The hepatic transcriptome of the turkey poult (Meleagris gallopavo) is minimally altered by high inorganic dietary selenium

Rachel M Taylor; Kristelle M Mendoza; Juan E Abrahante; Kent M Reed; Roger A Sunde

doi:10.1371/journal.pone.0232160

. 2020 May 7;15(5):e0232160. doi: 10.1371/journal.pone.0232160

The hepatic transcriptome of the turkey poult (Meleagris gallopavo) is minimally altered by high inorganic dietary selenium

Rachel M Taylor ¹, Kristelle M Mendoza ², Juan E Abrahante ³, Kent M Reed ², Roger A Sunde ^1,^*

Editor: Juan J Loor⁴

PMCID: PMC7205448 PMID: 32379770

Abstract

There is interest in supplementing animals and humans with selenium (Se) above Se-adequate levels, but the only good biomarker for toxicity is tissue Se. We targeted liver because turkeys fed 5 μg Se/g have hepatic Se concentrations 6-fold above Se-adequate (0.4 μg Se/g) levels without effects on growth or health. Our objectives were (i) to identify transcript biomarkers for high Se status, which in turn would (ii) suggest proteins and pathways used by animals to adapt to high Se. Turkey poults were fed 0, 0.025, 0.4, 0.75 and 1.0 μg Se/g diet in experiment 1, and fed 0.4, 2.0 and 5.0 μg Se/g in experiment 2, as selenite, and the full liver transcriptome determined by RNA-Seq. The major effect of Se-deficiency was to down-regulate expression of a subset of selenoprotein transcripts, with little significant effect on general transcript expression. In response to high Se intake (2 and 5 μg Se/g) relative to Se-adequate turkeys, there were only a limited number of significant differentially expressed transcripts, all with only relatively small fold-changes. No transcript showed a consistent pattern of altered expression in response to high Se intakes across the 1, 2 and 5 μg Se/g treatments, and there were no associated metabolic pathways and biological functions that were significant and consistently found with high Se supplementation. Gene set enrichment analysis also found no gene sets that were consistently altered by high-Se and supernutritional-Se. A comparison of differentially expressed transcript sets with high Se transcript sets identified in mice provided high Se (~3 μg Se/g) also failed to identify common differentially expressed transcript sets between these two species. Collectively, this study indicates that turkeys do not alter gene expression in the liver as a homeostatic mechanism to adapt to high Se.

Introduction

Selenium (Se) was first identified as a toxic element for livestock in the 1930s [1], and then reported to be carcinogenic for rats [2], thus lumping Se as a poison with other elements like arsenic. In 1957, however, Se was found to be an essential nutrient for rats [3] and chickens [4], and soon after was shown to prevent a cardiomyopathy in humans called Keshan disease [5]. In 1973, Se was identified as an essential cofactor for glutathione peroxidase (GPX), establishing a biochemical role for Se as a required essential trace element [6]. Tissue levels of GPX and many other selenoenzymes fall dramatically in animals fed Se-deficient diets, reach a well-defined plateau above a breakpoint, and thus are good biomarkers for assessment of Se-deficient vs. Se-adequate status [7]. Using this approach, the minimum dietary Se requirement in turkey poults is 0.4 μg Se/g [8–10], which is 4-fold higher than minimum Se requirements in rats [11]. Thus tissue Se concentrations, signs of Se-specific disease, and selenoenzyme activities can be good biomarkers of Se status and Se deficiency.

Co-translational synthesis and insertion of the Se cofactor, selenocysteine, into GPX and other selenoproteins requires an in-frame UGA and a 3’UTR stemloop (the SECIS element) in the transcript to facilitate incorporation of selenocysteine at what otherwise would be a stop codon [12]. Using these transcript elements as a signature to screen annotated genomes for selenoproteins, 25 human, 24 rodent, and 24 avian selenoproteins–the entire selenoproteome–have been characterized [13–15]. No selenoenzyme activity, however, has been identified that increases substantially above levels in Se-adequate animals when diets are supplemented with Se at levels above the requirement [7].

In rats, there is no effect on growth in male rats fed 2 μg Se/g, but a 20% decrease in growth in rats fed 5 μg Se/g [16], whereas poultry appear less susceptible to Se toxicity. Chicks fed 1 and 5 μg Se/g for 4 wk show no effect on body weight as compared to chicks supplemented with Se-adequate diets [17,18], and the same is observed in turkey poults fed 1–5 μg Se/g [9,10,19]. In both avians and rats, higher dietary Se levels and longer exposures do result in toxicity [20–22]. In these species, blood and tissue Se concentrations increase above the Se requirement in a near linear fashion with increasing Se intake, but do not plateau; at 5 μg Se/g, rat kidney accumulates more Se than liver whereas in the turkey poult, the liver accumulates substantially more Se than the kidney [9–11,23–26].

Transcript levels have potential as biomarkers for trace element status. Homeostatic mechanisms are present for altering uptake/export transporters, mineral carriers, and/or storage forms for Zn, Fe and Cu [27–29] across the range from deficiency to excess. In Se deficiency, a subset of selenoprotein transcripts decrease dramatically in Se deficiency in rodents, and can be used as biomarkers for Se deficiency [26]. We initially thought that animals might up-regulate expression of selenoprotein transcripts upon supernutritional Se supplementation, but found that this did not occur in rats, mice, chickens, turkeys or C. elegans [10,16,18,30]. Characterization of the effect of up to 2 μg Se/g diet as selenite in rats resulted in fewer than 10 transcripts with significant differential gene expression. Feeding 5 μg Se/g diet, however, significantly altered the expression of 4% of the rat transcriptome suggesting that the resulting transcript set might contain more transcripts that those specifically responding to high Se status, as this level of Se supplementation was associated with a 20% decrease in growth [16]. In a similar study which provided mice with 1.6 μg Se/ml in the drinking water for 16 wk (equivalent to feeding ~3 μg Se/g diet), gene set enrichment analysis also identified a small set of 18 liver transcripts with differential expression p < 0.05 [31]. We hypothesized that the turkey, apparently more resistant to high Se, might be a better model to identify transcripts (potential biomarkers) associated with changes in gene expression related to adaptation to high Se status.

To assess the full turkey transcriptome for the effect of Se status from Se-deficient to high-Se, we conducted RNA-Seq analysis on liver mRNA from turkey poults fed 0, 0.025, 0.4, 0.75 and 1.0 μg Se/g diet in experiment 1, and fed 0.4, 2.0 and 5.0 μg Se/g in experiment 2. Our objectives were (i) to identify transcript biomarkers for high Se status, which in turn (ii) would suggest proteins and pathways used by animals to adapt to high Se. We targeted the liver because turkey poults fed 5 μg Se/g have hepatic Se concentrations 6-fold above levels in poults fed Se-adequate diet (0.4 μg Se/g) but without apparent effects on growth or health [10].

Methods

Reagents

Molecular biology reagents were purchased from Promega (Madison, WI, USA), Invitrogen (Carlsbad, CA, USA), or Sigma (St. Louis, MO, USA). All other chemicals were of molecular biology or reagent grade.

Animals and diets

In both experiments 1 and 2, day-old male Nicholas White-derived turkey poults (kindly donated by Jennie-O Turkey Store, Barron, WI) were allocated randomly to treatment (5 poults/Se supplementation level) and housed in battery cages with raised wire floors and 24-hr lighting. The treatment and diets have been described in detail previously [8–10]. Briefly, the basal Se-deficient torula yeast-based diet (0.005 μg Se/g) was supplemented with 7.0% crystalline amino acids, including 0.93% methionine, to better match NRC recommendations for protein and amino acids [32]. Vitamin E (as all-rac-α-tocopherol acetate) was supplemented at 150 mg/kg (12.5X NRC requirement). In experiment 1, the basal diet was supplemented with 0, 0.025, 0.4, 0.75 or 1 μg Se/g diet as Na₂SeO₃. In experiment 2, the basal diet was supplemented with 0.4, 2 or 5 μg Se/g diet as Na₂SeO₃. The animal protocol was approved by the Research Animal Resources Committee at the University of Wisconsin-Madison (protocol no. A005368).

Tissue Se and Selenoenzymes

Poults in both experiments were killed at 28 days by terminal CO₂ overexposure followed by exsanguination. Entire livers were collected and immediately frozen at -80°C until analysis. Liver Se concentrations (n = 4) and selenoenzyme activities (n = 5) for plasma GPX3, and liver GPX1, GPX4, and thioredoxin reductase (TXNRD) were measured as described previously [8–10]. Data were analyzed by one-way ANOVA and variance equality was tested using Levene’s test for homogeneity of variances [33]. When the main effect of diet was significant, differences between means were assessed by Duncan’s multiple range test (p < 0.05) with Kramer’s modification for unequal class sizes when necessary [34].

RNA isolation and high-throughput sequencing

Total RNA from liver tissues (n = 4/treatment) was isolated with TRIzol Reagent (Invitrogen, catalog no. 15596–026) following the manufacturer’s protocol. The RNA pellet was dissolved in 300 μL diethylpyrocarbonate-treated water and quantitated using a DeNovix DS-11+ spectrophotometer (DeNovix Inc., Wilmington, DE). Library preparation and sequencing were done at the University of Wisconsin-Madison Biotechnology Center. Samples were prepared for sequencing with the Illumina TruSeqv4 sample preparation kit, according to manufacturer’s instructions. Libraries were multiplexed and sequenced on an Illumina HiSeq 2500. At the University of Minnesota, the 125-bp paired-end reads (27.2 ± 1.97 M reads/sample) were trimmed to removed sequence adaptors and low-quality bases (resulting 189 bp average inset mean), and mapped using Bowtie (v2.2.4.0) to the turkey genome build 5.0 Annotation 102, as described previously [35], resulting in an average of 16,132 genes detected (min 15,685, max 16,567). The original RNA Seq data discussed in this publication have been deposited in NIH's NCBI’s Sequence Read Archive (SRA) and are accessible through SRA accession PRJNA614544 (available at https://www.ncbi.nlm.nih.gov/sra/PRJNA614544).

A number of the selenotranscripts in the NCBI turkey Annotation 102 are mis-annotated, mis-assembled, or missing. For the 24 selenoproteins plus selenophosphate synthase1 (SEPHS1, not a selenoprotein in the turkey), corrected full-length selenotranscript sequences were generated based on the turkey Annotation 102, combined with our cloned and sequenced turkey selenotranscripts [14] and with unplaced turkey genome scaffold sequences, to include the selenocysteine UGA codon, the SECIS element, and the full 5’ and 3’UTR when sequences were available, and then assembled into a pseudo genome and remapped. Briefly in Annotation 102, SELENOU (Selenoprotein U) is designated as FAM213A ID:100550701, SELENOP1 (Selenoprotein P1) is misassembled and portions designated as SELENOP ID:100547086 and as LOC104914865 ID:104914865, SELENOP2 is designated as LOC100546913 ID:100546913, TXNRD1 (Thioredoxin reductase 1) is designated as LOC100545667 ID: 100545667, and SELENOW (Selenoprotein W) is not present in Annotation 102. Many others are mis-annotated. Based on raw counts (≥20), 21 of the 24 turkey selenotranscripts were judged to be present in this mapping, with DIO3 (Deiodinase 3), GPX2, and SELENOH (Selenoprotein H) not present.

Functional analysis

Principal component analysis (PCA) was performed with the R package FactoMineR [36] to identify and quantify variability in the data. Volcano plots were constructed to visualize the range of differential expression and significance across treatment groups. For differential expression analysis, pairwise comparisons between the Se-adequate control group (0.4 μg Se/g diet) and all other treatment groups were made in the R package DESeq2 [37], following the standard workflow. The false discovery rate (FDR) significance is designated as q.

Hierarchical cluster analysis was conducted, using fold-change values of differentially expressed (DE) transcripts (q < 0.1) in liver of poults fed graded levels of dietary Se (n = 4), as compared to Se-adequate (0.4 μg Se/g) poults, as generated by the likelihood ratio test (LRT) in DESeq2, subjected to centroid-linked Pearson correlation clustering implemented in Cluster 3.0 and visualized with Java Treeview [38].

Pathway analysis and gene ontology analysis

NCBI homolgene website (https://www.ncbi.nlm.nih.gov/homologene) was used to identify mammalian homologs for turkey genes with DE transcripts q < 0.05. This resulted in gene sets containing 7, 12 and 57 characterized genes for low Se (0 and 0.025 μg Se/g), moderately high Se (1 μg Se/g), and high Se (2 and 5 μg Se/g) treatments, respectively. These gene sets were analyzed using Enrichr for Human WikiPathways 2019 and for Biological Process 2018 [39].

Gene Set Enrichment Analysis (GSEA) [40,41] was used to assess the effect of Se status on two pre-determined sets of genes (Molecular Signatures Database v. 7.0 (MSigDB), http://software.broadinstitute.org/gsea/msigdb/index.jsp): the Hallmark gene set of 50 well-defined biological states or processes; the KEGG gene set of 186 gene sets derived from the KEGG pathway database; plus an additional Selenoprotein gene set of 25 turkey selenogenes (24 selenoproteins plus selenophosphate synthetase-1 [SEPHS1]) [13]. For this analysis, the full turkey transcriptome was reduced to 11,612 human gene transcripts. For each experiment, each Se treatment was analyzed by comparison with the corresponding Se-adequate (0.4 μg Se/g) treatment, and the datasets trimmed to transcripts with at least one sample >14 counts, and then normalized prior to GSEA (4 replicates/treatment). An FDR (q) < 0.25 is considered to be significant [41]. In addition, GSEA was conducted using the total MSigDB gene ontology (GO) collection of 9996 gene sets for biological processes, cellular components, and molecular functions.

Results

Se status

The response of the liver transcriptome in turkey poults was studied in two experiments that fed graded levels of inorganic Se (as selenite) in diets supplemented with vitamin E at 12.5X the NRC requirement [32] so that direct effects of altered Se status could be studied without downstream secondary effects (disease) arising from physiological effects of Se deficiency or toxicity in combination with marginal vitamin E status. Characterizations of poult growth, selenoenzyme activities and selenoprotein transcript expression from experiment 1 [8,9] and experiment 2 [10] have been reported in detail previously. Briefly, there were no overt signs of impaired health due to Se supplementation levels and no significant effects on final body weight (Table 1). In 0 μg Se/g poults, liver Se fell to 6% of levels in Se-adequate poults (0.4 μg Se/g), and selenoenzyme activities for GPX1, GPX4, and TXNRD decreased to 3–15% of Se-adequate levels, showing that the poults fed the 0 μg Se/g diet were Se deficient. In experiment 2, 5 μg Se/g raised liver Se to 6X the level in Se-adequate poults, thus confirming that these poults received high Se diets. These high Se diets, however, did not significantly increase levels of selenoenzyme activity above Se-adequate levels (Table 1). Collectively, the biomarkers of Se status confirm that these dietary treatments resulted in animals that were of Se-deficient to high-Se status.

Table 1. Liver selenoenzyme activities and Se concentration in turkey poults fed graded levels of dietary Se.

Expt	Se treatment	Final weight	Se conc.	GPX1 activity	GPX4 activity	TXNRD activity
	(μg/g)	(g)	(μg/g wet wt)	(EU/g)	(EU/g)	(EU/g)
Expt 1	0	526 ± 77^*	0.38 ± 0.06^d	3.2 ± 1.3^b	4.4 ± 0.4^d	0.41 ± 0.17^c
	0.025	569 ± 67	1.12 ± 0.18^d	4.6 ± 1.8^b	8.3 ± 1.0^d	1.31 ± 0.12^c
	0.4	717 ± 113	6.03 ± 0.11^c	95.1 ± 11.0^a	73.0 ± 5.1^a	2.70 ± 0.18^b
	0.75	765 ± 76	8.55 ± 0.60^c	100.2 ± 10.2^a	64.4 ± 4.1^ab	2.79 ± 0.28^b
	1	647 ± 158	8.40 ± 0.18^c	90.9 ± 20.0^a	57.9 ± 8.2^b	2.73 ± 0.08^b
Expt 2	0.4	742 ± 61	5.37 ± 0.48^c	73.4 ± 10.1^a	32.7 ± 3.9^c	3.08 ± 0.24^ab
	2	569 ± 50	12.53 ± 0.26^b	92.8 ± 17.1^a	39.4 ± 5.8^c	2.97 ± 0.28^ab
	5	588 ± 70	31.85 ± 3.12^a	103.9 ± 8.4^a	43.7 ± 2.4^bc	3.48 ± 0.19^a
	ANOVA p-value	0.366	4.61E-11	4.37E-06	2.34E-10	7.05E-12

Open in a new tab

*Values are mean ± SEM (n = 4 for Se; n = 5 for other biomarkers). Values not sharing a common letter are significantly different (p< 0.05).

PCA plots

PCA was performed either using all ~17,100 transcripts expressed at or above the level of detection in each sample from both experiments or using the 500 highest-expressed transcripts in each sample; no significant clustering was observed between treatment groups. When PCA was conducted for the effect of low Se by comparing 0 and 0.025 vs. 0.4 μg Se/g groups in experiment 1, poults supplemented with 0 μg Se/g diet tended to cluster together and PCA accounted for 53% of the variation (Fig 1A). When PCA was conducted for the effect of high Se by comparing 2 and 5 vs. 0.4 μg Se/g groups in experiment 2, there was no notable clustering of samples within treatment groups and PCA could account for only 24% of the variation (Fig 1B). These PCA analyses indicate that high Se did not exert major effects on general transcript expression in these studies.

Fig 1 — (A) Analysis of low-Se treatments in experiment 1; (B) Analysis of high Se treatments in experiment 2.

Volcano plots

For each Se treatment, volcano plots were constructed to show average fold changes and significance relative to the Se-adequate average for that experiment (Fig 2). In Se-deficient poults, 10 transcripts were significantly DE, including 4 down-regulated selenotranscripts: DIO1, SELENOP1, SELENOP2, and SELENOU (FAM213A) (Fig 2A). In poults supplemented with 0.025 μg/g, 23 transcripts were DE, including SELENOP1 (Fig 2B). Supplementing with 0.75, 1 and 2 μg/g resulted in 0, 218 and 2 DE transcripts, respectively (Fig 2C–2E). High-Se poults had 45 DE liver transcripts (Fig 2F), demonstrating that increasing Se supplementation did not increasingly enlarge the total number of DE transcripts. In addition, few transcript in any treatment had fold-changes >|±2|.

Fig 2 — Plotted are log₁₀(unadjusted p-value) vs. fold change (FC). (A) 0 vs 0.4 μg/g, (B) 0.025 vs 0.4 μg/g, (C) 0.75 vs 0.4 μg/g, (D) 1 vs 0.4 μg/g, (E) 2 vs 0.4 μg/g, (F) 5 vs 0.4 μg/g. Values are means (n = 4). Orange, FC ≥ |±2| only; Red, q ≤ 0.1 only; Green, q ≤ 0.1 and FC ≥ |±2|; Black, neither condition satisfied.

Differential expression

Se-deficient and low Se

Pairwise analysis of the 0 vs. 0.4 μg Se/g treatments found just 10 DE transcripts with q < 0.05, including 4 down-regulated selenotranscripts: SELENOU, DIO1, SELENOP1 and SELENOP2. Non-selenoprotein transcripts HSPB11, RASGRP3, and SGTA were also significantly down-regulated by Se deficiency; SOCS3 and 2 uncharacterized transcripts were significantly up-regulated (Table 2). HSPB11 is a heat shock protein family B (small) member 11, RASGRP3 is a RAS guanyl releasing protein 3, SGTA is a small glutamine rich tetratricopeptide repeat containing alpha, SOCS3 is a suppressor of cytokine signaling 3, LOC104909734 is a sulfotransferase family cytosolic 2B member 1-like, and LOC100545600 is a C-C motif chemokine 3-like protein. The function and NCBI Gene ID for all significant DE transcripts found in the 0, 0.025, 2, and 5 μg Se/g comparisons are given in S1 Table. Analysis of the 0.025 vs. 0.4 μg Se/g treatments found just 2 DE transcripts with q < 0.05; this included significant down-regulation of SELENOP1 expression and up-regulation of BLCAP expression. BLCAP is a bladder cancer-associated protein. For 0 or 0.025 μg Se/g, 3 additional selenoprotein transcripts, TXNRD3, GPX4, and GPX1, had DE with unadjusted p < 0.05 but q > 0.1 (Table 2). The only characterized transcripts up-regulated by low Se were SOCS3 and BLCAP by 0 and 0.025 μg Se/g, respectively. SOCS3 is a suppressor of cytokine signaling 3. Clearly low Se status targeted a subset of selenoprotein transcripts for down-regulation, but there was little additional impact of Se deficiency on the transcriptome.

Table 2. Transcripts DE by 0 and 0.025 μg Se/g vs. Se-adequate (Exp 1)^*.

Symbol	Gene ID	Fold	p-value	q-value	0.75	1	2	5
0 μg Se/g Differentially Expressed Transcripts					μg Se/g	μg Se/g	μg Se/g	μg Se/g
HSPB11	100543060	-2.52	5.02E-12	8.21E-08
SELENOU	100550701	-2.62	1.64E-10	1.34E-06
RASGRP3	100539029	-2.84	4.32E-08	0.0002		↓	↓
DIO1	100546412	-2.21	4.43E-08	0.0002
SELENOP1	100547086	-2.07	1.12E-07	0.0004
SELENOP2	100546913	-2.00	6.59E-07	0.0018
LOC104909734	104909734	2.36	9.30E-07	0.0022
SGTA	100543804	-2.31	7.27E-06	0.0132
SOCS3	100547594	2.28	8.07E-06	0.0132
LOC100545600	100545600	2.21	8.08E-06	0.0132
TXNRD3	100539983	-1.60	0.0005	0.3657
GPX4	100379155	-1.89	0.0009	0.5592
0.025 μg Se/g Differentially Expressed Transcripts
SELENOP1	100547086	-1.85	7.47E-06	0.0138
BLCAP	100547906	1.42	2.68E-05	0.0248		↑		↑
SELENOU	100550701	-1.66	0.0007	0.0956
GPX1	100379156	-1.47	0.0033	0.1429

Open in a new tab

*Fold changes, p-value, and FDR q-value for DE transcripts (q < 0.05) by 0 and 0.025 vs. 0.4 μg Se/g. Also listed are selected additional selenotranscripts with q >0.05. Arrows in columns (on right) indicate direction of fold-change for transcripts that were also DE (q < 0.05) by 0.75, 1, 2, and 5 vs 0.4 μg Se/g

High Se

Pairwise comparison of high dietary Se supplementation at 2 and 5 μg Se/g in experiment 2 only significantly altered gene expression for 2 and 14 transcripts, respectively; all changes were < |±2.5| fold, and there were no transcripts that overlapped between these two sets (Table 3). Moderate high Se (2 μg Se/g) significantly up-regulated SDPR, a serum deprivation response factor, and down-regulated RASGRP3, which was also one of the non-selenoprotein transcripts significantly down-regulated by Se deficiency. Dietary Se at 5 μg Se/g significantly up-regulated expression of 10 transcripts and down-regulated 4 transcripts; all but one change was < |±2.0| fold, and no selenotranscripts were significantly regulated. Included in up-regulated transcripts were genes for ATP synthases, kinases, a phosphatidylethanolamine binding protein, a mitochondrial CytC oxidase, and BLCAP which was also up-regulated by 0.025 but not 0 μg Se/g (Tables 2 and 3).

Table 3. Transcripts DE by 2.0 and 5.0 μg Se/g vs. Se-adequate (Exp 2)^*.

Symbol	Gene ID	Fold	p-value	q-value	0	0.025	0.75	1	2	5
2,0 μg Se/g Differentially Expressed Transcripts					μg Se/g	μg Se/g	μg Se/g	μg Se/g	μg Se/g	μg Se/g
RASGRP3	100539029	-2.50	1.51E-06	2.57E-02	↓			↓	↓
SDPR	100545738	1.83	4.11E-06	3.50E-02					↑
SELENOI	100548385	-1.44	8.75E-03	0.9530
SEPHS1	100550629	-1.36	2.58E-02	1.0000
5.0 μg Se/g Differentially Expressed Transcripts
ATP5G1	100548571	1.78	3.79E-09	0.0000				↑		↑
ATPAF2	100544773	2.21	5.19E-09	0.0000						↑
BLCAP	100547906	1.55	1.59E-07	0.0003		↑		↑		↑
LOC100545745	100545745	1.55	2.60E-06	0.0032						↑
RNH1	100548440	1.43	2.71E-06	0.0032				↑		↑
DAD1	100547759	1.47	0.0000	0.0158				↑		↑
RPS27A	100542056	1.65	0.0000	0.0365						↑
BRI3	104913503	1.77	5.92E-05	0.0431						↑
CCND1	100539759	-1.54	8.20E-05	0.0472						↓
CHCHD2	100550053	1.44	1.12E-04	0.0472						↑
PDPK1	100541724	-1.51	0.0001	0.0472				↓		↓
CDC42BPB	100550337	-1.39	0.0001	0.0472						↓
PEBP1	100545499	1.46	0.0001	0.0472				↑		↓
LOC104912775	104912775	-1.80	9.09E-05	0.0472						↓
SEPHS1	100550629	1.41	0.0232	1.0000

Open in a new tab

*Fold changes, p-value, and FDR q-value for DE transcripts (q < 0.05) by 2 and 5 vs. 0.4 μg Se/g. Also listed are selected additional selenotranscripts with q >0.05. Arrows in columns (on right) indicate direction of fold-change for transcripts that were also DE (q < 0.05) by 0, 0.025, 0.75, 1, 2, and 5 vs 0.4 μg Se/g

Supernutritional Se

Pairwise comparison of 0.75 μg Se/g with Se-adequate treatment in experiment 1 did not result in any DE transcripts. This indicates that this level of supernutritional Se supplementation had little or no effect on the general transcriptome (S2 Table).

Supplementation with 1 μg Se/g in experiment 1 resulted in the largest set of DE transcripts (S2 Table): 50 up- and 38 down-regulated. These transcripts included BLCAP which was also up regulated by 0.025 but not 0 μg Se/g, and RASGRP3 which was also down-regulated by 0 but not 0.025 μg Se/g. These U-shaped and inverted-U-shaped responses show that expression of these transcripts was not changed solely in response to increased Se status. The set of 88 DE transcripts for 1 μg Se/g, however, did include 6 transcripts which were also significantly up-regulated and 1 transcript that was also significantly down-regulated by 5 μg Se/g, but not by 2 μg Se/g or by low Se treatment (Table 3). Amongst these 7 transcripts with significant DE by 1 and 5 μg Se/g, no transcript also had DE with unadjusted p < 0.05 for 2 μg Se/g, indicating that changes in DE of these transcripts were not uniquely associated with high Se status.

Hierarchical cluster analysis

The overall pattern of regulation by Se-deficient and low-Se status was further examined by hierarchical cluster analysis on an expanded set of 33 DE transcripts with q < 0.1 (Fig 3A). Across the Se treatments (0–0.4 μg Se/g), there was a cluster of down-regulated transcript that included four selenotranscripts (SELENOP1, SELENOP2, DIO1, SELENOU), and a cluster of four up-regulated transcripts. This analysis indicated that almost half of these transcripts showed near 2-fold regulation, up or down, by 0 μg Se/g vs. Se-adequate 0.4. In addition, a third large cluster of transcripts showed inverted U-shaped expression with levels at 0.025 μg Se/g higher than at 0 or 0.4 μg Se/g (Fig 3A). When the remapped selenoprotein transcripts were subjected to cluster analysis (Fig 3B), a cluster of four selenotranscripts were down-regulated by both 0 and 0.025 μg Se/g, and a cluster of 14 additional selenotranscripts were down-regulated only by 0 μg Se/g, similar to our previous reports [8–10].

Cluster analysis of the high Se treatments revealed 3 clusters, with only modest fold-changes relative to Se-adequate treatment (Fig 3C). The largest cluster showed down-regulation of transcripts by 2 μg Se/g but not by 5 or 0.4 μg Se/g–an inverted U-shaped response. A second cluster showed down-regulation by 2 and 5 μg Se/g vs 0.4 μg Se/g. Lastly, a small cluster of 2 transcripts showed a U-shaped response, i.e. up-regulation by 2 μg Se/g. Thus, there was not a consistent pattern of regulation in high Se and few of these transcripts appeared to be good potential biomarkers for high Se status.

Pathway and gene ontology analysis

Transcripts sets with significant DE associated with low Se, high Se, and supernutritional Se were used to identify homologous human genes that were assessed for enriched pathways and biological functions (Enrichr) [39]. For the low Se set containing 7 humanized genes, selenoprotein metabolism and Se nutrient network were nearly significant, with 0.05 < q < 0.10 (S3 Table), but with no significant enrichment of biological processes (S4 Table). Analysis of high Se treatment in experiment 2 with 12 humanized genes revealed no significant pathway enrichment (all q > 0.2) or biological function enrichment (q = 1). Lastly, moderate high Se at 1 μg Se/g, with 57 DE orthologous genes, had significant enrichment in proteins associated with cytosolic ribosomal proteins (q = 0.003) and nearly significant enrichment of regulation of glycolysis (q = 0.052), with significant biological functions including protein synthesis and targeting, nuclear transcribed RNA processing, and response to EGF stimulation; none of these biological functions were significant for high Se (2 and 5 μg Se/g) treatments.

GSEA

Gene set enrichment analysis (GSEA) is an approach that evaluates transcript expression data at the level of gene sets to detect changes in pathways and biological processes that are coordinated at a more subtle level than found by DE analysis of individual genes [40,41]. In a recent report, this approach identified three significant pathways (q < 0.05) in liver that had increased DE of transcripts in mice supplemented with high Se; these were Hallmark gene sets for cholesterol homeostasis, pancreatic β-cell signaling, and fatty acid metabolism [31].

GSEA with the turkey selenoprotein gene set (22 of 25 were present) found that the selenoprotein gene set was significantly down-regulated (q = 0.023) by 0 vs. 0.4 μg Se/g supplementation, with 13 transcripts significantly contributing to the leading edge (Fig 4A). GSEA, however, did not find significant regulation of the turkey selenoprotein gene set by 0.025 vs. 0.4 μg Se/g (q = 0.454). These analyses affirm the results as assessed by individual DE analysis as well as previous analysis by qPCR [8,10], and demonstrate the effectiveness of the GSEA approach for this turkey RNA Seq dataset.

Fig 4 — Shown are resulting enrichment plots (left) and fold-change heat maps (right) for resulting leading-edge genes. (A) Turkey selenogene set; (B) Hallmark angiogenesis gene set; (C) Hallmark hedgehog signaling gene set. Enrichment score is plotted vs. the rank in the ordered dataset of 11,612 humanized turkey transcripts (trimmed to include >14 counts, and normalized). Bars mark the leading-edge genes shown in the corresponding heat map. FDR q value shown for each plot.

Using the full Se-deficient turkey liver transcriptome, however, GSEA identified no pathways that were significantly increased (all q > 0.9) or significantly decreased (all q > 0.4) when comparing Se-deficient (0 μg Se/g) vs. Se-adequate (0.4 μg Se/g) transcripts amongst the 50 Hallmark gene sets (Table 4). For low-Se (0.025 μg Se/g) vs. Se-adequate (0.4 μg Se/g), there also were no significantly increased Hallmark transcript pathways (all q > 0.3); two Hallmark pathways (angiogenesis and hedgehog signaling) were significantly decreased (q = 0.109 and 0.136, respectively) by low-Se treatment (Fig 4B and 4C; Table 4). Analyses using the set of 186 KEGG gene sets yielded similar results for up-regulated transcripts (Se-deficient vs Se-adequate, all q > 0.9); there were two KEGG pathways (pentose phosphate and ether lipid metabolism) significantly increased (q = 0.234 and 0.249, respectively) by 0.025 μg Se/g treatment (S5 Table). GSEA identified 12 KEGG pathways significantly decreased by Se-deficiency (q < 0.25), but no pathways significantly decreased by 0.025 μg Se/g (all q > 0.3). This application of GSEA was readily able to detect altered liver selenotranscript expression in Se deficiency even in turkeys with good vitamin E supplementation, but there were almost no Hallmark gene sets significantly affected by Se deficiency, and only a few KEGG gene sets were significantly regulated by low-Se; there were no gene sets in either group significantly regulated by both Se-deficiency and low-Se. What about high Se?

Table 4. GSEA Se-deficient and low-Se most significant Hallmark gene sets^a.

Down-Regulated States or Processes	0 μg Se/g				0.025 μg Se/g
	No	ES	p-value	q-value	No	ES	p-value	q-value
ANGIOGENESIS	23	-0.41	0.162	0.467	24	-0.50	0.030	0.109
HEDGEHOG SIGNALING	22	-0.40	0.468	0.466	23	-0.65	0.000	0.136
EPITHELIAL MESENCHYMAL TRANSITION	106	-0.34	0.371	0.504	119	-0.46	0.149	0.253
TGF BETA SIGNALING	39	-0.41	0.057	0.546	39	-0.44	0.045	0.300
TNFA SIGNALING VIA NFKB	113	-0.37	0.240	0.453	115	-0.36	0.136	0.307
PANCREAS BETA CELLS	20	-0.56	0.000	0.792	20	-0.46	0.105	0.319
APICAL JUNCTION	99	-0.31	0.317	0.491	101	-0.39	0.175	0.320
IL2 STAT5 SIGNALING	123	-0.36	0.054	0.589	122	-0.31	0.168	0.320
KRAS SIGNALING UP	106	-0.37	0.189	0.514	109	-0.38	0.191	0.331
WNT BETA CATENIN SIGNALING	29	-0.33	0.451	0.474	29	-0.38	0.197	0.354
MYOGENESIS	99	-0.39	0.202	0.461	98	-0.39	0.279	0.384
UV RESPONSE DN	108	-0.43	0.025	0.572	108	-0.42	0.124	0.395
INFLAMMATORY RESPONSE	82	-0.42	0.000	0.481	82	-0.28	0.176	0.422
ANDROGEN RESPONSE	89	-0.35	0.227	0.429	89	-0.21	0.755	0.839
FATTY ACID METABOLISM	118	-0.34	0.194	0.437
CHOLESTEROL HOMEOSTASIS	49	-0.35	0.373	0.450	49	-0.33	0.425	0.598
Up-Regulated States or Processes	0 μg Se/g				0.025 μg Se/g
	No	ES	p-value	q-value	No	ES	p-value	q-value
XENOBIOTIC METABOLISM					124	0.42	0.032	0.337
IL6 JAK STAT3 SIGNALING	41	0.24	0.745	1	40	0.35	0.175	0.482
OXIDATIVE PHOSPHORYLATION					151	0.38	0.049	0.490
DNA REPAIR					106	0.31	0.325	0.500
MYC TARGETS V1					164	0.35	0.330	0.504
MTORC1 SIGNALING					157	0.35	0.177	0.533
UV RESPONSE UP					85	0.31	0.262	0.544
BILE ACID METABOLISM					72	0.35	0.145	0.548
GLYCOLYSIS					138	0.27	0.267	0.563
ADIPOGENESIS					142	0.31	0.112	0.632
HYPOXIA					141	0.24	0.458	0.652
FATTY ACID METABOLISM					118	0.34	0.144	0.678
PEROXISOME					73	0.26	0.587	0.678
HEME METABOLISM					141	0.25	0.457	0.683
ALLOGRAFT REJECTION					89	0.25	0.557	0.687

Open in a new tab

^aShown are the 15 Hallmark gene sets with the lowest q-values for 0 (middle) and 0.025 (right) vs. 0.4 μg Se/g for down- (top) and up-regulated (bottom) gene sets, along with number of genes found (No), enrichment score (ES) and unadjusted p-values. Also shown are selected additional gene sets.

For high-Se supplementation, GSEA identified no Hallmark or KEGG gene sets that were significantly increased (all q > 0.9) when comparing 0.75, 2, or 5 0 μg Se/g vs. Se-adequate (0.4 μg Se/g) transcripts (Table 5, S6 Table). For the Hallmark gene set, only supplementation with 1 μg Se/g yielded 7 significantly up-regulated sets, including fatty acid metabolism (q = 0.201, Fig 5), and none of the Hallmark gene sets were significantly down-regulated by any of the four super- plus high-Se treatments (all q >0.9). Amongst KEGG gene sets, there were no significantly up-regulated sets (q > 0.25), and none were significantly down-regulated by 0.75 and 5 μg Se/g (all q >0.3, S6 Table). Ribosome and glyoxylate & dicarboxylate metabolism gene sets were significantly down-regulated by 2 μg Se/g (q = 0.169 and 0.182, respectively, S2A and S2B Fig) but not by the other Se treatments. Tight junction pathway was significantly down regulated by 1 μg Se/g (q = 0.217, S2C Fig) but not by the other Se treatments. Just as in Se-deficiency, GSEA found that high Se also non-significantly decreased expression of the three gene sets reported to be increased significantly by high Se in mouse liver (S3 Fig). Overall, GSEA did identify a very limited number of gene sets that were significantly altered by one level of super- or high-Se supplementation, but there were no pathways amongst the Hallmark or KEGG gene sets that were consistently altered by Se supplementation above Se-adequate levels.

Table 5. GSEA high-Se most significant Hallmark gene sets^a.

Down-regulated States or Processes	0.75 μg Se/g				1 μg Se/g				2 μg Se/g				5 μg Se/g
	No	ES	p-value	q-value	No	ES	p-value	q-value	No	ES	p-value	q-value	No	ES	p-value	q-value
ANGIOGENESIS	23	-0.48	0.156	0.811	23	-0.53	0.148	0.378	23	-0.19	0.926	0.966
INFLAMMATORY RESPONSE	81	-0.25	0.527	0.688	84	-0.35	0.152	0.453	84	-0.23	0.646	0.891	84	-0.28	0.248	0.578
APICAL SURFACE	17	-0.23	0.972	0.922	17	-0.41	0.396	0.479	19	-0.28	0.689	0.823	18	-0.40	0.222	0.486
TNFA SIGNALING VIA NFKB	114	-0.37	0.202	0.679	116	-0.32	0.166	0.497	115	-0.25	0.577	0.800	114	-0.26	0.433	0.580
XENOBIOTIC METABOLISM	124	-0.18	0.720	0.816					124	-0.32	0.268	0.841	124	-0.31	0.187	0.500
HEDGEHOG SIGNALING	24	-0.34	0.413	0.697	22	-0.42	0.270	0.502					23	-0.42	0.051	0.789
MITOTIC SPINDLE	158	-0.32	0.298	0.690	159	-0.30	0.413	0.503	160	-0.25	0.591	0.851	160	-0.37	0.313	0.508
WNT BETA CATENIN SIGNALING	29	-0.43	0.198	0.753	29	-0.49	0.089	0.507	29	-0.28	0.736	0.790	29	-0.52	0.000	1.000
IL2 STAT5 SIGNALING	124	-0.33	0.000	0.824	123	-0.27	0.311	0.511	126	-0.26	0.294	0.827	125	-0.23	0.559	0.568
PEROXISOME	73	-0.22	0.612	0.708					74	-0.35	0.240	0.839	73	-0.37	0.206	0.512
INTERFERON ALPHA RESPONSE	51	-0.27	0.559	0.690	51	-0.33	0.309	0.569	52	-0.28	0.568	0.891	52	-0.38	0.343	0.518
INTERFERON GAMMA RESPONSE	109	-0.31	0.326	0.674	110	-0.31	0.352	0.521	112	-0.30	0.428	0.865	113	-0.34	0.099	0.602
MYC TARGETS V1	164	-0.47	0.000	0.775					164	-0.42	0.152	0.994	164	-0.39	0.407	0.529
PI3K AKT MTOR SIGNALING	72	-0.31	0.189	0.664	72	-0.27	0.490	0.570	73	-0.31	0.408	0.779	72	-0.35	0.259	0.541
ADIPOGENESIS	142	-0.27	0.235	0.665					142	-0.36	0.100	1.000	142	-0.30	0.261	0.550
PANCREAS BETA CELLS	19	-0.44	0.193	0.869	20	-0.39	0.118	0.551	19	-0.29	0.713	0.795	18	-0.36	0.335	0.585
FATTY ACID METABOLISM	118	-0.22	0.565	0.673					118	-0.37	0.167	0.716	117	-0.23	0.689	0.762
CHOLESTEROL HOMEOSTASIS	49	-0.15	0.952	1.000					49	-0.27	0.597	0.805	49	-0.14	0.970	0.985
Up-regulated States or Processes	0.75 μg Se/g				1 μg Se/g				2 μg Se/g				5 μg Se/g
	No	ES	p-value	q-value	No	ES	p-value	q-value	No	ES	p-value	q-value	No	ES	p-value	q-value
XENOBIOTIC METABOLISM					124	0.50	0.000	0.144
MTORC1 SIGNALING					157	0.46	0.028	0.177
OXIDATIVE PHOSPHORYLATION					151	0.52	0.028	0.180
ADIPOGENESIS					142	0.39	0.000	0.187
FATTY ACID METABOLISM					118	0.50	0.000	0.201
REACTIVE OXYGEN SPECIES PATHWAY					38	0.46	0.063	0.231
PEROXISOME					74	0.43	0.061	0.249
UV RESPONSE UP					85	0.39	0.026	0.278
UNFOLDED PROTEIN RESPONSE					86	0.38	0.109	0.306
BILE ACID METABOLISM					72	0.38	0.098	0.326
GLYCOLYSIS					138	0.35	0.101	0.335
HYPOXIA					125	0.30	0.136	0.416
CHOLESTEROL HOMEOSTASIS					51	0.28	0.562	0.699
MYC TARGETS V1					164	0.31	0.603	0.722
ANDROGEN RESPONSE					76	0.24	0.581	0.745
ANGIOGENESIS													23	0.16	0.970	0.947
EPITHELIAL MESENCHYMAL TRANSITION									122	0.20	0.913	1.000
HEDGEHOG SIGNALING									24	0.22	0.837	1.000
MYOGENESIS									101	0.11	0.978	1.000

Open in a new tab

^aShown are the 15 Hallmark gene sets with the lowest q-values for 0.75, 1, 2, and 5 vs. 0.4 μg Se/g treatment, along with number of genes found (No), enrichment score (ES) and unadjusted p-values. Also shown are selected additional gene sets.

Fig 5 — **0.4 μg Se/g.** Shown are resulting enrichment plots from use of the 50 Hallmark gene sets. (A) Fatty Acid Metabolism; (B) Xenobiotic Metabolism; (C) Oxidative Phosphorylation; (D) mTORC1 Signaling; (E) Adipogenesis; (F) Reactive Oxygen Species Pathway; (G) Peroxisome; (H) Cholesterol Homeostasis. Bars mark the leading-edge genes. FDR q value shown for each plot.

GSEA was also conducted using the total MSigDB gene ontology collection of 9996 gene sets for biological processes, cellular components, and molecular functions, comparing each dietary Se treatment vs. Se-adequate in experiment 1 and 2. GSEA GO analysis found no significant gene sets when transcripts for 0, 0.025, 0.75, 2 and 5 μg Se/g were compared to Se-adequate (lowest q values: 0.609, 0.369, 0.485, 0.475, 0.484, respectively, S7 Table). For 1 μg Se/g, only 2 gene sets were significant: amide transmembrane transporter activity (q = 0.243) and cellular respiration (q = 0.248). Thus GO analysis by GSEA also found virtually no processes and biological functions consistently altered by high Se. Because Se is readily metabolized by sulfur metabolism enzymes (see Discussion), the GO results from GSEA were filtered for gene sets related to sulfur or thiol metabolism. The result was 21 GO gene sets that appeared in at least 1 comparison (S7 Table). Except for the 1 μg Se/g transcript set, no GO gene set was significant (q > 0.490); for 1 μg Se/g, glutathione metabolic process was non-significantly up-regulated (q = 0.290) but non-significantly down-regulated for 2 and 5 μg Se/g (q ≥ 0.49). Clearly, changes in Se status did not significantly affect sulfur-related transcript expression in this study.

The GO results from GSEA were also filtered for gene sets related to oxidative stress. None of the eight oxidative stress gene sets was significant for any of the pair-wise comparison with Se-adequate treatment (all q > 0.480) (S8 Table).

Discussion

Overall gene expression effects

When Se-deficient and low-Se poults (0 and 0.025 μg Se/g) were compared to Se-adequate poults (0.4 μg Se/g), the majority of the mapped selenotranscripts were down-regulated in Se-deficiency, consistent with our previous reports [7]. These results are very similar to results in rats and mice fed Se-deficient diets as assessed using Affymetrix microarray analysis [26]. When poults fed high dietary Se (2 and 5 μg Se/g) were compared to Se-adequate poults (0.4 μg Se/g), only 2 and 14 transcripts, out of the 17,100 transcripts were DE. These small sets of DE transcript are similar in scope to the minimal effects observed when rats were fed 2 μg Se/g [16]. In the rat study, however, a 20% reduction in growth was observed when rats were fed 5 μg Se/g, which was associated with significant DE of over 1100 transcripts (4% of the transcriptome) [16]. These small sets of DE transcripts also contrast with studies in turkeys treated with a known toxin, aflatoxin, which caused differential expression of 2% of the shared transcripts [42]. In the present turkey study, where there were no signs of overt toxicity, there were few significant DE transcripts with both 2 and 5 μg Se/g, showing that the turkey poult adapts to high Se status without major changes in gene expression.

As illustrated in the volcano plots (Fig 2) and cluster analysis (Fig 3), there were few large fold differences in gene expression observed in these studies. In Se-deficiency, all significant transcript changes were < |±3|-fold. Similarly, with 5 μg Se/g, all significant transcript changes were < |±2.2|-fold. The paucity of large changes in differential gene expression further indicates that dramatic changes in gene expression do not play a role in how the turkey adapts to high Se status.

Se-deficient and low-Se transcript expression

Se deficiency in this study resulted in significantly altering expression of 10 transcripts (0.06% of the mapped transcripts, q < 0.05), similar to what we found in rodents. Remapping against updated selenoprotein transcripts found 4 selenotranscripts significantly down-regulated by Se-deficiency (q < 0.05), with a total of 12 selenotranscripts down-regulated with a q < 0.2, similar to our previous qPCR analysis [7]. GSEA similarly found that a set of 13 selenotranscripts were significantly down-regulated in Se-deficient vs. Se-adequate liver; the selenotranscript gene set was not significantly down-regulated by 0.025 μg Se/g in GSEA. In previous studies in rats and mice, only 4 and 5 liver transcripts, respectively–all selenoproteins–were down-regulated by Se deficiency, and only 2 and 3 transcripts, respectively, were up regulated [16,26]. Clearly Se deficiency in this study results in major effects on expression of a subset but not all of the selenotranscripts, just as reported in most studies in poultry, rodents, and other species.

Analysis by DE, gene ontology, or GSEA, however, did not reveal a focused pattern of metabolic changes due to Se deficiency. The three non-selenoprotein DE transcripts down-regulated by Se-deficiency were HSPB11, an intraflagellar transport protein 25 homolog, SGTA, a small glutamine-rich tetratricopeptide repeat-containing protein alpha, and RASGRP3 (which was also significantly down-regulated by high Se); significantly up-regulated transcripts by Se-deficiency or low Se were SOCS3, a negative regulator of cytokine signaling through the JAK-STAT pathway which responds to a wide variety of stimuli [43], LOC104909734, encoding a sulfotransferase family cytosolic 2B member 1-like, LOC100545600, encoding a C-C motif chemokine 3-like protein, and BLCAP, encoding a bladder cancer associated protein (which was also significantly up-regulated by high Se). None of these transcripts were found to be associated with enriched GO pathways (S3 Table) or biological processes (S4 Table). Similarly, GSEA found no Hallmark gene sets were significantly regulated up or down by Se deficiency, and no KEGG gene sets were significantly up-regulated by Se deficiency. GSEA did yield 10 KEGG states and processes that were significantly down-regulated in Se-deficiency (q ≤ 0.25), but these gene sets were not significantly affected by 0.025 μg Se/g (S5 Table). GSEA did identify two Hallmark gene sets (angiogenesis and hedgehog signaling, Table 4) that were significantly down-regulated by 0.025 μg Se/g but not by 0 μg Se/g. These diverse states and processes, including angiogenesis, signaling, infection, disease states, and metabolic pathways, do not seem to reveal a unified change in metabolism, suggesting that these changes are more associated with an assortment of unhealthy biological conditions that are not unique to Se deficiency but rather are downstream effects due to loss of selenoprotein activity.

High Se transcript expression

High dietary Se resulted in only small sets of 2 and 14 significant DE transcript in poults fed 2 or 5 μg Se/g in experiment 2, respectively. This is similar to what we observed in rats fed 2 μg Se/g, where only 5 and 1 transcripts were up- and down-regulated, respectively [16]; none of the genes found in rats overlap with the significantly regulated genes in the present turkey study. The turkey gene sets contain one transcript, BLCAP, which showed an apparent U-shaped response, and one transcript, RASGRP3 that showed an inverted-U-shaped response across the spectrum from Se-deficient to high-Se (Tables 2 and 3); each of these changes was only significant in one of the two high-Se treatments. RASGRP3 was reported to be to be up-regulated 1.65X in rats fed 5 μg Se/g diet [16], not down-regulated as in the turkey; RASGRP3 binds to dynein light chain 1 (DLC1/DYNLL1, LOC104913587 in the turkey) [44] and is an NRF2 (NFE2L2) target, suggesting that these U-shaped responses may be associated with down-stream effects of Se excess or Se deficiency, perhaps due to increased reactive oxygen species at either extreme of Se status [16]. Just as with Se-deficiency, none of the significant DE transcripts were found to be significantly enriched in GO pathways (S3 Table) or biological processes (S4 Table). Remarkably, the GSEA approach also found no Hallmark or KEGG gene sets significantly regulated by 2 or 5 μg Se/g vs. 0.4 μg Se/g in experiment 2. Thus by individual transcript DE, by GO, or by GSEA, high dietary Se had minimal effect on the hepatic turkey transcriptome, and no individual genes or pathways were found that could consistently be used as biomarkers for high Se or that suggested a role in adaptation to high Se in the turkey.

Supernutritional Se expression

Treatment with supernutritional levels of dietary Se (0.75 and 1 μg Se/g) resulted in disparate results. With 0.75 μg Se/g, there were no significant DE transcripts in pairwise analysis relative to Se-adequate treatment. The minimal number of genes with big fold changes in the 0.75 μg Se/g group shown by cluster analysis (Fig 3, q < 0.1), by DE in pairwise comparison (S2 Table, q < 0.05), and by a lack of enriched GSEA gene sets, as well as by differences in growth and selenoenzyme expression (Table 1), indicate strongly that feeding 0.75 μg Se/g as selenite to the young turkey poult has no adverse effects as compared to feeding Se-adequate diets (0.4 μg Se/g).

The incremental increase from 0.75 to 1 μg Se/g in experiment 1 did affect transcript DE as there were 50 up and 38 down significant DE transcripts (S2 Table) with 1 vs. 0.4 μg Se/g supplementation, respectively. As discussed in Results and shown in Table 3, none of these transcripts were consistently altered by 2 and 5 μg Se/g, and some were also DE in the same direction by 0.025 or 0 μg Se/g (Tables 2 and 3), indicating that changes in DE of these transcripts were not uniquely associated with high Se status. As discussed in Results, GSEA found no significant KEGG gene sets for 1 vs 0.4 μg Se/g supplementation (S6 Table), but found significant increases for 6 Hallmark gene sets, including fatty acid metabolism and cholesterol homeostasis, 2 of the 3 up-regulated pathways reported in liver in mice supplemented with high Se [31]. Fatty acid metabolism, however, was non-significantly down-regulated (q >0.8) with 0.75, 2 and 5 μg Se/g treatment in this turkey study. GSEA on 1 vs 0.4 μg Se/g also found no significant down-regulated Hallmark gene sets (the pancreatic β-cell set was down-regulated non-significantly, q = 551). The only significant down-regulated KEGG gene set was the tight junction set (q = 0.217), which was not significantly regulated by 0.75, 2 or 5 μg Se/g. Thus this study did not identify one or more over-arching processes–including lipid and energy metabolism–that were consistently associated with supernutritional or high Se supplementation in turkey poults.

Turkeys vs. rodents

As discussed above, a recent study found that GSEA identified lipid and energy metabolism as pathways in mouse liver that were significantly up-regulated by high Se (equivalent to ~3 μg Se/g diet) [31]. As reviewed above, the present turkey study provides little evidence that up-regulation of lipid and energy metabolism is associated with high Se status in the turkey. We also compared DE transcript sets for 2 and 5 μg Se/g vs. Se-adequate in the present turkey study with rat liver DE transcript sets (2 μg Se/g vs. Se-adequate) in our previous study [16]. Even when these DE transcript sets were expanded to transcript sets with unadjusted p < 0.05, there were no transcript changes associated with high Se status that were in-common between turkey poults and rats, thus providing additional support for our contention that animals do not adapt to high Se status by altering gene expression (manuscript in preparation). Reduced susceptibility to Se toxicity by the turkey as well as animal age, supplementation with high vitamin E, and duration of high Se treatment, are all likely to have contributed to the lack of common DE transcripts between rodents and turkeys in these two studies.

Additional transcriptomics studies

A number of previous studies have assessed the effect of Se status on the full transcriptome in different species. Many of these studies, however, compared animals fed Se-adequate diets vs. Se-deficient or low-Se diets. Three of these studies in rat liver [45] and mouse liver [46,47] reported <50 significant DE transcripts q < 0.05 with selenite-supplementation vs. low Se treatment, but two studies in rat heart [48] and in chick oviduct [49] reported 4931 and 2196 transcripts, respectively, with differential expression. The cause of these elevated numbers of DE transcripts is unclear but may be due to differences in species, tissue, age of the animals, and diet including vitamin E supplementation (see below). Other Se transcriptomics studies involved other factors such as aged mice/Se-enriched rice [50], diabetic mice [51], immune challenge in trout [52], transcripts in offspring after high-fat feeding to mouse dams [53], or Se-enriched worms fed to trout [54]. In livestock, high Se supplementation as selenized yeast to cattle [55] and sheep [56] relative to Se-adequate controls reported 139 and 1186 DE transcripts in cattle liver and sheep blood, respectively. Interestingly, the cattle study used 2 different forms of Se; filtering the transcripts to those DE with both Se forms trimmed the Se-responsive set from 139 to down to 33 DE transcripts [55]. Most of the above studies only compared two treatments, whereas our study used filtering individual transcripts sets for significant transcripts DE altered in the same direction in the 5, 2 and 1 μg Se/g datasets. The result was a paucity of high-Se responsive transcripts.

Vitamin E

Our studies in the turkey and the rat were focused on characterizing the impact of changes just due to Se status. Thus the diets used in these studies were supplemented with 12.5X (turkey) and 2X (rat) the NRC requirements for vitamin E. Older studies nicely demonstrated that the Se requirement for prevention of gizzard myopathy in turkeys was increased from 0.18 to 0.28 μg Se/g when poults were not supplemented with vitamin E [57]. In our previous rat study, feeding 5 μg Se/g resulted in 20% growth depression and significant DE of 4% of the transcriptome. A substantial number of these transcripts were NRF2 (NFE2L2) targets indicating that increased oxidative stress was associated with the effects of Se toxicity [16]. In the present turkey study, however, GSEA did not find any significant GO gene sets related to oxidative stress (S8 Table).

There are many studies in the literature reporting that expression of all or nearly all of the selenoprotein transcripts are significantly decreased in chickens by Se deficiency [58–59]. These studies, however, typically did not supplement the diets with vitamin E such that clinical and sub-clinical disease associated with the combined loss of Se and vitamin E likely underlie the substantial number of reported changes in gene expression. Other studies using large numbers of chicks fed Se-deficient corn-soy diets did find significant decreases in growth in Se-deficient chicks supplemented with 1X to 5X the NRC chicken vitamin E requirement [60–61]; these studies reported that 7–9 selenotranscripts were significantly decreased by Se deficiency. Supernutritional Se supplementation studies in chicks fed corn-soy diets have reported that high dietary Se decreases [62] or increases growth [63] along with increasing expression of subsets of selenotranscripts. Our studies in the young chick [18] and turkey poult [7] only found significant changes in 8 and 4 selenoprotein transcripts out of 24, respectively, in the livers of birds fed Se-deficient torula yeast diets for 4 weeks. Differences in species, diet, and vitamin E supplementation clearly contribute to differences in growth and selenotranscript expression. For the present study, diets were supplemented with elevated dietary vitamin E so that we could study the effect of high dietary Se on transcript expression without disease complications, thus focusing on the molecular effects of Se status alone.

Other possibilities

The paucity of transcription changes associated with Se status raises several interesting alternative hypotheses. The large number of DE transcripts (S2 Table) and the seven significant Hallmark gene sets (Table 5) for 1 but not 0.75 or 2 μg Se/g might suggest that that 1 μg Se/g tweaks metabolism sufficiently to alter transcription of a number of genes, but that by 2 μg Se/g, Se metabolism readjusts to cope with the increased incoming Se. The smaller changes in liver Se concentrations in turkeys fed 0.4 to 2 μg Se/g as compared to 5 μg Se/g (Table 1), however, suggests that these DE transcripts are not the driving force used to adapt to additional dietary Se over this range. One attractive hypothesis is the “response to high Se” gene set is simply missing from Hallmark and KEGG gene sets; this is illustrated by GSEA using the “turkey selenoprotein” gene set which was highly significant, q = 0.023, and thus far more significant than any of the curated MSigDB gene sets. In part, these hypothetical genes involved in response to high Se may be amongst the uncharacterized LOC genes in the turkey as well as human transcriptome.

A second hypothesis is that Se is metabolized via other pathways where transcript expression is unaltered by Se status; there may have been little evolutionary pressure in mammals to enhance a pathway (gene set) to cope with high Se, perhaps mainly because Se is readily metabolized by sulfur metabolizing enzymes in animals [64] and plants [65]. Se, present at concentrations in animals over a 1000-fold less than for sulfur, simply “hitch-hikes” along these sulfur pathways, leading to formation of selenium analogs of sulfur metabolites, mixed selenodisulfides, and even a selenosugar (seleno-N-acetyl-galactosamine) in animals [66–69]. In yeast and plants, this results in synthesis of selenomethionine from inorganic Se, but this does not occur in monogastric animals [65,70,71]. In the present study, however, there was only 1 sulfur metabolism transcript with significant DE (S2 Table); LOC104909734, a sulfotransferase, was increased >2-fold in 0 vs. 0.4 μg Se/g, but not DE by high Se treatment. GSEA similarly did not find any significant gene ontology gene sets for 2 or 5 μg Se/g treatment related to sulfur/thiol metabolism (S7 Table). In other studies, we have evidence that Se accumulates in turkeys fed 5 μg Se/g as a selenosugar linked to low molecular weight thiols as well as general body proteins [71]. The pathway for formation of the selenosugar or a homologous sulfur-sugar has not been identified.

Conclusions

These experiments studied the effect of graded levels of dietary Se fed to turkey poults in diets supplemented with vitamin E so that direct effects of altered Se status could be studied. The study affirmed that the major effect of Se-deficiency is to down-regulate expression of a subset of selenoprotein transcripts in liver, with little significant effect on general transcript expression. In response to high Se intake (2 and 5 μg Se/g) relative to Se-adequate turkeys, there were only a limited number of significant DE transcripts, all with only relatively small fold-changes; no transcript showed a consistent pattern of altered expression in response to high Se intakes across the 1, 2 and 5 μg Se/g treatments, and there were no associated metabolic pathways and biological functions that were significant and consistently found with high Se supplementation. GSEA found no gene sets that were consistently altered by high-Se and supernutritional-Se. A comparison of these DE transcript sets with those identified in high Se transcript sets parallel studies conducted in rats (2 μg Se/g) and in mice (~3 μg Se/g) fed high Se also failed to identify common DE transcripts between these three species. Collectively, this study indicates that turkeys do not alter gene expression as a homeostatic mechanism to adapt to high Se. Targeting the nature of metabolites present in liver of turkeys fed high Se has the potential to identify biomarkers for high Se status and to suggest metabolic pathways used by turkeys to adapt to high Se intake.

Supporting information

S1 Fig. GSEA of liver transcripts in turkeys fed Se-deficient vs Se-adequate diet.

Shown are resulting enrichment plots (left) and fold-change heat maps (right) for resulting leading-edge genes. (A) Hallmark cholesterol homeostasis gene set; (B) Hallmark pancreatic β-cell gene set; (C) Hallmark fatty acid metabolism gene set. Enrichment score is plotted vs. the rank in the ordered dataset of 11,612 humanized turkey transcripts (trimmed to include >14 counts, and normalized). Bars marks the leading-edge genes shown in the corresponding heat map. FDR q value shown for each plot.

(PDF)

Click here for additional data file.^{(1,011.1KB, pdf)}

S2 Fig. GSEA of liver transcripts in turkeys fed 2 or 1 μg Se/g vs. 0.4 μg Se/g.

Shown are resulting enrichment plots (left) and fold-change heat maps (right) for resulting leading-edge genes. (A) KEGG ribosome gene set for 2 μg Se/g; (B) KEGG glyoxylate and dicarboxylate gene set for 2 μg Se/g; (C) KEGG tight junction gene set for 1 μg Se/g; Enrichment score is plotted vs. the rank in the ordered dataset of 11,612 humanized turkey transcripts (trimmed to include >14 counts, and normalized). Bars mark the leading-edge genes shown in the corresponding heat map. FDR q value shown for each plot.

(PDF)

Click here for additional data file.^{(988.6KB, pdf)}

S3 Fig. GSEA of liver transcripts in high-Se turkeys fed 5 μg Se/g vs. 0.4 μg Se/g.

Shown are resulting enrichment plots (left) and fold-change heat maps (right) for resulting leading-edge genes. (A) Hallmark cholesterol homeostasis gene set; (B) Hallmark pancreatic β-cell gene set; (C) Hallmark fatty acid metabolism gene set. Enrichment score is plotted vs. the rank in the ordered dataset of 11,612 humanized turkey transcripts (trimmed to include >14 counts, and normalized). Bar marks the leading-edge genes shown in the corresponding heat map. FDR q value shown for each plot.

(PDF)

Click here for additional data file.^{(988.6KB, pdf)}

S1 Table. Gene IDs and functions of significant DE transcripts by 0, 0.025, 2, and 5 μg Se/g.

(PDF)

Click here for additional data file.^{(28.9KB, pdf)}

S2 Table. Transcripts differentially expressed by 0.75 and 1.0 μg Se/g vs. Se-adequate (Exp 1).

(PDF)

Click here for additional data file.^{(76.5KB, pdf)}

S3 Table. Enriched gene ontology pathways.

(PDF)

Click here for additional data file.^{(61.6KB, pdf)}

S4 Table. Enriched gene ontology biological processes.

(PDF)

Click here for additional data file.^{(55.4KB, pdf)}

S5 Table. GSEA Se-deficient and low-Se most significant KEGG gene sets.

(PDF)

Click here for additional data file.^{(80.7KB, pdf)}

S6 Table. GSEA high-Se most significant KEGG gene gets.

(PDF)

Click here for additional data file.^{(81.6KB, pdf)}

S7 Table. Effect of dietary Se vs. Se-adequate on GO biological processes, GO cellular components, and GO molecular functions involving sulfur or thiols, from GSEA for 9996 gene sets.

(PDF)

Click here for additional data file.^{(33.9KB, pdf)}

S8 Table. Effect of dietary Se vs. Se-adequate on GO biological processes, GO cellular components, and GO molecular functions involving oxidative stress, from GSEA for 9996 gene sets.

(PDF)

Click here for additional data file.^{(33.1KB, pdf)}

S9 Table. Human symbols raw counts.

(XLSX)

Click here for additional data file.^{(2.2MB, xlsx)}

Data Availability

The original RNA Seq data are deposited in NIH's NCBI’s Sequence Read Archive (SRA) and are accessible through SRA accession PRJNA614544 (available at https://www.ncbi.nlm.nih.gov/sra/PRJNA614544). Additional relevant data are within the manuscript and its Supporting Information files.

Funding Statement

This Research was supported by the National Institute of Food and Agriculture, United States Department of Agriculture (accession number 1016808, at: https://portal.nifa.usda.gov/web/crisprojectpages/1016808-advanced-technologies-for-the-genetic-improvement-of-poultry.html), Hatch project 1004389 to RAS, by the Wisconsin Alumni Foundation (http://www.uwalumni.com) Selenium Nutrition Research Fund No. 12046295 to RAS, and by the USDA National Institute of Food and Agriculture Animal Genome Program project 2013-01043 to KMR.

References

1.Rosenfeld I, Beath OA. Selenium poisoning in animals In: Selenium: Geobotany, Biochemistry, Toxicity and Nutrition. New York: Academic Press; 1964. pp. 142–213. [Google Scholar]
2.Nelson AA, Fitzhugh OG, Calvery HO. Liver tumors following cirrhosis caused by selenium in rat. Cancer Res. 1943; 3:230–236 [Google Scholar]
3.Schwarz K, Foltz CM. Selenium as an integral part of factor 3 against dietary necrotic liver degeneration. J Am Chem Soc. 1957; 79:3292–3293 [PubMed] [Google Scholar]
4.Scott ML, Bieri JG, Briggs GM, Schwarz K. Prevention of exudative diathesis by Factor 3 in chicks on vitamin E-deficient Torula yeast diets. Poultry Sci. 1957; 36:1155. [Google Scholar]
5.Chen X, Yang G, Chen J, Wen Z, Ge K. Studies on the relations of selenium and Keshan disease. Biol Tr Elem Res. 1980; 2:91–107 [DOI] [PubMed] [Google Scholar]
6.Rotruck JT, Pope AL, Ganther HE, Swanson AB, Hafeman DG, Hoekstra WG. Selenium: biochemical role as a component of glutathione peroxidase. Science. 1973; 179:588–590. 10.1126/science.179.4073.588 [DOI] [PubMed] [Google Scholar]
7.Sunde RA, Li JL, Taylor RM. Insights for setting of nutrient requirements, gleaned by comparison of selenium status biomarkers in turkeys and chickens versus rats, mice, and lambs. Adv Nutr. 2016; 7:1129–1138. 10.3945/an.116.012872 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Taylor RM, Sunde RA. Selenoprotein transcript level and enzyme activity as biomarkers for selenium status and selenium requirements of turkeys (Meleagris gallopavo). PLoS ONE. 2016; 11:e0151665 10.1371/journal.pone.0151665 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Taylor RM, Sunde RA. Selenium requirements based on muscle and kidney selenoprotein enzyme activity and transcript level in the turkey poult (Meleagris gallopavo). PLoS ONE. 2017; 12:e0189001 10.1371/journal.pone.0189001 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Taylor RM, Bourget VG, Sunde RA. High dietary inorganic selenium has minimal effects on turkeys and selenium status biomarkers. Poult Sci. 2019; 98:855–865. 10.3382/ps/pey413 [DOI] [PubMed] [Google Scholar]
11.Barnes KM, Evenson JK, Raines AM, Sunde RA. Transcript analysis of the selenoproteome indicates that dietary selenium requirements in rats based on selenium-regulated selenoprotein mRNA levels are uniformly less than those based on glutathione peroxidase activity. J Nutr. 2009; 139:199–206. 10.3945/jn.108.098624 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Labunskyy VM, Hatfield DL, Gladyshev VN. Selenoproteins: molecular pathways and physiological roles. Physiol Rev. 2014; 94:739–777. 10.1152/physrev.00039.2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Kryukov GV, Castellano S, Novoselov SV, Lobanov AV, Zehtab O, Guigo R et al. Characterization of mammalian selenoproteomes. Science (Washington, DC). 2003; 300:1439–1443. 10.1126/science.1083516 [DOI] [PubMed] [Google Scholar]
14.Sunde RA, Sunde GR, Sunde CM, Sunde ML, Evenson JK. Cloning, sequencing, and expression of selenoprotein transcripts in the turkey (Meleagris gallopavo). PLoS ONE. 2015; 10:e0129801 10.1371/journal.pone.0129801 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Gladyshev VN, Arner ES, Berry MJ, Brigelius-Flohe R, Bruford EA, Burk RF et al. Selenoprotein gene nomenclature. J Biol Chem. 2016; 291:24036–24040. 10.1074/jbc.M116.756155 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Raines AM, Sunde RA. Selenium toxicity but not deficient or super-nutritional selenium status vastly alters the transcriptome in rodents. BMC Genomics. 2011; 12:26 10.1186/1471-2164-12-26 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Nesheim MC, Scott ML. Studies of the nutritive effects of selenium for chicks. J Nutr. 1958; 65:601–618. 10.1093/jn/65.4.601 [DOI] [PubMed] [Google Scholar]
18.Li JL, Sunde RA. Selenoprotein transcript level and enzyme activity as biomarkers for selenium status and selenium requirements of chickens (Gallus gallus). PLoS ONE. 2016; 11:e0152392 10.1371/journal.pone.0152392 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Walter ED, Jensen LS. Effectiveness of selenium and non-effectiveness of sulfur amino acids in preventing muscular dystrophy in the turkey poult. J Nutr. 1963; 80:327–331. 10.1093/jn/80.3.327 [DOI] [PubMed] [Google Scholar]
20.Tinsley IJ, Harr JR, Bone JF, Weswig PH, Yamamoto RS. Selenium toxicity in rats. I. Growth and longevity In: Selenium in Biomedicine. Westport CN: AVI Publishing Co; 1967. pp. 141–152. [Google Scholar]
21.Harr JR, Bone JF, Tinsley IJ, Weswig PH, Yamamoto RS. Selenium toxicity in rats. II. Histopathology In: Muth OH, Oldfield JE, Weswig PH, editors. Selenium in Biomedicine. Westport, CT: AVI Publishing Co; 1967. pp. 153–178. [Google Scholar]
22.El-Begearmi MM, Combs GF Jr. Dietary effects on selenite toxicity in the chick. Poult Sci. 1982; 61:770–776. 10.3382/ps.0610770 [DOI] [PubMed] [Google Scholar]
23.McAdam PA, Levander OA. Chronic toxicity and retention of dietary selenium fed to rats as D- or L-selenomethionine, selenite, or selenate. Nutr Res. 1987; 7:601–610. [Google Scholar]
24.Whanger PD, Butler JA. Effects of various dietary levels of selenium as selenite or selenomethionine on tissue selenium levels and glutathione peroxidase activity in rats. J Nutr. 1988; 118:846–852 10.1093/jn/118.7.846 [DOI] [PubMed] [Google Scholar]
25.Salbe AD, Levander OA. Effect of various dietary factors on the deposition of selenium in the hair and nails of rats. J Nutr. 1990; 120:200–206 10.1093/jn/120.2.200 [DOI] [PubMed] [Google Scholar]
26.Sunde RA, Raines AM, Barnes KM, Evenson JK. Selenium status highly-regulates selenoprotein mRNA levels for only a subset of the selenoproteins in the selenoproteome. Biosci Rep. 2009; 29:329–338. 10.1042/BSR20080146 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Maret W. Zinc in Cellular Regulation: The Nature and Significance of "Zinc Signals". Int J Mol Sci. 2017; 18:ijms18112285 10.3390/ijms18112285 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Muckenthaler MU, Rivella S, Hentze MW, Galy B. A Red Carpet for Iron Metabolism. Cell. 2017; 168:344–361. 10.1016/j.cell.2016.12.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Kim BE, Nevitt T, Thiele DJ. Mechanisms for copper acquisition, distribution and regulation. Nat Chem Biol. 2008; 4:176–185. 10.1038/nchembio.72 [DOI] [PubMed] [Google Scholar]
30.Boehler CJ, Raines AM, Sunde RA. Toxic-selenium and low-selenium transcriptomes in Caenorhabditis elegans: toxic selenium up-regulates oxidoreductase and down-regulates cuticle-associated genes. PLoS ONE. 2014; 9:e101408 10.1371/journal.pone.0101408 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Hu X, Chandler JD, Orr ML, Hao L, Liu K, Uppal K et al. Selenium supplementation alters hepatic energy and fatty acid metabolism in mice. J Nutr. 2018; 148:675–684. 10.1093/jn/nxy036 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.National Research Council. Nutrient Requirements of Poultry. Washington D.C.: National Academy Press; 1994. 155 p. [Google Scholar]
33.Brown MB, Forsyth AB. Robust tests for equality of variances. J Amer Stat Assn. 1974; 69:364–367. 10.1080/01621459.1974.10482955 [DOI] [Google Scholar]
34.Steel R. G. D. and Torrie J. H. Principles and Procedures of Statistics. New York, NY: McGraw-Hill Book; 1960. [Google Scholar]
35.Reed KM, Mendoza KM, Abrahante JE, Barnes NE, Velleman SG, Strasburg GM. Response of turkey muscle satellite cells to thermal challenge. I. transcriptome effects in proliferating cells. BMC Genomics. 2017; 18:352–3740. 10.1186/s12864-017-3740-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Lê S, Josse J, Husson F. An R package for multivariate analysis. J Stat Softw. 2008; 25:1–18 [Google Scholar]
37.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15:550–0550. 10.1186/s13059-014-0550-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Saldanha AJ. Java Treeview—extensible visualization of microarray data. Bioinformatics. 2004; 20:3246–3248. 10.1093/bioinformatics/bth349 [DOI] [PubMed] [Google Scholar]
39.Kuleshov MV, Jones MR, Rouillard AD, Fernandez NF, Duan Q, Wang Z et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acid Res. 2016; W44:W90–97. https://amp.pharm.mssm.edu/Enrichr/ [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005; 102:15545–15550. 10.1073/pnas.0506580102 [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Mootha VK, Lindgren CM, Eriksson KF, Subramanian A, Sihag S, Lehar J et al. PGC-1-alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet. 2003; 34:267–273. 10.1038/ng1180 [DOI] [PubMed] [Google Scholar]
42.Monson MS, Settlage RE, McMahon KW, Mendoza KM, Rawal S, El-Nezami HS et al. Response of the hepatic transcriptome to aflatoxin B1 in domestic turkey (Meleagris gallopavo). PLoS ONE. 2014; 9:e100930 10.1371/journal.pone.0100930 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Yin Y, Liu W, Dai Y. SOCS3 and its role in associated diseases. Hum Immunol. 2015; 76:775–780. 10.1016/j.humimm.2015.09.037 [DOI] [PubMed] [Google Scholar]
44.Okamura SM, Oki-Idouchi CE, Lorenzo PS. The exchange factor and diacylglycerol receptor RasGRP3 interacts with dynein light chain 1 through its C-terminal domain. J Biol Chem. 2006; 281:36132–36139. 10.1074/jbc.M605093200 [DOI] [PubMed] [Google Scholar]
45.Bosse AC, Pallauf J, Hommel B, Sturm M, Fischer S, Wolf NM et al. Impact of selenite and selenate on differentially expressed genes in rat liver examined by microarray analysis. Biosci Rep. 2010; 30:293–306. 10.1042/BSR20090089 [DOI] [PubMed] [Google Scholar]
46.Mallonee DH, Crowdus CA, Barger JL, Dawson KA, Power RF. Use of stringent selection parameters for the identification of possible selenium-responsive marker genes in mouse liver and gastrocnemius. Biol Trace Elem Res. 2011; 143:992–1006. 10.1007/s12011-010-8894-8 [DOI] [PubMed] [Google Scholar]
47.Tsuji PA, Carlson BA, Anderson CB, Seifried HE, Hatfield DL, Howard MT. Dietary selenium levels affect selenoprotein expression and support the interferon-gamma and IL-6 immune response pathways in mice. Nutrients. 2015; 7:6529–6549. 10.3390/nu7085297 [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Feng Y, Xing Y, Liu Z, Yang G, Niu X, Gao D. Integrated analysis of microRNA and mRNA expression profiles in rats with selenium deficiency and identification of associated miRNA-mRNA network. Sci Rep. 2018; 8:6601–24826. 10.1038/s41598-018-24826-w [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Brennan KM, Crowdus CA, Cantor AH, Pescatore AJ, Barger JL, Horgan K et al. Effects of organic and inorganic dietary selenium supplementation on gene expression profiles in oviduct tissue from broiler-breeder hens. Anim Reprod Sci. 2011; 125:180–188. 10.1016/j.anireprosci.2011.02.027 [DOI] [PubMed] [Google Scholar]
50.Zeng R, Liang Y, Farooq MU, Zhang Y, Ei HH, Tang Z et al. Alterations in transcriptome and antioxidant activity of naturally aged mice exposed to selenium-rich rice. Environ Sci Pollut Res Int. 2019; 26:17834–17844. 10.1007/s11356-019-05226-7 [DOI] [PubMed] [Google Scholar]
51.Ort JF, Latshaw JD. The toxic level of sodium selenite in the diet of laying chickens. J Nutr. 1978; 108:1114–1120. 10.1093/jn/108.7.1114 [DOI] [PubMed] [Google Scholar]
52.Pacitti D, Lawan MM, Feldmann J, Sweetman J, Wang T, Martin SA et al. Impact of selenium supplementation on fish antiviral responses: a whole transcriptomic analysis in rainbow trout (Oncorhynchus mykiss) fed supranutritional levels of Sel-Plex(R). BMC Genomics. 2016; 17:116 10.1186/s12864-016-2418-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Barnett MP, Bermingham EN, Young W, Bassett SA, Hesketh JE, Maciel-Dominguez A et al. Low folate and selenium in the mouse maternal diet alters liver gene expression patterns in the offspring after weaning. Nutrients. 2015; 7:3370–3386. 10.3390/nu7053370 [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Knight R, Marlatt VL, Baker JA, Lo BP, deBruyn AMH, Elphick JR et al. Dietary selenium disrupts hepatic triglyceride stores and transcriptional networks associated with growth and Notch signaling in juvenile rainbow trout. Aquat Toxicol. 2016; 180:103–114. 10.1016/j.aquatox.2016.09.014 [DOI] [PubMed] [Google Scholar]
55.Elgendy R, Giantin M, Castellani F, Grotta L, Palazzo F, Dacasto M et al. Transcriptomic signature of high dietary organic selenium supplementation in sheep: A nutrigenomic insight using a custom microarray platform and gene set enrichment analysis. J Anim Sci. 2016; 94:3169–3184. 10.2527/jas.2016-0363 [DOI] [PubMed] [Google Scholar]
56.Matthews JC, Zhang Z, Patterson JD, Bridges PJ, Stromberg AJ, Boling JA. Hepatic transcriptome profiles differ among maturing beef heifers supplemented with inorganic, organic, or mixed (50% inorganic:50% organic) forms of dietary selenium. Biol Trace Elem Res. 2014; 160:321–339. 10.1007/s12011-014-0050-4 [DOI] [PubMed] [Google Scholar]
57.Scott ML, Olson G, Krook L, Brown WR. Selenium-responsive myopathies of myocardium and of smooth muscle in the young poult. J Nutr. 1967; 91:573–583. 10.1093/jn/91.4.573 [DOI] [PubMed] [Google Scholar]
58.Liu CP, Fu J, Lin SL, Wang XS, Li S. Effects of dietary selenium deficiency on mRNA levels of twenty-one selenoprotein genes in the liver of layer chicken. Biol Trace Elem Res. 2014; 159:192–198. 10.1007/s12011-014-0005-9 [DOI] [PubMed] [Google Scholar]
59.Zhao X, Yao H, Fan R, Zhang Z, Xu S. Selenium deficiency influences nitric oxide and selenoproteins in pancreas of chickens. Biol Trace Elem Res. 2014; 161:341–349. 10.1007/s12011-014-0139-9 [DOI] [PubMed] [Google Scholar]
60.Huang JQ, Li DL, Zhao H, Sun LH, Xia XJ, Wang KN et al. The selenium deficiency disease exudative diathesis in chicks is associated with downregulation of seven common selenoprotein genes in liver and muscle. J Nutr. 2011; 141:1605–1610. 10.3945/jn.111.145722 [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Xu J, Wang L, Tang J, Jia G, Liu G, Chen X et al. Pancreatic atrophy caused by dietary selenium deficiency induces hypoinsulinemic hyperglycemia via global down-regulation of selenoprotein encoding genes in broilers. PLoS One. 2017; 12:e0182079 10.1371/journal.pone.0182079 [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Huang X, Tang J, Xu J, Jia G, Liu G, Chen X et al. Supranutritional dietary selenium induced hyperinsulinemia and dyslipidemia via affected expression of selenoprotein genes and insulin signal-related genes in broiler. RSC Adv. 2016; 6:84990–84998 [Google Scholar]
63.Xu JX, Zhang C, Cao CY, Zhu SY, Li H, Sun YC et al. Dietary Selenium Status Regulates the Transcriptions of Selenoproteome and Activities of Selenoenzymes in Chicken Kidney at Low or Super-nutritional Levels. Biol Trace Elem Res. 2016; 170:438–448. 10.1007/s12011-015-0470-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Sunde RA. Selenium In: Coates PM, editors. Dietary Supplements. New York: Marcel Dekkar; 2010. pp. 715–722. [Google Scholar]
65.White PJ. Selenium metabolism in plants. Biochim Biophys Acta Gen Subj. 2018; 1892:2333–2343. 10. 10.1016/j.bbagen.2018.05.006 [DOI] [PubMed] [Google Scholar]
66.Kobayashi Y, Ogra Y, Ishiwata K, Takayama H, Aimi N, Suzuki KT. Selenosugars are key and urinary metabolites for selenium excretion within the required to low-toxic range. Proc Natl Acad Sci U S A. 2002; 99:15932–15936. 10.1073/pnas.252610699 [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Bierla K, Szpunar J, Yiannikouris A, Lobinski R. Compehensive speciation of selenium in selenium-rich yeast. Trends Anal Chem. 2012; 41:122–132 [Google Scholar]
68.Takahashi K, Suzuki N, Ogra Y. Effect of administration route and dose on metabolism of nine bioselenocompounds. J Trace Elem Med Biol. 2018; 49:113–118. 10.1016/j.jtemb.2018.05.007 [DOI] [PubMed] [Google Scholar]
69.Takahashi K, Ogra Y. Identification of the biliary selenium metabolite and the biological significance of selenium enterohepatic circulation. Metallomics. 2019; 10 10.1039/c9mt00274j [DOI] [PubMed] [Google Scholar]
70.Sunde RA. Selenium In: Shils ME, Shike M, Ross CA, Caballero B, Cousins RJ, editors. Modern Nutrition in Health and Disease. Philadelphia: Lippincott Williams and Wilkins; 2012. pp. 225–237. [Google Scholar]
71.Bierla K, Taylor RM, Szpunar J, Lobinski R, Sunde RA. Identification and determination of selenocysteine, selenosugar, and other selenometabolites in turkey liver. Metallomics. 2020; 12: 10.1039/d0mt00040j [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Fig. GSEA of liver transcripts in turkeys fed Se-deficient vs Se-adequate diet.

(PDF)

Click here for additional data file.^{(1,011.1KB, pdf)}

S2 Fig. GSEA of liver transcripts in turkeys fed 2 or 1 μg Se/g vs. 0.4 μg Se/g.

(PDF)

Click here for additional data file.^{(988.6KB, pdf)}

S3 Fig. GSEA of liver transcripts in high-Se turkeys fed 5 μg Se/g vs. 0.4 μg Se/g.

Shown are resulting enrichment plots (left) and fold-change heat maps (right) for resulting leading-edge genes. (A) Hallmark cholesterol homeostasis gene set; (B) Hallmark pancreatic β-cell gene set; (C) Hallmark fatty acid metabolism gene set. Enrichment score is plotted vs. the rank in the ordered dataset of 11,612 humanized turkey transcripts (trimmed to include >14 counts, and normalized). Bar marks the leading-edge genes shown in the corresponding heat map. FDR q value shown for each plot.

(PDF)

Click here for additional data file.^{(988.6KB, pdf)}

S1 Table. Gene IDs and functions of significant DE transcripts by 0, 0.025, 2, and 5 μg Se/g.

(PDF)

Click here for additional data file.^{(28.9KB, pdf)}

S2 Table. Transcripts differentially expressed by 0.75 and 1.0 μg Se/g vs. Se-adequate (Exp 1).

(PDF)

Click here for additional data file.^{(76.5KB, pdf)}

S3 Table. Enriched gene ontology pathways.

(PDF)

Click here for additional data file.^{(61.6KB, pdf)}

S4 Table. Enriched gene ontology biological processes.

(PDF)

Click here for additional data file.^{(55.4KB, pdf)}

S5 Table. GSEA Se-deficient and low-Se most significant KEGG gene sets.

(PDF)

Click here for additional data file.^{(80.7KB, pdf)}

S6 Table. GSEA high-Se most significant KEGG gene gets.

(PDF)

Click here for additional data file.^{(81.6KB, pdf)}

S7 Table. Effect of dietary Se vs. Se-adequate on GO biological processes, GO cellular components, and GO molecular functions involving sulfur or thiols, from GSEA for 9996 gene sets.

(PDF)

Click here for additional data file.^{(33.9KB, pdf)}

S8 Table. Effect of dietary Se vs. Se-adequate on GO biological processes, GO cellular components, and GO molecular functions involving oxidative stress, from GSEA for 9996 gene sets.

(PDF)

Click here for additional data file.^{(33.1KB, pdf)}

S9 Table. Human symbols raw counts.

(XLSX)

Click here for additional data file.^{(2.2MB, xlsx)}

Data Availability Statement

[pone.0232160.ref001] 1.Rosenfeld I, Beath OA. Selenium poisoning in animals In: Selenium: Geobotany, Biochemistry, Toxicity and Nutrition. New York: Academic Press; 1964. pp. 142–213. [Google Scholar]

[pone.0232160.ref002] 2.Nelson AA, Fitzhugh OG, Calvery HO. Liver tumors following cirrhosis caused by selenium in rat. Cancer Res. 1943; 3:230–236 [Google Scholar]

[pone.0232160.ref003] 3.Schwarz K, Foltz CM. Selenium as an integral part of factor 3 against dietary necrotic liver degeneration. J Am Chem Soc. 1957; 79:3292–3293 [PubMed] [Google Scholar]

[pone.0232160.ref004] 4.Scott ML, Bieri JG, Briggs GM, Schwarz K. Prevention of exudative diathesis by Factor 3 in chicks on vitamin E-deficient Torula yeast diets. Poultry Sci. 1957; 36:1155. [Google Scholar]

[pone.0232160.ref005] 5.Chen X, Yang G, Chen J, Wen Z, Ge K. Studies on the relations of selenium and Keshan disease. Biol Tr Elem Res. 1980; 2:91–107 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref006] 6.Rotruck JT, Pope AL, Ganther HE, Swanson AB, Hafeman DG, Hoekstra WG. Selenium: biochemical role as a component of glutathione peroxidase. Science. 1973; 179:588–590. 10.1126/science.179.4073.588 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref007] 7.Sunde RA, Li JL, Taylor RM. Insights for setting of nutrient requirements, gleaned by comparison of selenium status biomarkers in turkeys and chickens versus rats, mice, and lambs. Adv Nutr. 2016; 7:1129–1138. 10.3945/an.116.012872 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref008] 8.Taylor RM, Sunde RA. Selenoprotein transcript level and enzyme activity as biomarkers for selenium status and selenium requirements of turkeys (Meleagris gallopavo). PLoS ONE. 2016; 11:e0151665 10.1371/journal.pone.0151665 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref009] 9.Taylor RM, Sunde RA. Selenium requirements based on muscle and kidney selenoprotein enzyme activity and transcript level in the turkey poult (Meleagris gallopavo). PLoS ONE. 2017; 12:e0189001 10.1371/journal.pone.0189001 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref010] 10.Taylor RM, Bourget VG, Sunde RA. High dietary inorganic selenium has minimal effects on turkeys and selenium status biomarkers. Poult Sci. 2019; 98:855–865. 10.3382/ps/pey413 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref011] 11.Barnes KM, Evenson JK, Raines AM, Sunde RA. Transcript analysis of the selenoproteome indicates that dietary selenium requirements in rats based on selenium-regulated selenoprotein mRNA levels are uniformly less than those based on glutathione peroxidase activity. J Nutr. 2009; 139:199–206. 10.3945/jn.108.098624 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref012] 12.Labunskyy VM, Hatfield DL, Gladyshev VN. Selenoproteins: molecular pathways and physiological roles. Physiol Rev. 2014; 94:739–777. 10.1152/physrev.00039.2013 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref013] 13.Kryukov GV, Castellano S, Novoselov SV, Lobanov AV, Zehtab O, Guigo R et al. Characterization of mammalian selenoproteomes. Science (Washington, DC). 2003; 300:1439–1443. 10.1126/science.1083516 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref014] 14.Sunde RA, Sunde GR, Sunde CM, Sunde ML, Evenson JK. Cloning, sequencing, and expression of selenoprotein transcripts in the turkey (Meleagris gallopavo). PLoS ONE. 2015; 10:e0129801 10.1371/journal.pone.0129801 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref015] 15.Gladyshev VN, Arner ES, Berry MJ, Brigelius-Flohe R, Bruford EA, Burk RF et al. Selenoprotein gene nomenclature. J Biol Chem. 2016; 291:24036–24040. 10.1074/jbc.M116.756155 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref016] 16.Raines AM, Sunde RA. Selenium toxicity but not deficient or super-nutritional selenium status vastly alters the transcriptome in rodents. BMC Genomics. 2011; 12:26 10.1186/1471-2164-12-26 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref017] 17.Nesheim MC, Scott ML. Studies of the nutritive effects of selenium for chicks. J Nutr. 1958; 65:601–618. 10.1093/jn/65.4.601 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref018] 18.Li JL, Sunde RA. Selenoprotein transcript level and enzyme activity as biomarkers for selenium status and selenium requirements of chickens (Gallus gallus). PLoS ONE. 2016; 11:e0152392 10.1371/journal.pone.0152392 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref019] 19.Walter ED, Jensen LS. Effectiveness of selenium and non-effectiveness of sulfur amino acids in preventing muscular dystrophy in the turkey poult. J Nutr. 1963; 80:327–331. 10.1093/jn/80.3.327 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref020] 20.Tinsley IJ, Harr JR, Bone JF, Weswig PH, Yamamoto RS. Selenium toxicity in rats. I. Growth and longevity In: Selenium in Biomedicine. Westport CN: AVI Publishing Co; 1967. pp. 141–152. [Google Scholar]

[pone.0232160.ref021] 21.Harr JR, Bone JF, Tinsley IJ, Weswig PH, Yamamoto RS. Selenium toxicity in rats. II. Histopathology In: Muth OH, Oldfield JE, Weswig PH, editors. Selenium in Biomedicine. Westport, CT: AVI Publishing Co; 1967. pp. 153–178. [Google Scholar]

[pone.0232160.ref022] 22.El-Begearmi MM, Combs GF Jr. Dietary effects on selenite toxicity in the chick. Poult Sci. 1982; 61:770–776. 10.3382/ps.0610770 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref023] 23.McAdam PA, Levander OA. Chronic toxicity and retention of dietary selenium fed to rats as D- or L-selenomethionine, selenite, or selenate. Nutr Res. 1987; 7:601–610. [Google Scholar]

[pone.0232160.ref024] 24.Whanger PD, Butler JA. Effects of various dietary levels of selenium as selenite or selenomethionine on tissue selenium levels and glutathione peroxidase activity in rats. J Nutr. 1988; 118:846–852 10.1093/jn/118.7.846 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref025] 25.Salbe AD, Levander OA. Effect of various dietary factors on the deposition of selenium in the hair and nails of rats. J Nutr. 1990; 120:200–206 10.1093/jn/120.2.200 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref026] 26.Sunde RA, Raines AM, Barnes KM, Evenson JK. Selenium status highly-regulates selenoprotein mRNA levels for only a subset of the selenoproteins in the selenoproteome. Biosci Rep. 2009; 29:329–338. 10.1042/BSR20080146 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref027] 27.Maret W. Zinc in Cellular Regulation: The Nature and Significance of "Zinc Signals". Int J Mol Sci. 2017; 18:ijms18112285 10.3390/ijms18112285 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref028] 28.Muckenthaler MU, Rivella S, Hentze MW, Galy B. A Red Carpet for Iron Metabolism. Cell. 2017; 168:344–361. 10.1016/j.cell.2016.12.034 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref029] 29.Kim BE, Nevitt T, Thiele DJ. Mechanisms for copper acquisition, distribution and regulation. Nat Chem Biol. 2008; 4:176–185. 10.1038/nchembio.72 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref030] 30.Boehler CJ, Raines AM, Sunde RA. Toxic-selenium and low-selenium transcriptomes in Caenorhabditis elegans: toxic selenium up-regulates oxidoreductase and down-regulates cuticle-associated genes. PLoS ONE. 2014; 9:e101408 10.1371/journal.pone.0101408 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref031] 31.Hu X, Chandler JD, Orr ML, Hao L, Liu K, Uppal K et al. Selenium supplementation alters hepatic energy and fatty acid metabolism in mice. J Nutr. 2018; 148:675–684. 10.1093/jn/nxy036 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref032] 32.National Research Council. Nutrient Requirements of Poultry. Washington D.C.: National Academy Press; 1994. 155 p. [Google Scholar]

[pone.0232160.ref033] 33.Brown MB, Forsyth AB. Robust tests for equality of variances. J Amer Stat Assn. 1974; 69:364–367. 10.1080/01621459.1974.10482955 [DOI] [Google Scholar]

[pone.0232160.ref034] 34.Steel R. G. D. and Torrie J. H. Principles and Procedures of Statistics. New York, NY: McGraw-Hill Book; 1960. [Google Scholar]

[pone.0232160.ref035] 35.Reed KM, Mendoza KM, Abrahante JE, Barnes NE, Velleman SG, Strasburg GM. Response of turkey muscle satellite cells to thermal challenge. I. transcriptome effects in proliferating cells. BMC Genomics. 2017; 18:352–3740. 10.1186/s12864-017-3740-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref036] 36.Lê S, Josse J, Husson F. An R package for multivariate analysis. J Stat Softw. 2008; 25:1–18 [Google Scholar]

[pone.0232160.ref037] 37.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15:550–0550. 10.1186/s13059-014-0550-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref038] 38.Saldanha AJ. Java Treeview—extensible visualization of microarray data. Bioinformatics. 2004; 20:3246–3248. 10.1093/bioinformatics/bth349 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref039] 39.Kuleshov MV, Jones MR, Rouillard AD, Fernandez NF, Duan Q, Wang Z et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acid Res. 2016; W44:W90–97. https://amp.pharm.mssm.edu/Enrichr/ [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref040] 40.Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005; 102:15545–15550. 10.1073/pnas.0506580102 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref041] 41.Mootha VK, Lindgren CM, Eriksson KF, Subramanian A, Sihag S, Lehar J et al. PGC-1-alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet. 2003; 34:267–273. 10.1038/ng1180 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref042] 42.Monson MS, Settlage RE, McMahon KW, Mendoza KM, Rawal S, El-Nezami HS et al. Response of the hepatic transcriptome to aflatoxin B1 in domestic turkey (Meleagris gallopavo). PLoS ONE. 2014; 9:e100930 10.1371/journal.pone.0100930 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref043] 43.Yin Y, Liu W, Dai Y. SOCS3 and its role in associated diseases. Hum Immunol. 2015; 76:775–780. 10.1016/j.humimm.2015.09.037 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref044] 44.Okamura SM, Oki-Idouchi CE, Lorenzo PS. The exchange factor and diacylglycerol receptor RasGRP3 interacts with dynein light chain 1 through its C-terminal domain. J Biol Chem. 2006; 281:36132–36139. 10.1074/jbc.M605093200 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref045] 45.Bosse AC, Pallauf J, Hommel B, Sturm M, Fischer S, Wolf NM et al. Impact of selenite and selenate on differentially expressed genes in rat liver examined by microarray analysis. Biosci Rep. 2010; 30:293–306. 10.1042/BSR20090089 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref046] 46.Mallonee DH, Crowdus CA, Barger JL, Dawson KA, Power RF. Use of stringent selection parameters for the identification of possible selenium-responsive marker genes in mouse liver and gastrocnemius. Biol Trace Elem Res. 2011; 143:992–1006. 10.1007/s12011-010-8894-8 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref047] 47.Tsuji PA, Carlson BA, Anderson CB, Seifried HE, Hatfield DL, Howard MT. Dietary selenium levels affect selenoprotein expression and support the interferon-gamma and IL-6 immune response pathways in mice. Nutrients. 2015; 7:6529–6549. 10.3390/nu7085297 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref048] 48.Feng Y, Xing Y, Liu Z, Yang G, Niu X, Gao D. Integrated analysis of microRNA and mRNA expression profiles in rats with selenium deficiency and identification of associated miRNA-mRNA network. Sci Rep. 2018; 8:6601–24826. 10.1038/s41598-018-24826-w [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref049] 49.Brennan KM, Crowdus CA, Cantor AH, Pescatore AJ, Barger JL, Horgan K et al. Effects of organic and inorganic dietary selenium supplementation on gene expression profiles in oviduct tissue from broiler-breeder hens. Anim Reprod Sci. 2011; 125:180–188. 10.1016/j.anireprosci.2011.02.027 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref050] 50.Zeng R, Liang Y, Farooq MU, Zhang Y, Ei HH, Tang Z et al. Alterations in transcriptome and antioxidant activity of naturally aged mice exposed to selenium-rich rice. Environ Sci Pollut Res Int. 2019; 26:17834–17844. 10.1007/s11356-019-05226-7 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref051] 51.Ort JF, Latshaw JD. The toxic level of sodium selenite in the diet of laying chickens. J Nutr. 1978; 108:1114–1120. 10.1093/jn/108.7.1114 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref052] 52.Pacitti D, Lawan MM, Feldmann J, Sweetman J, Wang T, Martin SA et al. Impact of selenium supplementation on fish antiviral responses: a whole transcriptomic analysis in rainbow trout (Oncorhynchus mykiss) fed supranutritional levels of Sel-Plex(R). BMC Genomics. 2016; 17:116 10.1186/s12864-016-2418-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref053] 53.Barnett MP, Bermingham EN, Young W, Bassett SA, Hesketh JE, Maciel-Dominguez A et al. Low folate and selenium in the mouse maternal diet alters liver gene expression patterns in the offspring after weaning. Nutrients. 2015; 7:3370–3386. 10.3390/nu7053370 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref054] 54.Knight R, Marlatt VL, Baker JA, Lo BP, deBruyn AMH, Elphick JR et al. Dietary selenium disrupts hepatic triglyceride stores and transcriptional networks associated with growth and Notch signaling in juvenile rainbow trout. Aquat Toxicol. 2016; 180:103–114. 10.1016/j.aquatox.2016.09.014 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref055] 55.Elgendy R, Giantin M, Castellani F, Grotta L, Palazzo F, Dacasto M et al. Transcriptomic signature of high dietary organic selenium supplementation in sheep: A nutrigenomic insight using a custom microarray platform and gene set enrichment analysis. J Anim Sci. 2016; 94:3169–3184. 10.2527/jas.2016-0363 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref056] 56.Matthews JC, Zhang Z, Patterson JD, Bridges PJ, Stromberg AJ, Boling JA. Hepatic transcriptome profiles differ among maturing beef heifers supplemented with inorganic, organic, or mixed (50% inorganic:50% organic) forms of dietary selenium. Biol Trace Elem Res. 2014; 160:321–339. 10.1007/s12011-014-0050-4 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref057] 57.Scott ML, Olson G, Krook L, Brown WR. Selenium-responsive myopathies of myocardium and of smooth muscle in the young poult. J Nutr. 1967; 91:573–583. 10.1093/jn/91.4.573 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref058] 58.Liu CP, Fu J, Lin SL, Wang XS, Li S. Effects of dietary selenium deficiency on mRNA levels of twenty-one selenoprotein genes in the liver of layer chicken. Biol Trace Elem Res. 2014; 159:192–198. 10.1007/s12011-014-0005-9 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref059] 59.Zhao X, Yao H, Fan R, Zhang Z, Xu S. Selenium deficiency influences nitric oxide and selenoproteins in pancreas of chickens. Biol Trace Elem Res. 2014; 161:341–349. 10.1007/s12011-014-0139-9 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref060] 60.Huang JQ, Li DL, Zhao H, Sun LH, Xia XJ, Wang KN et al. The selenium deficiency disease exudative diathesis in chicks is associated with downregulation of seven common selenoprotein genes in liver and muscle. J Nutr. 2011; 141:1605–1610. 10.3945/jn.111.145722 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref061] 61.Xu J, Wang L, Tang J, Jia G, Liu G, Chen X et al. Pancreatic atrophy caused by dietary selenium deficiency induces hypoinsulinemic hyperglycemia via global down-regulation of selenoprotein encoding genes in broilers. PLoS One. 2017; 12:e0182079 10.1371/journal.pone.0182079 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref062] 62.Huang X, Tang J, Xu J, Jia G, Liu G, Chen X et al. Supranutritional dietary selenium induced hyperinsulinemia and dyslipidemia via affected expression of selenoprotein genes and insulin signal-related genes in broiler. RSC Adv. 2016; 6:84990–84998 [Google Scholar]

[pone.0232160.ref063] 63.Xu JX, Zhang C, Cao CY, Zhu SY, Li H, Sun YC et al. Dietary Selenium Status Regulates the Transcriptions of Selenoproteome and Activities of Selenoenzymes in Chicken Kidney at Low or Super-nutritional Levels. Biol Trace Elem Res. 2016; 170:438–448. 10.1007/s12011-015-0470-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref064] 64.Sunde RA. Selenium In: Coates PM, editors. Dietary Supplements. New York: Marcel Dekkar; 2010. pp. 715–722. [Google Scholar]

[pone.0232160.ref065] 65.White PJ. Selenium metabolism in plants. Biochim Biophys Acta Gen Subj. 2018; 1892:2333–2343. 10. 10.1016/j.bbagen.2018.05.006 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref066] 66.Kobayashi Y, Ogra Y, Ishiwata K, Takayama H, Aimi N, Suzuki KT. Selenosugars are key and urinary metabolites for selenium excretion within the required to low-toxic range. Proc Natl Acad Sci U S A. 2002; 99:15932–15936. 10.1073/pnas.252610699 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0232160.ref067] 67.Bierla K, Szpunar J, Yiannikouris A, Lobinski R. Compehensive speciation of selenium in selenium-rich yeast. Trends Anal Chem. 2012; 41:122–132 [Google Scholar]

[pone.0232160.ref068] 68.Takahashi K, Suzuki N, Ogra Y. Effect of administration route and dose on metabolism of nine bioselenocompounds. J Trace Elem Med Biol. 2018; 49:113–118. 10.1016/j.jtemb.2018.05.007 [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref069] 69.Takahashi K, Ogra Y. Identification of the biliary selenium metabolite and the biological significance of selenium enterohepatic circulation. Metallomics. 2019; 10 10.1039/c9mt00274j [DOI] [PubMed] [Google Scholar]

[pone.0232160.ref070] 70.Sunde RA. Selenium In: Shils ME, Shike M, Ross CA, Caballero B, Cousins RJ, editors. Modern Nutrition in Health and Disease. Philadelphia: Lippincott Williams and Wilkins; 2012. pp. 225–237. [Google Scholar]

[pone.0232160.ref071] 71.Bierla K, Taylor RM, Szpunar J, Lobinski R, Sunde RA. Identification and determination of selenocysteine, selenosugar, and other selenometabolites in turkey liver. Metallomics. 2020; 12: 10.1039/d0mt00040j [DOI] [PubMed] [Google Scholar]

PERMALINK

The hepatic transcriptome of the turkey poult (Meleagris gallopavo) is minimally altered by high inorganic dietary selenium

Rachel M Taylor

Kristelle M Mendoza

Juan E Abrahante

Kent M Reed

Roger A Sunde

Roles

Abstract

Introduction

Methods

Reagents

Animals and diets

Tissue Se and Selenoenzymes

RNA isolation and high-throughput sequencing

Functional analysis

Pathway analysis and gene ontology analysis

Results

Se status

Table 1. Liver selenoenzyme activities and Se concentration in turkey poults fed graded levels of dietary Se.

PCA plots

Fig 1. Principal component analysis (PCA) of liver transcripts (n = 17,100) from poults supplemented with graded levels of Se for 28 d.

Volcano plots

Fig 2. Volcano plots depicting the effect of dietary Se on differential expression of liver transcripts (n = 17,100) in poults supplemented with graded levels of Se for 28 d.

Differential expression

Se-deficient and low Se

Table 2. Transcripts DE by 0 and 0.025 μg Se/g vs. Se-adequate (Exp 1)*.

High Se

Table 3. Transcripts DE by 2.0 and 5.0 μg Se/g vs. Se-adequate (Exp 2)*.

Supernutritional Se

Hierarchical cluster analysis

Fig 3. Hierarchical cluster analysis of significant DE transcripts (q < 0.1 in one or more treatments).

Pathway and gene ontology analysis

GSEA

Fig 4. GSEA of liver transcripts in turkeys fed Se-deficient vs Se-adequate diet.

Table 4. GSEA Se-deficient and low-Se most significant Hallmark gene setsa.

Table 5. GSEA high-Se most significant Hallmark gene setsa.

Fig 5. GSEA of liver transcripts in supernutritional-Se turkeys fed 1 μg Se/g vs.

Discussion

Overall gene expression effects

Se-deficient and low-Se transcript expression

High Se transcript expression

Supernutritional Se expression

Turkeys vs. rodents

Additional transcriptomics studies

Vitamin E

Other possibilities

Conclusions

Supporting information

Data Availability

Funding Statement

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Table 2. Transcripts DE by 0 and 0.025 μg Se/g vs. Se-adequate (Exp 1)^*.

Table 3. Transcripts DE by 2.0 and 5.0 μg Se/g vs. Se-adequate (Exp 2)^*.

Table 4. GSEA Se-deficient and low-Se most significant Hallmark gene sets^a.

Table 5. GSEA high-Se most significant Hallmark gene sets^a.