Peroxisome Proliferator-Activated Receptor Activation in Precision-Cut Bovine Liver Slices Reveals Novel Putative PPAR Targets in Periparturient Dairy Cows

Abstract

Metabolic challenges experienced by dairy cows during the transition between pregnancy and lactation (also known as peripartum), are of considerable interest from a nutrigenomic perspective. The mobilization of large amounts of non-esterified fatty acids (NEFA) leads to an increase in NEFA uptake in the liver, the excess of which can cause hepatic accumulation of lipids and ultimately fatty liver. Interestingly, peripartum NEFA activate the Peroxisome Proliferator-activated Receptor (PPAR), a transcriptional regulator with known nutrigenomic properties. The study of PPAR activation in the liver of periparturient dairy cows is thus crucial; however, current in vitro models of the bovine liver are inadequate, and the isolation of primary hepatocytes is time consuming, resource intensive, and prone to errors, with the resulting cells losing characteristic phenotypical traits within hours. The objective of the current study was to evaluate the use of precision-cut liver slices (PCLS) from liver biopsies as a model for PPAR activation in periparturient dairy cows. Three primiparous Jersey cows were enrolled in the experiment, and PCLS from each were prepared prepartum (-8.0 ± 3.6 DIM) and postpartum (+7.7± 1.2 DIM) and treated independently with a variety of PPAR agonists and antagonists: the PPARα agonist WY-14643 and antagonist GW-6471; the PPARδ agonist GW-50156 and antagonist GSK-3787; and the PPARγ agonist rosiglitazone and antagonist GW-9662. Gene expression was assayed through RT-qPCR and RNAseq, and intracellular triacylglycerol (TAG) concentration was measured. PCLS obtained from postpartum cows and treated with a PPARγ agonist displayed upregulation of ACADVL and LIPC while those treated with PPARδ agonist had increased expression of LIPC, PPARD, and PDK4. In PCLS from prepartum cows, transcription of LIPC was increased by all PPAR agonists and NEFA. TAG concentration tended to be larger in tissue slices treated with PPARδ agonist compared to CTR. Use of PPAR isotype-specific antagonists in PCLS cultivated in autologous blood serum failed to decrease expression of PPAR targets, except for PDK4, which was confirmed to be a PPARδ target. Transcriptome sequencing revealed considerable differences in response to PPAR agonists at a false discovery rate-adjusted p-value of 0.2, with the most notable effects exerted by the PPARδ and PPARγ agonists. Differentially expressed genes were mainly related to pathways involved with lipid metabolism and the immune response. Among differentially expressed genes, a subset of 91 genes were identified as novel putative PPAR targets in the bovine liver, by cross-referencing our results with a publicly available dataset of predicted PPAR target genes, and supplementing our findings with prior literature. Our results provide important insights on the use of PCLS as a model for assaying PPAR activation in the periparturient dairy cow.

Keywords: PCLS; PPAR; dairy cows; liver; nutrigenomics; peripartum.

Figure 1

Relative normalized expression of ACADVL, FABP1, HES6, LIPC, PDK4, PGC1A, PPARA, PPARD and PPARG in response to synthetic PPAR agonists (100 μM WY-14643, a PPARα agonist; 50 μM GW-501516, a PPARδ agonist; and 100 μM Rosiglitazone, a PPARγ agonist), 200 μM palmitic acid, or 100 μM NEFA isolated from each respective animal, in PCLS obtained from prepartum [(A), −10 DIM] and postpartum [(B), +10 DIM] cows, and cultured for 18 h in William's Medium E. *indicates significant differences (P < 0.05) when compared to WME control.

Figure 2

Amount of triacylglycerol in PCLS from postpartum cows (+10 DIM) treated with PPAR isotype-specific agonists (100 μM WY-14643, a PPARα agonist; 50 μM GW-501516, a PPARδ agonist; and 100 μM Rosiglitazone, a PPARγ agonist), 200 μM palmitic acid, or 100 μM NEFA isolated from each respective animal. Results are presented as fold change vs. each animal's control group (PCLS cultured in William's Medium E only). ^#indicates tendencies (P < 0.1) when compared to control. Incubation time was 18 h.

Figure 3

Relative normalized expression of ACADVL, FABP1, HES6, LIPC, PDK4, PGC1A, PPARA, PPARD and PPARG in response to synthetic PPAR antagonists (50 μM GW-6471, a PPARα antagonist; 50 μM GSK-3787, a PPARδ antagonist; and 50 μM GW-9662, a PPARγ antagonist) or 200 μM palmitic acid, in PCLS obtained from prepartum (A, −10 DIM) and postpartum (B, +10 DIM) cows, and cultured in blood serum. *Indicates significant differences (P < 0.05) when compared to serum control; Incubation time was 18 h.

Figure 4

Main categories and subcategories of KEGG pathways induced via the use of isotype-specific PPAR agonists in PCLS, as summarized by the Dynamic Impact Approach. Blue bars refer to the impact in terms of overrepresented genes in the pathways, while the shaded cell denotes the flux, i.e., the overall effect on the pathway, with red denoting activation and green denoting inhibition.

Figure 5

Main KEGG pathways highlighted by DIA, divided by subcategory. Triangles (color-coded) represent the impact in terms of overrepresented genes in the pathways, whereas bars (also color-coded) refer to the overall flux of the pathway, with negative numbers indicating inhibition of the pathway, and positive numbers indicating induction of the pathway.

Figure 6

Overrepresented gene ontology (GO) terms with a p-value <0.01, according to DAVID, by the three PPAR agonists. Pathways associated with positive numbers are overrepresented in the subset of genes significantly upregulated by the treatment, whereas pathways represented by a negative number are overrepresented in the genes significantly downregulated by the treatment. Absolute numbers correspond to fold enrichment of the pathway.

Figure 7

Genes predicted as PPAR target by the PPARgene algorithm, which were also differentially expressed by the three treatments in the current study. The color of the words corresponds to the degree of confidence of the prediction, while the color of the underlining line indicates the direction of the differential expression, with red indicating upregulation in response to the treatment, and green indicating downregulation.