Identification of a novel antisense RNA that regulates growth hormone receptor expression in chickens

ABSTRACT

Natural antisense transcripts (NATs) are widely present in mammalian genomes and act as pivotal regulator molecules of gene expression. However, studies on NATs in the chicken are relatively rare. We identified a novel antisense transcript in the chicken, designated GHR-AS-EST, transcribed from the growth hormone receptor (GHR) locus, which encodes a well-known regulatory molecule of muscle development and fat deposition. GHR-AS-EST is predominantly expressed in the chicken liver and muscle tissues. GHR-AS-EST sequence conservation among vertebrates is weak. GHR-AS-EST forms an RNA–RNA duplex with GHBP to increase its stability, and regulates the expression of GHR sense transcripts at both the mRNA and protein levels. Further, GHR-AS-EST promotes cell proliferation by stimulating the expression of signaling factors in the JAK2/STAT pathway, and contributes to fat deposition via downregulating the expression of signaling factors in the JAK2/SOCS pathway in LMH hepatocellular carcinoma cells. We expect that the discovery of a NAT for a regulatory gene associated with cell proliferation and lipolysis will further our understanding of the molecular regulation of both muscle development and fat deposition.

Introduction

Antisense RNAs, also called natural antisense transcripts (NATs), are transcribed from the DNA strand opposite to the strand of their specific protein-coding or noncoding gene and overlap with the sense transcript [1,2]. Some NATs can form double-stranded RNAs with other transcripts according to the base complementary rule. Based on whether they work on cis or trans, NATs are grouped into cis-NATs and trans-NATs [3]. NATs belong to the family of long non-coding RNAs (lncRNAs) and have been divided into head-to-head, tail-to-tail, and fully overlapped categories in reference to their sense transcripts [4–8]. Various techniques and methods have been applied to identify NATs and to characterize their functions based on transcriptome data [3,9,10], chromatin features, genome distribution, expression pattern, and subcellular localization [6,11]. NATs regulate target genes expression at several levels, including transcription, mRNA processing, splicing, RNA stability, cellular transport and translational regulation [3,12,13]. Both sense and antisense RNAs can encode proteins or be non-protein-coding transcripts [2]. However, the functions of NATs are mostly unknown. Identifying functional coding or non-coding antisense RNAs and inferring biological pathways in which they act represent major challenges in understanding genome complexity and RNA-mediated gene regulation.

The actions of growth hormone (GH) are mediated via the growth hormone receptor (GHR) [14]. Both human Laron-type dwarfism and sex-linked dwarfism in the chicken are caused by GHR mutation [15–17]. Chicken GHR produces several sense transcripts through alternative usage of different 5′ untranslated regions (UTRs) and a functional polyadenylation signal [16,18–20] (Fig. S1). GHR mRNA encodes a full-length GHR protein that contains an extracellular domain, a transmembrane domain, and an intercellular domain [21], a truncated GH-binding protein (GHBP), that comprises only the extracellular hormone-binding domain of GHR [20], and a truncated 0.7-kb transcript that might encode a 27.5-kDa protein [18]. GHBP can recognize and bind to GH, which reduces the amount of GH bound to GHR [20]. In the liver, GH and GHR together with insulin-like growth factor (IGF) form the GH-GHR-IGF signal pathway that regulates various physiological processes, including cell proliferation and differentiation, as well as sugar, protein, fat metabolism, through promoting the transcription of related genes, such as insulin receptor substrate (IRS), cytokine signaling 3 (SOCS3), and leptin receptor (LEPR), which ultimately determine animal growth and development [15,16,22].

NATs have been reported for numerous genes [7,23]. In our previous study, we identified NATs originating from more than 7,200 genes that were extensively expressed in the chicken liver by using digital gene expression sequencing of RNA [24]. In addition to known expressed NATs, a 4,337 nucleotide (nt)-long NAT transcribed from the opposite strand of the GHR 3ʹUTR to intron 6 (Fig. S1B), which we termed GHR-AS [16]. In this study, we identified and characterized a novel intronic NAT of GHR transcribed from the opposite strand of GHR gene intron 5 to the 5ʹUTR, termed GHR-AS-EST. It highly expresses in normal chicken liver and leg muscle tissues. GHR-AS-EST suppresses GHR mRNA expression by repressing the transcriptional activity of the GHR V1 promoter. Surprisingly, the expression of GHBP and 0.7-kb GHR was stimulated when the transcriptional activity of the V1 promoter was reduced. We report that GHR-AS-EST does not stabilize GHR mRNA and the 0.7-kb transcript, but it does stabilize GHBP by forming an RNA–RNA duplex in vitro and in vivo, and that it decreases GHR expression, but increases the expression of GHBP and the 0.7-kb transcript at both the RNA and the protein level. GHR-AS-EST contributes to cell proliferation and fat deposition in LMH hepatocellular carcinoma cells by binding GHBP to influence GH-GHR signaling, promoting JAK2/STAT and suppressing JAK2/SOCS signaling. Moreover, the chicken GHR-AS-EST sequence shows weak similarity with those of other vertebrates, suggesting that it might be a species-specific NAT.

Results

GHR-AS-EST is strongly expressed in chicken liver and muscle tissues

From the expressed sequence tag (EST) sequences in the NCBI UniGene database (http://www.ncbi.nlm.nih.gov/UniGene), BX263735.2 was identified as a NAT that is transcribed from the negative strand of the Gallus gallus GHR locus and comprises more than three exons of the GHR gene (exons 1, 2, 4 and a portion of exon 5). To confirm that BX263735.2 originates from the reverse strand of the GHR locus, we conducted strand-specific reverse transcription (RT-)PCR, rapid amplification of cDNA ends (RACE) and northern blot analyses of White Recessive Rock (WRR) chicken liver tissues. As shown in Figs. S2 and S3, BX263735.2 was successfully amplified (Fig. S3A–C), and its entire sequence (688 bp with GC-rich  in 3'UTR) matched the sequence ‘CATGGATCCCAGTTTGACTAG’ sequence (Fig. S3D), which was previouly identified from the GHR negative strand at exon 6 [24]. These findings confirmed that BX263735.2 is a novel NAT of the GHR locus in the chicken genome. Moreover, this NAT was complementary to GHR pre-mRNA from the 5′UTR to exon 6. Therefore, we referred to BX263735.2 as the GHR antisense EST (GHR-AS-EST) transcript. GHR-AS-EST was classified as an antisense lncRNA based on its low coding potential as calculated using Coding Potential Calculator (CPC) software [25,26] (Fig. S3E). However, based on analysis of LMH cells transfected with pCMV-GHR-AS-EST-C-Flag vector, the transcript might encode a 17-kDa protein (Figure 1A). The transcript is expressed in both cytoplasmic and nuclear fractions of chicken primary myoblasts (Figure 1B). Next, we quantified GHR-AS-EST expression by RT-qPCR in 13 tissues from six 16-day-old WRR chicken embryos. GHR-AS-EST expression was the highest in the heart, followed by the liver, glandular stomach, and leg muscle, and it was the lowest in gizzard tissue (Figure 1C).

Figure 1.

Chicken GHR-AS-EST does encode a protein.

(A) Western blot analysis of the coding ability of GHR-AS-EST. The upper panel shows a schematic representation of pCMV-GHR-AS-EST-C-Flag vector. The whole sequence of GHR-AS-EST was cloned into the eukaryotic expression vector, pCMV-C-Flag. The lower panel showed the western blot results. LMH cells transfected with β-tubulin were used as a positive control (N = 3). (B) RT-qPCR detection of GHR-AS-EST in cytoplasmic and nuclear fractions of chicken primary myoblasts. GAPDH and U6 serve as cytoplasmic and nuclear localization controls, respectively. (C) GHR-AS-EST expression profile in WRR chicken tissues. Data are representative of three independent assays using tissues from six female 16-day-old WRR chickens embryos and shown as the mean ± SEM.

GHR-AS-EST suppresses GHR mRNA expression, but promotes the expression of both GHBP and 0.7-kb GHR

The V1 GHR sense promoter reportedly drives the transcription of GHR sense transcript [27]. In this study, we identified an antisense promoter (ASP) of GHR located in intron 5 and 99 bp away from the 5′ end of GHR-AS-EST, which drives the expression of GHR-AS-EST (Figures 2A and S4). Given the reports on NAT regulation of host gene expression [16,28–30], and the potential overlap region between the GHR sense transcripts and GHR-AS-EST (Figure 2A), we predicted that the overlap between GHR-AS-EST and GHR was indicative of an underlying regulatory effect. To confirm this, we constructed an interfering oligo (si-GHR-AS-EST) and a GHR-AS-EST overexpression vector (pc3.1-GHR-AS-EST), and transfected them into LMH cells. After 48 h, the relative mRNA levels of GHR and IGF1 were significantly enhanced after GHR-AS-EST knockdown (p < 0.05) and supressed by >55% following GHR-AS-EST overexpression (p < 0.01) when compared to levels in the respective control cells, as indicated by RT-qPCR (Figure 2B and E). In contrast, the relative RNA levels of GHBP and 0.7-kb GHR were significantly lower after GHR-AS-EST knockdown (p < 0.05) and significantly higher after GHR-AS-EST overexpression than in the respective control cells (Figure 2C and D)). Protein expression levels of GHR, GHBP, and 0.7-kb GHR were consistent with the RNA levels (Figure 2F–H). These results indicated that GHR-AS-EST suppresses GHR mRNA and protein expression, but promotes RNA and protein expression of GHBP and 0.7-kb GHR.

Figure 2.

GHR-AS-EST inhibits GHR expression, but promotes the expression of 0.7-kb GHR and GHBP.

(A) Schematic representation of the structure and transcriptional orientation of GHR sense transcript (GHR-S-RNA) or antisense RNA (GHR-AS-EST). Green and orange boxes: GHR sense promoter V1 and antisense promoter (ASP), respectively. E: exon. UTR: untranslated region. AUG in GHR-S-RNA represents the translation initiation codon. The 5ʹ end of GHR-AS-EST is 99 bp away from ASP. Relative RNA expression levels of GHR mRNA (B), 0.7-kb GHR (C), IGF1 (D) and GHBP (E) in treated LMH cells. Data are the mean ± SEM and are representative of triplicate experiments. Protein expression levels of GHR (F), 0.7-kb GHR (G), and GHBP (H) in treated LMH cells. *p < 0.05; **p < 0.01.

GHR-AS-EST inhibits GHR mRNA expression by suppressing the transcriptional activity of the V1 sense promoter

To evaluate the regulatory functions of GHR-AS-EST in the expression of the various GHR transcripts, we examined the transcriptional activity of the GHR V1 promoter in LMH cells. A pEGFP reporter gene assay revealed that the transcriptional activity of GHR V1 was significantly increased after treatment of LMH cells with si-GHR-AS-EST when compared with the control (si-NC) (p < 0.05) (Fig. S5A). Opposite findings were obtained in an overexpression assay (p< 0.05) (Fig. S5B). A pGL3-basic promoter reporter assay revealed similar results: GHR V1 transcriptional activity was enhanced by 39% in GHR-AS-EST knockdown LMH cells as compared to control cells (p < 0.05), and was suppressed by 38% in GHR-AS-EST-overexpressing LMH cells as compared to control cells (p < 0.01) (Figure 3). Together, these results demonstrated that GHR-AS-EST suppresses GHR V1 transcriptional activity, thus affecting GHR mRNA expression. Unexpectedly, GHBP and 0.7-kb GHR expressions were increased when GHR V1 transcriptional activity was decreased. Based on these results, we hypothesized that GHR-AS-EST downregulates the transcriptional activity of GHR V1 to inhibit GHR mRNA expression, whereas expression of GHBP and 0.7-kb GHR is upregulated through another mechanism.

Figure 3.

GHR-AS-EST decreases the transcriptional activity of the GHR V1 promoter.

Relative transcriptional activity changes of the GHR V1 promoter after knockdown and overexpression of GHR-AS-EST. Data are shown as fold induction compared with the activity of LMH cells transfected with pGL3-basic (promoter-less) luciferase reporter vector, and are the mean ± SEM of at least three experiments. *p < 0.05; ***p < 0.001.

GHR-AS-EST upregulates GHBP expression by forming an RNA duplex in vivo and in vitro

Antisense RNAs reportedly form duplexes with their sense mRNA partners [12,13,16,31,32], and these duplexes regulate the stability of the coding mRNA and thus, the protein level [2,33]. To determine whether GHR-AS-EST forms a duplex with GHR sense transcripts, including GHR mRNA, GHBP and 0.7-kb GHR, we carrried out an endogenous RNase protection assay using RNA extracted from liver and leg muscle tissues, LMH cells, and primary myoblasts. Only the overlapping region at exons 2-6 of GHR mRNA was retained (Figure 4A and B)), indicating that GHR-AS-EST and GHR sense transcript form a 537-bp RNA duplex in vivo and in vitro. Next, we conducted an mRNA stability assay to investigate whether GHR-AS-EST overexpression would enhance GHR sense transcripts stability. Compared to cells transfected with an empty vector (pcDNA 3.1), GHR-AS-EST-overexpressing (pc3.1-GHR-AS-EST) cells demonstrated increased stability of GHBP, but not GHR mRNA and 0.7-kb GHR (Figure 4C), which was consistent with the results on the protein levels of both GHBP and GHR (Figure 2F and H). These findings suggested that GHR-AS-EST enhances the stability of GHBP RNA by forming a duplex, and thus regulate its protein level.

Figure 4.

GHR-AS-EST upregulates GHBP expression by forming an RNA duplex in exons 2–6.

(A) Chicken GHR exons 2–6 RNA duplex detected in liver tissue and LMH cells. (B) Chicken GHR exons 2–6 RNA duplex detected in leg muscle tissue and primary myoblasts. (a) lane 1: DNA template as a positive control; lanes 2 and 3, the template is cDNA that is strand-specifically reverse-transcribed from 12-day-old WRR chicken embryo liver tissues (A) or leg muscle tissue (B); lanes 4 and 5, the template is cDNA that is strand-specifically reverse-transcribed from LMH cells (A) or primary myoblasts (B). (b–f) lane 1: cDNA is the template as a positive control; lanes 2–5, the templates are the same as those in panel (a). Marker (M): 2000, 1200, 750, 500, 250, and 100 bp. The black triangle indicated the RNA duplex formed by GHR-S and GHR-AS-EST with a total length of 537 bp. (C) mRNA stability assay results. Data are shown as the mean ± SEM of at least three experiments. *p < 0.05; **p < 0.01; ***p < 0.001.

Localization of GH and GH dimerization with different GHR proteins in living cells

To directly investigate the localization and dimerization of GH and/or different GHR proteins in living cells, we conducted fluorescence colocalization assay by transfecting LMH cells with GH_CFP, GHR_EGFP, GHBP_YFP, 0.7-kb GHR_BFP, or GHR-AS-EST_DsRed (Fig. S6A-B), or GH_YFP/GHBP_DsRed, 0.7-kb GHR_YFP/GHR-AS-EST_DsRed, GH_YFP/0.7-kb GHR_DsRed, or GH_YFP/GHR-AS-EST_DsRed (Fig. S6C). Strikingly, GH, GHR and GHBP were expressed only in the cytosol (Fig. S3B and D(a)), whereas 0.7-kb GHR and GHR-AS-EST were expressed in both the cytosol and the nucleus (Fig. S6D(c), (d)). Fluorescence resonance energy transfer (FRET) was used to visually monitor the dimerization of the six above pairs in living LMH cells. We first optimized the filter settings to make sure no crossover excitation of blue fluorescent protein (BFP) occurred when DsRed was excited. For this purpose, we individually transfected GH_BFP, GHBP_BFP, GHR-AS-EST_BFP, GHBP_DsRed, 0.7-kb GHR_DsRed, or GHR-AS-EST_DsRed into chicken primary myoblasts cells. Cells transfected with GH_BFP, GHBP_BFP or GHR_BFP alone exhibited blue fluorescence under excitation at 382 nm and emission at 435–465 nm, and no fluorescence signal was detected in the bandwidth of 570–600 nm (GHBP_DsRed, 0.7-kb GHR_DsRed, GHR-AS-EST_DsRed channel) (Figure 5A–F, upper panels). Similarly, cells transfected GHBP_DsRed, 0.7-kb GHR_DsRed, or GHR-AS-EST_DsRed alone exhibited red fluorescence under excitation at 558 nm and emission at 570–600 nm, without crossover to the BFP channel (emission at 435–465 nm) (Figure 5A–F, middle panels). At the optimal FRET setting, excitation of GH_BFP, GHBP_BFP, or GHR-AS-EST_BFP (donor) at 382 nm produced a diffuse image of GHBP_DsRed, 0.7-kb GHR_DsRed, or GHR-AS-EST_DsRed (acceptor) at 570–600 nm (Figure 5A–F), bottom panels), suggesting that FRET occurred from GH_BFP to GHBP_DsRed, 0.7-kb GHR_DsRed, or GHR-AS-EST_DsRed, GHBP_BFP to 0.7-kb GHR_DsRed or GHR-AS-EST_DsRed, and GHR-AS-EST_BFP to 0.7-kb GHR_DsRed.

Figure 5.

FRET analysis.

Chicken primary myoblasts were transfected with GH_BFP, GHBP_BFP, GHR-AS-EST_BFP, GHBP_DsRed, 0.7-kb GHR_DsRed, or GHR-AS-EST_DsRed, and 48 h later, the cells were observed under a confocal microscope with optimal filter sets. Cells containing GH_BFP, GHBP_BFP and GHR-AS-EST_BFP were observed and imaged with excitation of BFP (A–F, upper left panels), with no crossover with the DsRed channel (A–F, upper second panels). Conversely, in cells containing GHBP_DsRed, 0.7-kb GHR_DsRed and GHR-AS-EST_DsRed, DsRed was excited and imaged in the DsRed channel only (A–F, middle left panels) without crossover with the BFP channel (A–F, middle second panels). Cells containing both GH_BFP and GHBP_DsRed, GH_BFP and 0.7-kb GHR_DsRed, GH_BFP and GHR-AS-EST_DsRed, GHBP_BFP and 0.7-kb GHR_DsRed, GHBP_BFP and GHR-AS-EST_DsRed, or GHR-AS-EST_BFP and 0.7-kb GHR_DsRed were observed at the excitation of the donor (BFP) and collection of both donor (BFP) (A–F, bottom left panels) and acceptor (DsRed) emission (A–F, bottom second panels) using optimal FRET settings. Corresponding phase-contrast and overlay images of the cells are presented in (A–F, third panels) and (A–F, right panels), respectively.

GHR-AS-EST blocks signal transmission from GH to GHR and from GHR to STAT5 by promoting GHBP expression

For signaling, GH needs to bind to a GHR dimer, which modifies the position of the GHR extracellular domains and activates JAK2-STAT signaling [17,34]. GHBP inhibits the binding of GH to GHR by forming a GH–GHBP complex [20]. Though GHBP and 0.7-kb GHR expression was upregulated by GHR-AS-EST (Figure 2), whether 0.7-kb GHR regulated GHR mRNA and GHR-AS-EST expression remained unknown. In LMH cells transfected with pc3.1-0.7-kb GHR overexpression plasmid, unexpectedly, GHR mRNA expression was not affected, whereas GHR-AS-EST expression was significantly enhanced (p < 0.01) (Fig. S7A). However, in LMH cells cotransfected with pc3.1-0.7 kb GHR and pc3.1-GHR-AS-EST, GHR mRNA expression was suppressed (p < 0.05) (Fig. S7B). We hypothesized that GHR-AS-EST can be induced by 0.7-kb GHR and affect the signal transduction from GHR to STAT5 by influencing GHBP expression. To confirm this hypothesis, we explored the RNA level of different factors involved in JAK2/STAT and JAK2/MAPK signaling (Figure 6A). Upon GHR-AS-EST knockdown, GH mRNA expression was increased, whereas the relative expression of STAT1-6, RAS guanyl releasing protein (RASGRP), B-Raf proto-oncogene (BRAF), Raf-1 proto-oncogene (RAF1), and mitogen-activated protein kinase kinase 2 (MAP2K2) was decreased (Figure 6B and D). In contrast, JAK2, STAT1, STAT2, STAT6 and MAP2K2 expression was appreciably enhanced after GHR-AS-EST overexpression (Figure 6C and E). Cell proliferation was strongly stimulated in expression LMH cells overexpressing GHR-AS-EST, as indicated by cell-cycle analysis, CCK8 and EdU assays (Figure 6F–K).

Figure 6.

GHR-AS-EST positively regulates JAK2/STAT and JAK2/MAPK signaling.

(A) Schematic diagram of the JAK2/STAT and JAK2/MAPK signaling pathways. (B–E) Relative expression levels of GH, JAK2, STAT1-6, and MAP2K2 in treated LMH cells. (F) and (G) Cell cycle analysis of treated LMH cells. (H) and (I) CCK8 assay results. (J) and (K) Proliferating cell fractions (%) in treated LMH cells (magnification, 400×). Data are the mean ± SEM of at least three experiments. *p < 0.05; **p < 0.01.

GHR-AS-EST stimulates fat deposition via JAK2/SOCS signaling

SOCS3 induces fat deposition by suppressing IRS1, LEPR and JAK expression. Additionally, the GHR gene regulates SOCS3 through JAK/STAT signaling [15,22]. This suggests that GHR plays a role in fat deposition. Hence, we deduced that GHR-AS-ETS might regulate fat deposition, too. To verify this, we evaluated the RNA levels of different factors involved in JAK2/SOCS signaling (Figure 7A). SOCS3, IRS1, and LEPR mRNA levels were appreciably suppressed after GHR-AS-EST overexpression as indicated by RT-qPCR (p < 0.01; Figure 7B–D). In contrast, mRNA expression of SOCS3 and LEPR was significantly enhanced after GHR-AS-EST knockdown (p < 0.05 and p < 0.01, respectively, Figure 7B–D). Oil Red O staining assay indicated that fat deposition in LMH cells was lowered upon GHR-AS-EST knockdown, but enhanced by GHR-AS-EST overexpression (Figure 7E and F). These findings showed that chicken GHR-AS-EST can facilitate fat deposition, and we suggest via either inhibition of signal transduction from GH to GHR and from GHR to SOCS3 via a decrease in GHR mRNA expression and an increase in GHBP expression; or through direct inhibition of signal transduction in JAK2/SOCS pathway. Unexpectedly, although chicken GHR-AS-EST contains the sequence that is complementary to the exons 2-6-coding region of GHR, it exhibites <60% sequence identity with sequences of other species, as indicated by MEGA6.0 analysis (Table 1), BLAST2P with E-value >0.001, and other methods (data not shown), indicating low sequence conservation.

Figure 7.

GHR-AS-EST increase fat deposition in LMH cells.

(A) Schematic diagram of JAK2/STAT and signaling pathway. (B–D) Relative expression levels of SOCS3, IRS1, and LEPR in treated LMH cells. Data are the mean ± SEM of at least three experiments. (E) and (F) Oil Red O staining analysis of fat deposition in treated LMH cells (magnification, 40×). *p < 0.05; **p < 0.01.

Table 1.

Pair distances of GHR-AS-EST among different species.

Divergence
Sheep	Norway rat	Pig AJ94061	Pig AJ687535	BY775223.1	BU757735.1	Human	frog	medaka	Dog	Cattle	Chicken	Pencent identity
350.0	40.5	191.0	148.9	350.0	109.1	350.0	151.2	277.3	41.4	190.5		Chiken
350.0	191.7	12.8	221.3	231.4	178.5	221.1	98.6	186.5	193.2		2.9	Cattle
350.0	23.8	212.4	157.3	350.0	119.3	350.0	181.5	224.8		4.1	59.9	Dog
350.0	186.4	171.5	350.0	350.0	180.3	99.5	150.4		11.5	3.7	4.5	Japaness medaka
350.0	181.5	97.9	192.1	297.3	242.4	218.9		2.8	2.8	14.4	5.7	African clawed frog
221.3	350.0	224.7	350.0	170.1	230.8		4.8	6.0	2.8	5.3	14.3	Human
350.0	95.2	174.7	151.3	350.0		4.6	3.8	4.7	19.4	4.5	28.8	House mouse BU757735.1
44.2	350.0	350.0	167.9		4.8	6.4	5.6	4.4	6.8	8.8	16.0	House mouse BY775223.1
350.0	138.6	203.4		7.2	14.7	4.5	11.2	4.8	4.8	5.4	5.4	Pig AJ687535
350.0	209.4		5.1	5.2	4.3	4.4	9.7	3.6	5.5	84.9	1.7	Pig AJ94061
350.0		4.8	8.6	8.4	32.6	3.7	2.9	2.7	80.0	3.4	57.6	Norway rat
	3.4	5.0	4.8	52.0	3.2	9.0	4.1	3.2	2.5	4.3	12.8	Sheep

Discussion

NATs can originate from an intron of the host gene [5,35]. Here, we report GHR-AS-EST, a 688-bp and GC-rich RNA transcribed in antisense direction from intron 5 of the GHR gene. GHR-AS-EST represses GHR expression and promotes GHBP and 0.7-kb GHR expression. This finding indicated that the antisense RNA could divergently regulate host gene expression. NATs having an inverse relationship with host genes have been reported in numerous studies. For example, p15AS could silence p15 in cis and in trans through heterochromatin formation [29]; the NAT of Apolipoprotein A1 (APOA1-AS) was shown to repress APOA1 expression in humans and monkeys [36]; nucleolar protein 14-AS1 anti-correlates with its sense gene [37]; and a NAT of low-density lipoprotein receptor-related protein 1 downregulates the latter’s expression [38]. However, some sense/antisense transcript pairs are concordantly expressed in a cell type or a tissue. For example, Sirtuin 1 is concordantly expressed with its antisense partner during myogenic differentiation of mouse C2C12 cells [25]. Additionally, a NAT as well as a short splice variant of PTEN induced putative kinase 1 coexist in human neuroblastoma cells and human skeletal muscle tissue [39].

RNA duplex formation may mask pivotal regulatory features within transcripts to restrain the binding of other trans-acting factors. RNA masking intervenes with mRNA splicing, transport, polyadenylation, translation, and degradation [4,40]. Some NATs pair with their host genes to regulate epigenetic silencing, transcription and mRNA stability [12]. Increasing numbers of functional NATs are being identified in animals, such as mice, pigs, sheep, and chickens. For example, a NAT targeting the mouse gene encoding ubiquitin carboxy-terminal hydrolase L1 which is involved in brain function and neurodegenerative diseases, has been showed to regulate the synthesis of corresponding protein at the post-transcriptional level [12]. A phospholipase A2 group XVI locus (PLA2G16) NAT transcribed from porcine PLA2G16 promotes PLA2G16 transcription and might stimulate adipogenesis in pigs [31]. We previously identified a GHR-AS that enhanced GHR mRNA stability by forming an RNA–RNA duplex at the last exon of GHR mRNA in chicken LMH cells [16]. In this study, we uncovered a novel antisense RNA of chicken GHR, GHR-AS-EST, which decreases GHR mRNA and 0.7-kb GHR stability but increases GHBP stability by forming an RNA–RNA duplex at the exons 2-6 of the GHR sense transcripts, as shown by in-vivo and in-vitro experiments. Exon 2 of GHR mRNA is necessary for GHR translation because it possesses the translation initiation codon (AUG) [20,41]. Exons 2-6 are required for GH–GHR binding [42,43]. GHBP prolongs the half-life of GH by forming a GHBP–GH complex. However, this complex formation inhibits GH–GHR binding [20]. Hence, hybridization of GHR-AS-EST to exons 2-6 of GHBP might supresses GH–GHBP binding and thus increase the probability for GH to bind with the GHR.

GH has more than 400 actions that are mediated by the GHR, which is widely expressed in nearly all body tissues [44,45]. GHR is highly expressed in liver and muscle tissues [15,16], and 0.7-kb GHR is highly expressed in liver, skeletal muscle, and adipose tissues of dwarf chicken [18]. The current study indicated that GHR-AS-EST regulates GHR RNA and protein expression, as well as the signalling transduction from GH to GHR and GHR to downstream JAK/STAT, and JAK/SOCS signaling via GHBP in LMH cells. Considering the biological function of GHR in chicken muscle development and fat deposition, GHR-AS-EST can be considered as a candidate regulatory element for muscle development, fat deposition, and fat metabolism in chickens. Further exploration of the underlying mechanism of GHR-AS-EST in chicken muscle development and fat deposition will be of interest. It will be very important to assess whether GHR-AS-EST-encoded protein exists in vivo. Both GH and IGF1 are involved in somatotropic axis (GH-GHR-IGF1). IGF1 is a very potent growth factor that stimulates growth in all cell types playing a critical role in pre- and postnatal growth [34]. Therefore, further exploration of changes in GH and IGF1 proteins caused by GHR-AS-EST might provide insights into the biological significance of GHR-AS-EST and details on its regulatory mechanism.

NATs are capable of controlling the epigenetic status of the promoter of their host genes [7,35,46]. Given that CpG sites of the GHR V1 promoter are more highly methylated in normal than dwarf WRR chicken livers [16] and that GHR-AS-EST is a GC-rich NAT located only 29 bp from the GHR V1 promoter, GHR-AS-EST might have a differential regulatory function in GHR V1 methylation between normal and dwarf WRR chickens. The chicken GHR gene produces different sense transcripts as a result of alternative usage of different 5′UTRs and functional polyadenylation signals [16,18–20]. Interestingly, our study identified two new GHR sense transcripts that have different 3′UTRs, and contain different 5′UTRs of which differ from that of GHR mRNA. Hence, in future, we plan to investigate the potential function of GHR-AS-EST in regulating the GHR V1 promoter methylation level and in alternative splicing involving the 5′UTR and polyadenylation signal of GHR sense transcripts.

Although several NATs are highly conserved in mammals [39,47], many NATs lack strong sequence conservation because their evolution does not involve the modification of corresponding protein-coding regions [48]. Interestingly, such weak evolutionary constraints not only allow the rapid generation of lncRNAs, but also contribute to rapid sequence evolution [49]. Rapid evolution of antisense lncRNAs enables rapid responses to and recovery from stimuli in organisms [48]. Our data revealed that chicken GHR-AS-EST exhibits low sequence conservation, and could not be found in other vertebrates included in the NCBI UniGene database. This may indicate low abundance or even total absence of GHR-AS-EAT in other species. Therefore, we infer that GHR-AS-EST is a species-specific antisense RNA.

In summary, we identified a novel chicken NAT, GHR-AS-EST, which shows very low sequence similarity with sequences in different other species, but strongly compliments its host gene, GHR, a primary regulator of muscle development and fat deposition. GHR-AS-EST might encode a ~ 17-kDa protein and regulates GHR RNA and protein expression via regulating the transcriptional activity of the GHR V1 promoter and forming an RNA duplex with GHBP. Although we evaluated the expression and function of GHR-AS-EST only in liver and muscle tissues of normal chicken, not sex-linked dwarf chickens, we carefully suggest that the differential fat deposition in these chicken lines might be affected by GHR-AS-EST. We believe that the discovery of GHR-AS-EST will aid further exploration into the molecular mechanisms of muscle development and fat deposition in chickens.

Materials and methods

Ethics statement

All animal procedures were authorized by the Animal Care Committee of South China Agricultural University (Guangzhou, China). Animals involved in the present study were humanely sacrificed and every attempt was made to minimize their suffering.

Animals and tissue collection

Six 16-day-old female normal embryonic WRR chickens bought from the Chicken Breeding Farm of South China Agricultural University (Guangzhou, China) were used. The following tissues were collected: cerebrum, cerebellum, hypothalamus, heart, liver, spleen, lung, kidney, breast muscle, leg muscle, glandular stomach, and gizzard. Tissue samples were immediately homogenized in liquid nitrogen.

RNA extraction and cloning of chicken GHR-AS

Total RNA was extracted from the chicken liver tissues using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA). RNA integrity was assessed by agarose electrophoresis. RACE of chicken GHR-AS-EST was conducted using a RACE kit (Clontech, Japan) per the manufacturer’s instruction, followed by nested PCRs of the cDNA copies. The PCR products were cloned and sequenced in the forward and reverse directions. All RACE primers are listed in Table S1.

Northern blotting

Northern blotting to evaluate GHR-AS-EST RNA expression in the chicken liver was conducted as described previously [16], with some modifications. RNA probe was designed to minimize non-specific hybridization through BLAST searches (http://blast.ncbi.nlm.nih.gov/). The selected 444-nt probe sequence for the detection of GHR-AS-EST (Table 2) was synthesized and labelled with biotin by Sangon Biotech Company (Shanghai, China).

Table 2.

The single strand biotin-labelled probe, used for Northern blot to detect chicken GHR-AS-EST.

Application

Probe sequence (5ʹ to 3ʹ, 444 bp)

GHR-AS-EST detection

actccttccattgggtctcattaacttatttgtactgcaattcatactccagagtaatccatcctttctgaacatctgctgttggtggtggatcccatcgtacttgaatatccccatggatcccagtttgactagtatttagcagagtccagttaaggtgcacagggggatcaggtagtactatttcatcaacactgaaacacttttcgtcaaatacttcatctttattggcaagcttaacacaatatggtatccaaatcgaggtgtaggatgtgttgaagtaacagctattttctcctgcagtgatataatccggacattctttccagtcttcatcactccttttcatatacaacagttgtattgttcctgaagtagtgaggtttccatcagtccaataacacgaaaatgtctccagctcaggtgacctgcacttgctgattt

Strand-specific RT-PCR

Total RNA was extracted using TRIzol (TaKaRa, Otsu, Japan) according to the manufacturer’s instructions, and treated with gDNA Eraser (TaKaRa) at 42°C for 5 min. One microgram of RNA was reverse-transcribed into cDNA utilizing RT reagent Kit (TaKaRa) in 20-μL reactions at 50°C for 15 min, 85°C for 5 s. The reverse primer of GHR-AS-EST (5 pmol) was used as the RT primer. PCRs were carried out in 20-μL reactions including 5 μL strand-specific cDNA, 10 μL 2 × Taq PCR MasterMix (Tiangen, Beijing, China), and 0.5 μM forward and reverse primer. Thermal cycles were as follows: 94°C for 3 min, 35 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 1 min, and finally, 72°C for 5 min.

Construction of expression plasmids

The sequences of GHR-AS-EST and the 0.7-kb GHR transcript were amplified by PCR, cleaved, and ligated into corresponding sites of pcDNA 3.1 to construct pc3.1-GHR-AS-EST and pc3.1-0.7-kb GHR. GHR V1 and ASP luciferase reporters constructs were generated by cloning the corresponding regions into the pGL3-basic vector by PCR. For qualitative analyses of the GHR V1 and ASP promoters, GFP luciferase reporters pV1-EGFP and pASP-EGFP were constructed by cloning the corresponding V1 and ASP regions into the pEGFP-N1 vector to substitute its CMV region, by PCR. The negative control vector pLinker-EGFP (without CMV region of pEGFP-N1) was constructed using deletion Ω-PCR as reported elsewhere [27]. To construct the Flag-tag expression plasmids, GH-CDS, GHR-FL-CDS, GHBP, 0.7-kb GHR and GHR-AS-EST were amplified by PCR from chicken liver cDNA. The PCR products were cleaved and ligated into the corresponding sites of pCMV-C-Flag (Beyotime Biotechnology, Guangzhou, China). To construct other tagged expression plasmids, cDNAs encoding full-length chicken GH, GHR-FL-CDS, GHBP, 0.7-kb GHR, and GHR-AS-EST, with their signal peptides if present, were separately amplified by PCR. The PCR products were inserted into the vector of pCMV-C-CFP, pCMV-C-EGFP, pCMV-C-YFP, pCMV-C-BFP, and/or pCMV-C-DsRed (Beyotime), producing X_Y constructs (X represents GH, GHR, GHBP, 0.7-kb GHR, or GHR-AS-EST, Y represents CFP, EGFP, YFP, BFP, or DsRed). Primers used in these experiments are listed in Table S1.

Cell culture and transfection

The chicken LMH cell line was a gift from the Harbin Veterinary Research Institute, Chinese Academy of Agricultural Science (Heilongjiang, China). Cells were cultured in high-glucose Dulbecco’s modified Eagle’s medium (Gibco, Grand Island, NY, USA) containing 10% fetal bovine serum (Invitrogen), penicillin (100 IU/mL), and streptomycin (100 μg/mL) (Invitrogen). Chicken primary myoblasts were isolated from 11-day-old chicken embryos and were cultured as described elsewhere [50]. All cells were maintained at 37°C in 5% CO₂.

For GHR-AS-EST overexpression and knockdown assays, cells were seeded in 12- or 24-well plates. When the cells reached 60–70% confluence, they were transfected with pc3.1-GHR-AS-EST or empty pcDNA3.1, and siRNA oligonucleotides or negative control (si-NC: GenePharma, Suzhou, China) (Table 3), using Lipofectamine 3000 (Invitrogen), following the manufacturer’s instructions. Forty-eight hours post-transfection, the cells were harvested and stored at −80°C for further analysis.

Table 3.

siRNA targeting chicken GHR-AS-EST.

	Sequence (5ʹ to 3ʹ)
si-RNA Name	Sense	Antisense
si-GHR-AS-EST	CCUUCCAUUGGGUCUCAUUTT	AAUGAGACCCAAUGGAAGGTT
si-Negative control (si-NC)	UUCUCCGAACGUGUCACGUTT	ACGUGACACGUUCGGAGAATT

Rt-qPCR

Total RNA was reverse-transcribed to cDNA using Oligo-dT and random 6mers, and the PrimeScript RT Master Mix kit (TaKaRa, Otsu, Japan) per the manufacturer’s instructions. qPCRs were run in triplicate on a Bio-Rad CFX96 Real-Time Detection system (Bio-Rad, Hercules, California, USA) using the SYBR Premix ExTaq kit (TaKaRa, Otsu, Japan). Primers used for RT-qPCR were designed by Primer5 software and sequences are shown in Table S1. Relative target gene expression levels were calculated using the 2^–ΔΔCt method and were normalized to chicken β-actin expression.

Reporter assays

For luciferase reporter assays, cells were seeded in triplicate into 24-well plates at 2.5 × 10⁵ cells/well and co-transfected with the indicated reporter plasmid, pRL-TK vector (Promega) encoding Renilla luciferase, empty pcDNA 3.1, or pc3.1-GHR-AS-EST overexpression vector using Lipofectamine 3000. At 48 h after transfection, cells were lysed with passive lysis buffer, and luciferase activity was measured using the Dual-Luciferase® Assay System (Promega) per the manufacturer’s protocols. For GFP detection, LMH cells were inoculated into 24-wells plates at 2.5 × 10⁵ cells/well. The cells were transfected with pV1-EGFP and pASP-EGFP using Lipofectamine 3000. At 48 h after transfection, fluorescence was observed under a Leica DMi8 system (Leica, Germany).

Endogenous ribonuclease protection assay

To detect the RNA duplex formed by GHR sense and antisense RNAs, endogenous ribonuclease protection assay was performed as previously described [16,31,40], with minor modifications, using total RNA from chicken liver and leg muscle tissues, LMH cells and primary myoblasts. To remove genomic DNA and single-strand RNA, the total RNA was treated with DNase I (TaKaRa, Otsu, Japan) at 37°C for 30 min and then digested with RNase A (20 ng/μL; TaKaRa) 37°C for 1 h. Following the RNase A protection assay, we used RT-PCR to detect duplex formation. Five sets of primers were designed for PCR: a first set was targeted to the GHR-AS-EST full-length region of the GHR sense and antisense transcripts, and the other sets were targeted to different overlapping regions of these transcripts (Table S1). PCRs were run in 35 cycles using 25-μL reactions and 2 × Taq PCR Master Mix (Tiangen). Products were checked by 2% agarose gel electrophoresis.

RNA stability assay

RNA stability was assayed as described previously [25]. Chicken primary myoblasts transfected with 1 μg of pc3.1-GHR-AS-EST or empty pcDNA3.1 were seeded into 12-well plates. After 48 h, the cells were treated with 2 μg/mL actinomycin D (Sigma-Aldrich, St. Louis, MO, USA) to inhibit transcription and harvested 0, 2, 4 h, 6 h, 8 h, and 10 h later. Total RNA was extracted for RT-PCR and residual mRNAs were detected by RT-qPCR. Three independent samples were detected for each data point, GAPDH was used as internal control gene because its half-life is more than 30 h [25].

Western blotting

Cellular protein was extracted using lysis buffer (0.5% Triton X-100, 150 mM NaCl, 50 mM Tris–HCl, 2 mM EDTA, and 1 mM PMSF). The proteins were separated by 8% SDS-PAGE and transferred onto a Hybond membrane (Amersham). Membranes were blocked with 5% milk in phosphate-buffered saline (PBS) for 1 h and then incubated with the primary antibodies anti-Flag (1:1000) and β-tubulin (1:1000) (Beyontime) for 2 h. Then, the membranes were probed with antibody anti-mouse IgG (Beyontime). Target proteins were detected using an enhanced chemiluminescence detection system (Pierce, Rockford, IL, USA).

Fluorescence co-localization assay

To generate cells co-expressing GH_CFP, GHR_EGFP, GHBP_YFP, 0.7-kb GHR_BFP, and GHR-AS-EST_DsRed, LMH cells were co-transfected with 0.5 μg of each of the corresponding plasmids using Lipofectamine 3000. After 48 h of culture, the cells were washed with cold PBS and fixed in cold 4% formaldehyde in PBS for 4 min at room temperature. Images were captured under a confocal laser scanning microscope (Leica). Filters used for observing GH_CFP allowed excitation at 458 nm and emission at 465–495; for GHR_EGFP, allowed excitation at 488 nm and emission at 495–525 nm; for GHBP_YFP, allowed excitation at 513 nm and emission at 518–548 nm; for 0.7 kb GHR_BFP, allowed excitation at 382 nm and emission at 435–465 nm; and for GHR-AS-EST_DsRed, allowed excitation at 558 nm and emission at 570–600 nm.

FRET measurements

To generate co-expressing GH_BFP/GHBP_DsRed, GH_BFP/0.7-kb GHR_DsRed, GH_BFP/GHR-AS-EST_DsRed, GHBP_BFP/0.7-kb GHR_DsRed, GHBP_BFP/GHR-AS-EST_DsRed, and GHR-AS-EST_BFP/0.7-kb GHR_DsRed cells, chicken primary myoblasts were co-transfected with 2 μg of each group of these six constructs using Lipofectamine 3000. After 48 h of culture, the cells were image under the Leica confocal laser scanning microscope as described above. Filters used for observing GH_BFP, GHBP_BFP, or GHR-AS-ESR_BFP allowed excitation at 382 nm and emission at 435–465 nm; for GHBP_DsRed, 0.7-kb GHR_DsRed, or GHR-AS-EST_DsRed excitation at 558 nm and emission at 570–600 nm was used; for FRET of these six constructs, excitation was at 382 nm and emission at 570–600 nm.

Cell proliferation assay

Cells transfected with pc3.1-GHR-AS-EST or si-GHR-AS-EST and control cells were harvested after 72 h of culture, washed with cold PBS, and fixed in 1 mL of cold 70% ethanol. After overnight incubation at – 20°C in ethanol, cells were washed in PBS and suspended in 500 mL propidine iodide (PI) solution for 30 min before flow cytometry. Populations in G1, S, and G2 phases were measured by flow cytometry (Beckman, CA, USA), and the data were analyzed using Multicycle-DNA Cell Cycle Analysis Software. Measurements were conducted in quadruplicate. Additionally, transfected cells were seeded in 96-well flat-bottom plates at 2 × 10³ cells/well in 200 μL of medium and were cultured for 12 h, 24 h, 36 h, and 48 h. The Cell Counting Kit-8 (CCK-8) (Transgen, Guangzhou, China) was used per the manufacturer’s instructions to assay proliferation. DNA synthesis in proliferating cells was determined by the EdU assay, using the Cell-Light^TM EdU Appolo567 In Vitro Kit (Ribobio, Shanghai, China) following the manufacturer’s protocol. Treated cells were observed under the Leica fluorescence microscope. Experiments were repeated at least three times in duplicate.

Oil Red O staining

To analyze the effect of GHR-AS-EST on fat deposition in the chicken liver, LMH cells transfected with si-GHR-AS-EST or pc3.1-GHR-AS-EST and control cells were stained with Oil Red O. After 48 h of transfection, cells were washed twice with cold PBS, fixed with 10% paraformaldehyde for 30 min, stained with Oil Red O at room temperature for 15 min, and observed under a light microscope (Nikon, Tokyo, Japan).

Statistical analysis

All experimental data are as means ± SEMs of at least three independent experiments. SPSS software was used for statistical analysis. Student’s t-test was used for individual comparisons. Difference between groups was considered statistically significant if p < 0.05.

GHR-AS-EST sequence similarity analysis

The whole sequence of chicken GHR-AS-EST was determined in this study, and GHR-AS sequences for nine other species (human: AA398260.1, cattle: CK848828.1, dog: CX009677.1, house mouse: BU757735 and BY775223.1, sheep: S78251, Japanese medaka: DC276396, pig: AJ940615 and AJ687535, African clawed frog: BI441774, Norway rat: AW533500) were downloaded from the NCBI UniGene database. The DNAStar-EdiSeq tool was used to translate the potential coding regions of each GHR-AS sequence. Similarity among either nucleotide or proteins sequences was assessed using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/) [51,52]. Nnucleotide similarity between two species was tested by pairwise searches using BLAST2N [53], and pair distances were analyzed using MEGA6.0 software [54]. Similarity of potential coding regions was assessed using BLAST Multiple Alignment Tool [55] and that between two species was tested using BLAST2P with a similarity E-value <0.001 [53].

Funding Statement

This work was funded by the China Agriculture Research System (CARS-41-G03), the National Natural Science Foundation of China (31672412), and the Graduate Student Overseas Study Program from South China Agricultural University (Grant No. 2017LHPY016).

Acknowledgments

We thank Prof. Yunfeng Wang and Yan Zhao (Harbin Veterinary Research Institute, Chinese Academy of Agricultural Science, Heilongjiang, China) for giving us LMH cells.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Author Contributions

SL wrote the paper, analyzed data and designed the figures; SL, WL, QN and XZ conceived and designed this study; SL, ZZ, and TX performed experiments; both BH and ZR helped a lot during these experiments; LZ, CL, and CL put forward many precious suggestions to improve the quality of this draft; X.Z. gave a lot of valuable opinions for modification and finalized the manuscript.