclear factor-κB transcription factor activity, oxidation-dependent catabolic processes, sphingomyelin biosynthesis, mitochondrial fragmentation, and lipophagy. In other words, follicles that start to grow in the presence of high antioxidant concentrations (βC + Vit. E) in the blood at wk 2 pp seem to exhibit less inflammatory responses and less cellular stress and catabolism.
      Figure thumbnail gr4
      Figure 4Functional annotation of differentially expressed genes (DEG) detected in the comparisons based on (A) blood β-carotene (βC) + vitamin E (Vit. E) concentration at wk 2 (w2), or (B) blood nonesterified fatty acids (NEFA) concentration at wk 2. Positive regulation of a pathway is labeled with (+) and a negative regulation is shown as (−). Q4 vs. Q1 = highest and lowest quartiles.
      In contrast, the DEG detected in GC samples from cows that had high blood wk 2 NEFA concentrations (0.86 ± 0.16 mM in Q4 vs. 0.30 ± 0.08 mM in Q1; Figure 4B) are related to response to ketones, response to lipid and lipid catabolism, immune response, and response to ROS (Figure 4B). These changes clearly indicate a high cellular stress response that is still detectable in the GC of the preovulatory follicle at the time of breeding. It is important to note that the average blood NEFA concentration at the time of sample collection at wk 8 had decreased in these Q4 cows from 0.86 ± 0.16 to 0.32 ± 0.13 mM, which rules out any direct NEFA effect at wk 8, because wk 2 Q1 cows had a similar level of blood NEFA (0.40 ± 0.23 mM) at wk 8. High wk 2 NEFA was also associated with transcriptomic changes that indicate a higher rate of cellular respiration and increased mitochondrial membrane potential and electron transport chain activity, which is known to increase ROS production and increase OS levels (
      • Aon M.A.
      • Bhatt N.
      • Cortassa S.C.
      Mitochondrial and cellular mechanisms for managing lipid excess.
      ). This may explain the detection of genes related to positive regulation of intrinsic apoptotic signaling pathway and cellular oxidant detoxification. Genes related to fatty acid β-oxidation and nitrogen compound metabolic processes were downregulated, which could be a compensatory response in the mitochondria to control cellular metabolism (
      • Runkel E.D.
      • Baumeister R.
      • Schulze E.
      Mitochondrial stress: Balancing friend and foe.
      ). DNA replication also appeared to be negatively affected in these follicles. In other words, follicles that start to grow in the presence of high NEFA in the blood at wk 2 pp seem to suffer from a persistent lipotoxic effect and altered cellular metabolic activity.
      Importantly, although 58 DEG were detected within the pre-LH cohort in association with FF E2:P2 ratio, none of these genes were in common with those detected in the antioxidant and NEFA comparisons at wk 2 or at wk 8. Hence, the differences due to E2:P4 ratio that persist in the pre-LH sample cohort do not confound the antioxidant and NEFA effects described above.

      Effect of High NEFA Concentration at wk 2 on the GC Transcriptomic Profile in Cows with Low Antioxidant Profile

      It has been shown previously that cows with high BCS loss during the transition period, and thus high NEFA concentrations at wk 2, have higher plasma pro-oxidant radicals and lower levels of antioxidants in the early postpartum period (
      • Bernabucci U.
      • Ronchi B.
      • Lacetera N.
      • Nardone A.
      Influence of body condition score on relationships between metabolic status and oxidative stress in periparturient dairy cows.
      ). Deficiency in antioxidants might lead to a higher sensitivity to metabolic and OS (
      • Bernabucci U.
      • Ronchi B.
      • Lacetera N.
      • Nardone A.
      Influence of body condition score on relationships between metabolic status and oxidative stress in periparturient dairy cows.
      ); however, the effect of antioxidant deficiency on the influence of high NEFA on GC gene expression has not been previously investigated. Vitamin E and βC concentrations were particularly low in cows with high NEFA concentrations during the early postpartum period (severe NEB;
      • Mary A.E.P.
      • Artavia Mora J.I.
      • Ronda Borzone P.A.
      • Richards S.E.
      • Kies A.K.
      Vitamin E and beta-carotene status of dairy cows: A survey of plasma levels and supplementation practices.
      ). To provide a more clinically relevant insight, we examined the potential effect of NEFA on the GC transcriptome within the cows with low antioxidant status (i.e., only in cows with low serum concentrations of βC or Vit. E) at wk 2 and at wk 8 (below median values; Table 2). We hypothesized that the NEFA concentration at wk 2 might be associated with a greater effect on the growing follicles and their GC in cows with antioxidant deficiencies. These comparisons were based on the median values (high vs. low) of each parameter as a cut-off to allow a sufficient number of cows in each subgroup (at least 3). The mean concentrations of NEFA and antioxidants in the compared subgroups were pathophysiologically and clinically relevant (see Supplemental Table S3; https://data.mendeley.com/datasets/s3c28xj492/2;
      • Marei W.
      JDS 21928 supplemental data. Version 2.
      ).
      Table 2The association between blood nonesterified fatty acid (NEFA) concentration (high vs. low based on median values) at wk 2 and the granulosa cell (GC) transcriptomic profile in cows with low blood antioxidant concentrations at wk 2 and wk 8
      Cow selection criteria
      βC = β-carotene; Vit. E = vitamin E.
      High vs. low NEFA at wk 2
      Data are shown as number of differentially expressed genes (DEG), where ↑ = upregulated and ↓ = downregulated.
      No. of DEGNo. up- or downregulated
      Wk 2
       Low βC638 ↑, 55 ↓
       Low Vit. E108 ↑, 2 ↓
       Low βC + Vit. E11 ↑, 0 ↓
      Wk 8
       Low βC6346 ↑, 17 ↓
       Low Vit. E148141 ↑, 7 ↓
       Low βC + Vit. E10687 ↑, 19 ↓
      The functional annotation of the DEG is described hereafter as a representative example of functional NEFA effects on GC gene expression in cows deficient in antioxidants (Figure 6).
      1 βC = β-carotene; Vit. E = vitamin E.
      2 Data are shown as number of differentially expressed genes (DEG), where ↑ = upregulated and ↓ = downregulated.
      3 The functional annotation of the DEG is described hereafter as a representative example of functional NEFA effects on GC gene expression in cows deficient in antioxidants (Figure 6).
      We found that the number of DEG associated with elevated wk 2 NEFA concentrations (high vs. low) was generally higher in cows with antioxidant deficiency at wk 8 (Table 2) compared with the effects of NEFA in cows with antioxidant deficiency at wk 2 (Table 2). Only a few of these DEG were in common with the 64 DEG initially detected in the wk 2 Q4 vs. Q1 NEFA comparison (regardless of antioxidant status; Table 1, Figure 5); nevertheless, they belonged functionally to similar pathways. As a representative example, the functional annotation of the DEG detected in the high versus low wk 2 NEFA comparison with low βC + Vit. E at wk 8 (106 DEG) is shown in Figure 6. In addition to genes involved in response to ketone and lipid metabolic processes detected in Figure 4B, we observed more oxidative stress–related pathways (H2O2 and superoxide radical–dependent responses), apoptosis, and catabolic processes (Figure 6) that were not detected in the initial NEFA comparison regardless of the antioxidants (Figure 4B). The effect of NEFA on the pathways related to mitochondrial activity and cellular metabolism in the GC also appeared to be more prominent in low antioxidant cows. Moreover, we detected downregulation of genes related to early endosomes and endosome transport to lysosomes, suggesting disruption of autophagy, which is a common consequence of oxidative damage in metabolic disorders (
      • Chiumia D.
      • Chagunda M.G.
      • Macrae A.I.
      • Roberts D.J.
      Predisposing factors for involuntary culling in Holstein-Friesian dairy cows.
      ). Altogether, these transcriptomic changes indicate high cellular stress levels and reduced cell survival in the GC of the preovulatory follicle in cows with high NEFA at wk 2. This carryover effect was stronger in cows that were deficient in antioxidants at wk 8.
      Figure thumbnail gr5
      Figure 5Common differentially expressed genes (DEG) between the comparisons shown in [within cows with low antioxidant concentrations at wk 2 or wk 8 (w8)] and the blood wk 2 (w2) nonesterified fatty acids (NEFA) comparison shown in (regardless of the antioxidant concentration; 64 DEG). The numbers of common genes are shown in circles on the connecting lines. βC = β-carotene; Vit. E = vitamin E; Q4 vs. Q1 = highest and lowest quartiles.
      Figure thumbnail gr6
      Figure 6Functional annotation of differentially expressed genes (DEG) detected in the high versus low wk 2 (w2) blood nonesterified fatty acids (NEFA) comparison in cows with low blood β-carotene (βC) + vitamin E (Vit. E) concentration at wk 8 (106 DEG, ). Positive regulation of a pathway is labeled with (+) and a negative regulation is shown as (−). w8 = wk 8.

      Potential Benefit of Blood Antioxidants on GC Transcriptomic Profile Within Cows with High NEFA at Wk 2

      It was surprising to see that the blood and FF antioxidant-based comparisons at wk 8 (Vit. E and βC + Vit. E) yielded a much smaller number of DEG compared with wk 2 antioxidant comparisons, regardless of blood NEFA concentrations (Table 1). To provide a more clinically relevant insight, we repeated these antioxidant comparisons but only within the cows with high NEFA concentration at wk 2 (severe NEB cows; n = 10). This was to examine whether optimal blood antioxidant concentration (at either wk 2 or wk 8) could positively influence the GC transcriptomic pattern of wk 8 preovulatory follicles in cows exposed to severe NEB during transition. For that, DESeq2 comparisons were performed between subgroups with high versus low blood βC, Vit. E, or both in blood at wk 2 and wk 8 (based on median values) only in the cows with high NEFA concentrations at wk 2 (above median concentration; 0.78 ± 0.19 mM; Table 3). All mean ± standard deviation values of these parameters in each subgroup are listed in Supplemental Table S4 (https://data.mendeley.com/datasets/s3c28xj492/2;
      • Marei W.
      JDS 21928 supplemental data. Version 2.
      ).
      Table 3The effect of high versus low levels of antioxidants in blood (at wk 2 and wk 8) on the granulosa cell (GC) transcriptomic profile only in cows that exhibited elevated wk 2 blood nonesterified fatty acids (NEFA)
      Cow selection criterionHigh vs. low antioxidant comparison
      βC = β-carotene; Vit. E = vitamin E.
      DEG
      Data are shown as number of differentially expressed genes (DEG), where ↑ = upregulated and ↓ = downregulated.
      No.No. up- or downregulated
      High blood NEFA at wk 2 (above median; 0.78 ± 0.19 mM, n = 10)Wk 2
       βC106 ↑, 4 ↓
       Vit. E2221 ↑, 1 ↓
       βC + Vit. E33 ↑, 0 ↓
      Wk 8
       βC18524 ↑, 161 ↓
       Vit. E11155 ↑, 56 ↓
       βC + Vit. E19448 ↑, 146 ↓
      The functional annotation of the DEG is described hereafter as a representative example of functional antioxidant effects on GC gene expression in negative energy balance cows with high NEFA at wk 2 (Figure 8).
      1 βC = β-carotene; Vit. E = vitamin E.
      2 Data are shown as number of differentially expressed genes (DEG), where ↑ = upregulated and ↓ = downregulated.
      3 The functional annotation of the DEG is described hereafter as a representative example of functional antioxidant effects on GC gene expression in negative energy balance cows with high NEFA at wk 2 (Figure 8).
      Blood antioxidant concentrations at wk 2 in the presence of high NEFA had relatively lower effects (fewer DEG; Table 3) compared with those associated with wk 2 antioxidant regardless of NEFA concentrations (Table 1). This illustrates how severe NEB may diminish the long-term beneficial effects of antioxidants. In contrast, high wk 8 antioxidant concentrations in cows with high wk 2 NEFA were linked to a greater effect (more DEG; Table 3) compared with wk 8 antioxidant effects, regardless of NEFA concentration (Table 1). This was true for βC, Vit. E, and βC + Vit. E. This is in line with observations in the previous section, focusing on wk 2 NEFA effects in antioxidant deficient cows, which were more prominent in cows with antioxidant deficiency at wk 8.
      Many of the DEG influenced by wk 8 blood βC + Vit. E or βC only in cows with high wk 2 NEFA (Table 3) were in common with the DEG detected in the wk 2 blood βC + Vit. E and wk 8 βC comparisons (shown in Table 1 regardless of NEFA comparisons; Figure 7). This illustrates that most of these DEG are specifically regulated in the GC under the influence of the antioxidant effect, and that this regulation may already be initiated by antioxidants at wk 2 but attenuated by high wk 2 NEFA. These genes can be seen as a mechanism by which antioxidants might protect (at wk 2) or rescue (at wk 8) GC functions from the effect of NEFA (for the full list of DEG, see supplemental data;
      • Marei W.
      JDS 21928 supplemental data. Version 2.
      ). Altogether, these results suggest a strong effect of antioxidant status (especially due to βC concentration) on GC functions in the preovulatory follicle at wk 8, even in cows that suffered from severe NEB and high lipid mobilization at wk 2.
      Figure thumbnail gr7
      Figure 7Common differentially expressed genes (DEG) between the comparisons shown in [within cows with high blood nonesterified fatty acids (NEFA) concentration at wk 2 (w2)] and blood wk 2 and wk 8 (w8) antioxidant comparisons shown in (regardless of NEFA concentration). The numbers of common genes are shown in circles on the connecting lines. βC = β-carotene; Vit. E = vitamin E.
      As a representative example, functional annotation of the 194 DEG in the high versus low βC + Vit. E comparison in high wk 2 NEFA cows is shown in Figure 8. The enriched upregulated pathways are related to activation of meiosis, MAPK signaling, IGF receptor signaling, regulation of EGF receptor signaling, fertilization and acrosome reaction, and embryo development, which are indicators of a better supportive capacity of these preovulatory follicles. These follicles also appear to have more active RNA biosynthetic process, amino acid and carbohydrate metabolism, and mitotic activity (cell proliferation), which also indicate better cell viability and follicle quality. The downregulated pathways suggest lower levels of inflammation and cellular stress because genes related to mitochondrial fragmentation, DNA breakdown, sphingomyelin biosynthesis, and apoptosis are downregulated. This indicates that GC from follicles exposed to elevated NEFA during their early growth phases showed lower cell stress levels and lower oxidative damage when the final follicle preovulatory development took place under optimal antioxidant concentrations.
      Figure thumbnail gr8
      Figure 8Functional annotation of differentially expressed genes (DEG) detected in the high versus low wk 8 (w8) blood β-carotene (βC) + vitamin E (Vit. E) comparison in cows with high blood nonesterified fatty acid concentration at wk 2 (w2; 194 DEG, ). Positive regulation of a pathway is labeled with (+) and a negative regulation is shown as (−).
      It can be seen from the functional annotations shown in Figures 4, 6, and 8 that alterations associated with antioxidant concentrations included changes in interleukin- and chemokine-mediated signaling in both directions (up- and downregulation). This is perhaps confusing; however, it is important to remember that inflammatory responses at basal levels play a crucial regulatory role in ovarian follicle physiology, particularly during ovulation (
      • Smolikova K.
      • Mlynarcikova A.
      • Scsukova S.
      Role of interleukins in the regulation of ovarian functions.
      ). In contrast, OS commonly occurs in dairy cows during the transition period and plays a crucial role in the initiation and progression of many diseases. Immune suppression and uncontrolled inflammatory tissue responses are important elements of the pathogenesis of these disorders. The primary effect of antioxidant supplementation in dairy cows appears to involve the activity and effectiveness of the immune system (particularly the activity of lymphocytes and neutrophils;
      • Allison R.D.
      • Laven R.A.
      Effect of vitamin E supplementation on the health and fertility of dairy cows: A review.
      ). Daily intake of Vit. E and βC in multiparous cows has been linked to early resumption of ovarian activity postpartum (
      • Aoki M.
      • Ohshita T.
      • Aoki Y.
      • Sakaguchi M.
      Plasma thiobarbituric acid reactive substances, vitamin A and vitamin E levels and resumption of postpartum ovarian activity in dairy cows.
      ). In contrast, cows with severe NEB have recently been shown to exhibit significantly higher gene expression of IL1A, IL1B, IL8, IL15, IL23, and TNFA and lower gene expression of IL4 in GC of preovulatory follicles compared with cows with no NEB (
      • Warma A.
      • Descarreaux M.
      • Chorfi Y.
      • Dupras R.
      • Remillard R.
      • Ndiaye K.
      Interleukins' expression profile changes in granulosa cells of preovulatory follicles during the postpartum period in dairy cows.
      ). It is also worth highlighting that the c-Jun N-terminal kinases (JNK) cascade was positively regulated in the antioxidant comparisons (Figure 4, Figure 8). Although stimulation of JNK has been shown to mediate NEFA/ROS–induced impairment of insulin signaling (
      • Xu L.
      • Wang W.
      • Zhang X.
      • Ke H.
      • Qin Y.
      • You L.
      • Li W.
      • Lu G.
      • Chan W.Y.
      • Leung P.C.K.
      • Zhao S.
      • Chen Z.J.
      Palmitic acid causes insulin resistance in granulosa cells via activation of JNK.
      ), it has also been linked to physiological activation of autophagy in GC in association with the process of luteinization and P4 production (
      • Gao H.
      • Lin L.
      • Haq I.U.
      • Zeng S.
      Inhibition of NF-κB promotes autophagy via JNK signaling pathway in porcine granulosa cells.
      ). In the data presented here, it is clear that follicles of cows with high antioxidant concentrations showed lower levels of stress, as documented by downregulation of ubiquitin-mediated protein catabolism, response to chemokines, mitochondrial fragmentation, and apoptosis (Figure 8), even if the cows had high NEFA at wk 2. This suggests that the immunomodulatory effects of the antioxidant illustrated in the up- or downregulation of IL-mediated signaling and the effect on JNK cascade are more likely to be protective and linked to higher follicular cell survival, steroidogenic activity, or both.
      Overall, many other possible biological interactions could exist between metabolic and antioxidant factors that cannot be covered in one study. However, we identified here a few key players using straightforward comparisons based on 1 or 2 factors each. Ultimately, it is very difficult to distinguish correlation from causation in this type of study, but the associations observed here clearly illustrate how the interplay between the metabolic and antioxidant profiles of the cow during the early postpartum period can affect ovarian follicle cell functions in the preovulatory follicle at the moment of breeding.

      CONCLUSIONS

      The data presented here strongly support the Britt hypothesis and show that a more severe NEB, estimated in our study by NEFA concentrations at wk 2 pp, appears to exert a long-term carryover effect detectable in the GC transcriptomic profile of follicles that reach the preovulatory stage at the time of breeding (wk 8), suggesting persistent cellular stress and oxidative damage. This effect was more pronounced in cows with antioxidant deficiency at wk 8. In contrast, optimal blood antioxidant concentrations at wk 2 were associated with prominent transcriptomic changes in GC at wk 8 that indicate lower cellular stress and higher cell viability and that suggest better support for oocyte developmental competence. Such beneficial effects of antioxidants at wk 2 pp were not obvious in the presence of elevated NEFA concentrations early postpartum. Nevertheless, high blood antioxidant concentrations at wk 8 pp were linked to pathways involved in enhanced follicular metabolic activity, reduced stress levels, and oocyte supportive capacity that were exclusively detectable in cows with severe NEB (high NEFA at wk 2 pp). These observations suggest that optimal antioxidant status of the cow around the time of breeding might alleviate, at least in part, the effect of NEB on the follicular microenvironment. The fundamental data and the research-based insights presented in this study can be useful to support herd management to better cope with the increasing metabolic needs of high-yielding dairy cows. Based on these results, that should be accomplished through adequate dietary energy and antioxidant supplementation strategies both during transition and at the moment of breeding to support preovulatory follicle physiology and maximize pregnancy success.

      ACKNOWLEDGMENTS

      This project was financially supported by the Province of Antwerp (Belgium). We thank Marijke Van Looveren, Els Stevens, and Katleen Geerinckx at Hooibeekhoeve (Geel, Belgium) for their help with the animal trial and data retrieval. We also thank Peter Vercauteren and Joost Klop at CRV (Arnhem, the Netherlands) for their technical support for ovum pick-up. The authors have not stated any conflicts of interest.

      REFERENCES

        • Aardema H.
        • Gadella B.M.
        • van de Lest C.H.
        • Brouwers J.F.
        • Stout T.A.
        • Roelen B.A.
        • Vos P.L.
        Free fatty acid levels in fluid of dominant follicles at the preferred insemination time in dairy cows are not affected by early postpartum fatty acid stress.
        J. Dairy Sci. 2015; 98 (25648816): 2322-2336
        • Aardema H.
        • van Tol H.T.A.
        • Vos P.
        An overview on how cumulus cells interact with the oocyte in a condition with elevated NEFA levels in dairy cows.
        Anim. Reprod. Sci. 2019; 207 (31227325): 131-137
        • Abuelo A.
        • Hernández J.
        • Benedito J.L.
        • Castillo C.
        Oxidative stress index (OSi) as a new tool to assess redox status in dairy cattle during the transition period.
        Animal. 2013; 7 (23510791): 1374-1378
        • Allison R.D.
        • Laven R.A.
        Effect of vitamin E supplementation on the health and fertility of dairy cows: A review.
        Vet. Rec. 2000; 147 (11140928): 703-708
        • Andrews S.
        FastQC: A Quality Control Tool for High Throughput Sequence Data.
        • Aoki M.
        • Ohshita T.
        • Aoki Y.
        • Sakaguchi M.
        Plasma thiobarbituric acid reactive substances, vitamin A and vitamin E levels and resumption of postpartum ovarian activity in dairy cows.
        Anim. Sci. J. 2014; 85 (24506096): 532-541
        • Aon M.A.
        • Bhatt N.
        • Cortassa S.C.
        Mitochondrial and cellular mechanisms for managing lipid excess.
        Front. Physiol. 2014; 5 (25132820): 282
        • Argov N.
        • Moallem U.
        • Sklan D.
        Lipid transport in the developing bovine follicle: messenger RNA expression increases for selective uptake receptors and decreases for endocytosis receptors.
        Biol. Reprod. 2004; 71 (15056566): 479-485
        • Baddela V.S.
        • Sharma A.
        • Vanselow J.
        Non-esterified fatty acids in the ovary: Friends or foes?.
        Reprod. Biol. Endocrinol. 2020; 18 (32505200): 60
        • Bernabucci U.
        • Ronchi B.
        • Lacetera N.
        • Nardone A.
        Influence of body condition score on relationships between metabolic status and oxidative stress in periparturient dairy cows.
        J. Dairy Sci. 2005; 88 (15905432): 2017-2026
        • Bols P.E.J.
        • Vandenheede J.M.M.
        • Van Soom A.
        • de Kruif A.
        Transvaginal ovum pick-up (OPU) in the cow: A new disposable needle guidance system.
        Theriogenology. 1995; 43 (16727659): 677-687
        • Brickell J.S.
        • Wathes D.C.
        A descriptive study of the survival of Holstein-Friesian heifers through to third calving on English dairy farms.
        J. Dairy Sci. 2011; 94 (21426972): 1831-1838
        • Britt J.H.
        Impacts of early postpartum metabolism on follicular development and fertility.
        in: Proc. American Association of Bovine Practitioners (AABP) Annu. Conf. AABP, 1992: 39-43
        • Brown J.A.
        • Eberhardt D.M.
        • Schrick F.N.
        • Roberts M.P.
        • Godkin J.D.
        Expression of retinol-binding protein and cellular retinol-binding protein in the bovine ovary.
        Mol. Reprod. Dev. 2003; 64: 261-269
        • Carvalho P.D.
        • Souza A.H.
        • Amundson M.C.
        • Hackbart K.S.
        • Fuenzalida M.J.
        • Herlihy M.M.
        • Ayres H.
        • Dresch A.R.
        • Vieira L.M.
        • Guenther J.N.
        • Grummer R.R.
        • Fricke P.M.
        • Shaver R.D.
        • Wiltbank M.C.
        Relationships between fertility and postpartum changes in body condition and body weight in lactating dairy cows.
        J. Dairy Sci. 2014; 97 (24731646): 3666-3683
        • Chen H.
        • Yang Y.
        • Wang Y.
        • He Y.
        • Duan J.
        • Cheng J.
        • Li Q.
        The effects of phospholipase C on oestradiol and progesterone secretion in porcine granulosa cells cultured in vitro.
        Reprod. Domest. Anim. 2019; 54 (31319005): 1236-1243
        • Chen H.
        • Yang Y.
        • Wang Y.
        • Li Y.
        • He Y.
        • Duan J.
        • Xu D.
        • Pei Y.
        • Cheng J.
        • Yang L.
        • Hua R.
        • Li X.
        • Wang J.
        • Jiang X.
        • He H.
        • Wu L.
        • Liu D.
        • Li Q.
        Phospholipase C inhibits apoptosis of porcine primary granulosa cells cultured in vitro.
        J. Ovarian Res. 2019; 12 (31554511): 90
        • Chen L.
        • Wang C.
        • Huang S.
        • Gong B.
        • Yu J.
        • Shi Q.
        • Chen G.
        Effects of individual and multiple fatty acids (palmitate, oleate and docosahaexenoic acid) on cell viability and lipid metabolism in LO2 human liver cells.
        Mol. Med. Rep. 2014; 10 (25241938): 3254-3260
        • Chiumia D.
        • Chagunda M.G.
        • Macrae A.I.
        • Roberts D.J.
        Predisposing factors for involuntary culling in Holstein-Friesian dairy cows.
        J. Dairy Res. 2013; 80 (23164053): 45-50
        • Da Broi M.G.
        • Giorgi V.S.I.
        • Wang F.
        • Keefe D.L.
        • Albertini D.
        • Navarro P.A.
        Influence of follicular fluid and cumulus cells on oocyte quality: Clinical implications.
        J. Assist. Reprod. Genet. 2018; 35 (29497954): 735-751
        • De Bie J.
        • Langbeen A.
        • Verlaet A.A.J.
        • Florizoone F.
        • Immig I.
        • Hermans N.
        • Fransen E.
        • Bols P.E.J.
        • Leroy J.
        The effect of a negative energy balance status on beta-carotene availability in serum and follicular fluid of nonlactating dairy cows.
        J. Dairy Sci. 2016; 99 (27157583): 5808-5819
        • De Bie J.
        • Proost K.
        • Van Loo H.
        • Callens J.
        • Bols P.E.J.
        • Fransen E.
        • Leroy J.L.M.R.
        β-carotene and vitamin E in the dairy industry: Blood levels and influencing factors—A case study in Flanders.
        Vlaams Diergeneeskd. Tijdschr. 2019; 88: 137-149
        • Ferst J.G.
        • Missio D.
        • Bertolin K.
        • Gasperin B.G.
        • Leivas F.G.
        • Bordignon V.
        • Goncalves P.B.
        • Ferreira R.
        Intrafollicular injection of nonesterified fatty acids impaired dominant follicle growth in cattle.
        Anim. Reprod. Sci. 2020; 219 (32828411)106536
        • Gao H.
        • Lin L.
        • Haq I.U.
        • Zeng S.
        Inhibition of NF-κB promotes autophagy via JNK signaling pathway in porcine granulosa cells.
        Biochem. Biophys. Res. Commun. 2016; 473 (27016483): 311-316
        • Gilbert I.
        • Robert C.
        • Dieleman S.
        • Blondin P.
        • Sirard M.A.
        Transcriptional effect of the LH surge in bovine granulosa cells during the peri-ovulation period.
        Reproduction. 2011; 141 (21123518): 193-205
        • Green A.S.
        • Fascetti A.J.
        Meeting the vitamin A requirement: The efficacy and importance of β-carotene in animal species.
        ScientificWorldJournal. 2016; 2016 (27833936)7393620
        • Kim D.
        • Paggi J.M.
        • Park C.
        • Bennett C.
        • Salzberg S.L.
        Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype.
        Nat. Biotechnol. 2019; 37: 907-915
        • Leroy J.L.
        • Opsomer G.
        • Van Soom A.
        • Goovaerts I.G.
        • Bols P.E.
        Reduced fertility in high-yielding dairy cows: Are the oocyte and embryo in danger? Part I. The importance of negative energy balance and altered corpus luteum function to the reduction of oocyte and embryo quality in high-yielding dairy cows.
        Reprod. Domest. Anim. 2008; 43 (18384499): 612-622
        • Leroy J.L.
        • Vanholder T.
        • Mateusen B.
        • Christophe A.
        • Opsomer G.
        • de Kruif A.
        • Genicot G.
        • Van Soom A.
        Non-esterified fatty acids in follicular fluid of dairy cows and their effect on developmental capacity of bovine oocytes in vitro.
        Reproduction. 2005; 130 (16183866): 485-495
        • Leroy J.L.
        • Vanholder T.
        • Van Knegsel A.T.
        • Garcia-Ispierto I.
        • Bols P.E.
        Nutrient prioritization in dairy cows early postpartum: Mismatch between metabolism and fertility?.
        Reprod. Domest. Anim. 2008; 43 (18638110): 96-103
        • Liang C.-G.
        • Su Y.-Q.
        • Fan H.-Y.
        • Schatten H.
        • Sun Q.-Y.
        Mechanisms regulating oocyte meiotic resumption: Roles of mitogen-activated protein kinase.
        Mol. Endocrinol. 2007; 21 (17536005): 2037-2055
        • Liao Y.
        • Smyth G.K.
        • Shi W.
        FeatureCounts: An efficient general purpose program for assigning sequence reads to genomic features.
        Bioinformatics. 2014; 30: 923-930
        • Liu C.
        • Su K.
        • Chen L.
        • Zhao Z.
        • Wang X.
        • Yuan C.
        • Liang Y.
        • Ji H.
        • Li C.
        • Zhou X.
        Prediction of oocyte quality using mRNA transcripts screened by RNA sequencing of human granulosa cells.
        Reprod. Biomed. Online. 2021; 43 (34400084): 413-420
        • Love M.I.
        • Huber W.
        • Anders S.
        Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2.
        Genome Biol. 2014; 15: 550
        • Macaulay A.D.
        • Gilbert I.
        • Caballero J.
        • Barreto R.
        • Fournier E.
        • Tossou P.
        • Sirard M.A.
        • Clarke H.J.
        • Khandjian E.W.
        • Richard F.J.
        • Hyttel P.
        • Robert C.
        The gametic synapse: RNA transfer to the bovine oocyte.
        Biol. Reprod. 2014; 91 (25143353): 90
        • Marei W.
        JDS 21928 supplemental data. Version 2.
        • Mary A.E.P.
        • Artavia Mora J.I.
        • Ronda Borzone P.A.
        • Richards S.E.
        • Kies A.K.
        Vitamin E and beta-carotene status of dairy cows: A survey of plasma levels and supplementation practices.
        Animal. 2021; 15 (34252721)100303
        • Miqueo E.
        • Chiarle A.
        • Giuliodori M.J.
        • Relling A.E.
        Association between prepartum metabolic status and resumption of postpartum ovulation in dairy cows.
        Domest. Anim. Endocrinol. 2019; 69 (31301559): 62-67
        • Mu Y.M.
        • Yanase T.
        • Nishi Y.
        • Tanaka A.
        • Saito M.
        • Jin C.H.
        • Mukasa C.
        • Okabe T.
        • Nomura M.
        • Goto K.
        • Nawata H.
        Saturated FFAs, palmitic acid and stearic acid, induce apoptosis in human granulosa cells.
        Endocrinology. 2001; 142 (11459807): 3590-3597
        • Paglia D.E.
        • Valentine W.N.
        Studies on the quantitative and qualitative characterization of erythrocyte glutathione peroxidase.
        Lab. Clin. Med. 1967; 70: 158-169
        • Roche J.R.
        • Burke C.R.
        • Crookenden M.A.
        • Heiser A.
        • Loor J.L.
        • Meier S.
        • Mitchell M.D.
        • Phyn C.V.C.
        • Turner S.A.
        Fertility and the transition dairy cow.
        Reprod. Fertil. Dev. 2018; 30 (29539306): 85-100
        • Runkel E.D.
        • Baumeister R.
        • Schulze E.
        Mitochondrial stress: Balancing friend and foe.
        Exp. Gerontol. 2014; 56 (24603155): 194-201
        • Sakaguchi M.
        Practical aspects of the fertility of dairy cattle.
        J. Reprod. Dev. 2011; 57 (21422734): 17-33
        • Sartori R.
        • Bastos M.R.
        • Wiltbank M.C.
        Factors affecting fertilisation and early embryo quality in single- and superovulated dairy cattle.
        Reprod. Fertil. Dev. 2010; 22 (20003858): 151-158
        • Shin E.K.
        • Jeong J.K.
        • Choi I.S.
        • Kang H.G.
        • Hur T.Y.
        • Jung Y.H.
        • Kim I.H.
        Relationships among ketosis, serum metabolites, body condition, and reproductive outcomes in dairy cows.
        Theriogenology. 2015; 84 (25872806): 252-260
        • Smolikova K.
        • Mlynarcikova A.
        • Scsukova S.
        Role of interleukins in the regulation of ovarian functions.
        Endocr. Regul. 2012; 46 (23127507): 237-253
        • Uyar A.
        • Torrealday S.
        • Seli E.
        Cumulus and granulosa cell markers of oocyte and embryo quality.
        Fertil. Steril. 2013; 99 (23498999): 979-997
        • Valckx S.D.
        • Van Hoeck V.
        • Arias-Alvarez M.
        • Maillo V.
        • Lopez-Cardona A.P.
        • Gutierrez-Adan A.
        • Berth M.
        • Cortvrindt R.
        • Bols P.E.
        • Leroy J.L.
        Elevated non-esterified fatty acid concentrations during in vitro murine follicle growth alter follicular physiology and reduce oocyte developmental competence.
        Fertil. Steril. 2014; 102 (e1): 1769-1776
        • Warma A.
        • Descarreaux M.
        • Chorfi Y.
        • Dupras R.
        • Remillard R.
        • Ndiaye K.
        Interleukins' expression profile changes in granulosa cells of preovulatory follicles during the postpartum period in dairy cows.
        Cytokine X. 2020; 2 (33604553)100022
        • Webb R.
        • Garnsworthy P.C.
        • Gong J.G.
        • Armstrong D.G.
        Control of follicular growth: Local interactions and nutritional influences.
        J. Anim. Sci. 2004; 82: E63-E74
        • Wiltbank M.C.
        • Carvalho P.D.
        • Barletta R.V.
        • Fricke P.M.
        • Shaver R.D.
        Relationships among changes in body condition score and reproductive efficiency in lactating dairy cows.
        in: Proc. 27th Annual Tri-State Dairy Nutrition Conference, Fort Wayne, IN. 2018: 121-138
        • Wonnacott K.E.
        • Kwong W.Y.
        • Hughes J.
        • Salter A.M.
        • Lea R.G.
        • Garnsworthy P.C.
        • Sinclair K.D.
        Dietary omega-3 and -6 polyunsaturated fatty acids affect the composition and development of sheep granulosa cells, oocytes and embryos.
        Reproduction. 2010; 139 (19789173): 57-69
        • Xu L.
        • Wang W.
        • Zhang X.
        • Ke H.
        • Qin Y.
        • You L.
        • Li W.
        • Lu G.
        • Chan W.Y.
        • Leung P.C.K.
        • Zhao S.
        • Chen Z.J.
        Palmitic acid causes insulin resistance in granulosa cells via activation of JNK.
        J. Mol. Endocrinol. 2019; 62 (30913535): 197-206
        • Yang R.
        • Le G.
        • Li A.
        • Zheng J.
        • Shi Y.
        Effect of antioxidant capacity on blood lipid metabolism and lipoprotein lipase activity of rats fed a high-fat diet.
        Nutrition. 2006; 22 (17095404): 1185-1191
        • Yoon H.
        • Shaw J.L.
        • Haigis M.C.
        • Greka A.
        Lipid metabolism in sickness and in health: Emerging regulators of lipotoxicity.
        Mol. Cell. 2021; 81 (34547235): 3708-3730