Introduction

Genomic imprinting is an essential epigenetic phenomenon in mammalian cells that causes monoallelic expression of a small amount of genes with a parent-specific pattern (Barlow and Bartolomei 2014). Gene imprinting regulates embryonic and placental growth, and affects animal postnatal development and the occurrence of disease (Peters 2014). To date, about 150 chromosomal loci are identified to be imprinted in humans (Wei et al. 2014; Partida et al. 2017). Aberrant regulation of the imprinted loci is associated 12 congenital diseases (Soellner et al. 2017). Compared to humans and mice, few imprinted genes have been reported in cattle. Most imprinted genes are conserved in humans and mice (Peters 2014). Genomic imprinting evolution and biological significance have been explained by various theories, and comparative expression and imprinting status analysis among species are useful tools for learning the biological significance of genomic imprinting.

Solute carrier family 38 member 4 (slc38a4), also named ATA3, is a member of the System A gene family and plays a major regulatory role in the sodium-dependent amino acid transport system (Sugawara et al. 2000). This gene is paternally expressed in all tested mice tissues (Mizuno et al. 2002); however, it shows complex isoform- and tissue-specific maternal expressions in swine (Bischoff et al. 2009). The first aim of this study was to assess the expression pattern and imprinting status of the SLC38A4 in bovine adult tissues and placenta.

Genomic imprinting is controlled by imprinting control regions (ICRs) that possess differentially methylated regions (DMRs) derived from the germline, which retains the modification in somatic tissues (Reik and Walter 2001; Docherty et al. 2014). A maternally methylated DMR exists at the mouse slc38a4 promoter region (Chotalia et al. 2009). The other objective of this study was to investigate the role of DNA methylation modifications in imprinting of the slc38a4 gene by analyzing the DNA methylation status of three CpG rich regions around the bovine SLC38A4 promoter.

Materials and Methods

Tissue Collection

Holsteins bovine tissue samples (n = 32) used in this study were collected from a local abattoir. 30 bovine term placental samples (consisting of multiple sites), with corresponding maternal whole-blood samples and paternal sperms’ samples being used for artificial insemination of normal pregnancies, were collected from a local cattle farm. Whatever the source, all collected samples were frozen in liquid nitrogen for later DNA and RNA extractions, respectively.

DNA Extraction and Single Nucleotide Polymorphism (SNP) Identification

Genomic DNA was extracted from liver and placental (including maternal blood and paternal sperms) samples using the DNA Extraction Kit (Sangon Biotech, Shanghai, China). The exonic SNP was determined by sequencing the PCR products directly. The primers SLC-Fg (5′-TCGTGGTTGGCATCATCT-3′) and SLC-Rg (5′-GGGGAGAGGGAAAAGAGTAT-3′) were designed based on the bovine SLC38A4 sequence (GenBank, NM_001205943). The 25 μL PCR volumes contained 1 μL of DNA template, 1 μL of each primer (10 μM), 9.5 μL of ddH2O and 12.5 μL of ES Tap Master Mix (CWBio, Beijing, China). The PCR was performed as follows: an initial denaturation at 94 °C for 4 min, followed by 30 cycles (at 94 °C for 30 s, 56 °C for 30 s, 72 °C for 30 s) and a final extension at 72 °C for 10 min. The 416 bp PCR products were purified using a UNIQ-10 column DNA gel extraction kit (Sangon Biotech, Shanghai, China) according to the manufacturer’s instructions and sequenced directly to be sent to Bio sequencing Corporation (BGI, Beijing, China).

RNA Preparation and Reverse Transcription

Total RNAs were isolated from the frozen tissues by a Trizol RNA extraction kit (Invitrogen, Carlsbad, CA, USA) following the instructions of the manufacture. To remove possible genomic contamination, total RNAs were treated with RNase-free DNase I. The quantity and quality of the RNA was assessed using NanoDrop ND-1000 spectrophotometer system.

Subsequently, Total RNA (2 μg) was reverse-transcribed to cDNA using a reverse transcription (RT) kit (Promega, Madison, WI) according to the manufacturer’s protocol.

Expression Analysis

Expression of the SLC38A4 gene in different tissues was detected by RT-PCR. Total RNA prepared from three individuals was reverse transcribed into cDNA for use as a template. The intron-spanning primers, SLC-F (5′-TACACATTTGACACCCCTCTC-3′) and SLC-R (5′-AATACATCATCACACACTGCTCTA-3′), were designed to analyze bovine SLC38A4 gene expression by RT-PCR. The GAPDH, a housekeeping gene, was amplified as the internal control with the primers Gap-F (5′-GCACAGTCAAGGCAGAGAAC-3′) and Gap-R (5′-GTGGCAGTGATGGTGGA-3′) designed according to the sequence (GenBank, BTU85042). The RT-PCR products were purified and directly sequenced.

Allelic Expression Analysis

Allelic expression was investigated by comparing the sequence results of PCR products from genomic DNA and cDNA of SLC38A4. Total RNA isolated from organ samples and placentas heterozygous individuals was reverse-transcribed into cDNA and used for the allelic transcription analysis of the bovine SLC38A4 gene.

Methylation Analysis

Three CGIs were found surrounding the SLC38A4 promoter and exon 1 using the online SoftBerry program (http://www.softberry.com/). For each CGIs, primers were designed for nested or seminested PCR with the online software for methprimer (http://www.urogene.org/methprimer). Primer sequences are shown in Table 1.

Table 1 Primers used for bisulfite sequencing
Full size table

DNA was extracted from three heterozygote placentas (6, 13 and 15) and their corresponding paternal sperms samples. Genomic DNA (800 ng) was bisulfate treated using the EZ DNA Methylation Kit (Zymo, Orange County, CA) following manufacturer’s guidelines. These bisulfite-converted DNAs were used as templates for amplification of specific CGIs. One microliter of tenfold dilution of the first-round PCR product was the template of the second-round PCR. PCR products of second round were purified and cloned into pMD19-T vectors (Takara). At least twenty clones containing the amplified fragments were sequenced. The SNPs in each CGI were used to distinguish the two parental strands. In each individual, the percentage of overall mCpGs of both parental strands was calculated, respectively. If the difference of methylation levels between the two parental strands is over 50% in a CGI, then this region was defined as a DMR.

Results

Biallelic Expression of SLC38A4 in Bovine Adult Tissues

A 683 bp fragment was amplified using primers SLC-F and SLC-R in eight tissues (heart, liver, spleen, lung, kidney, muscle, fat. and brain), indicated that SLC38A4 was expressed in all detected tissues and showed a higher expression level in the liver than that in other tissues (Fig. 1a).

Fig. 1
figure 1

Biallelic expression of SLC38A4 in adult bovine tissues. a Expression of the bovine SLC38A4 transcripts obtained from eight adult tissues analyzed by RT-PCR. M, DL2000 (2000, 1000, 750, 500). Lanes 1–8: RT-PCR products of heart, liver, spleen, lung, kidney, muscle, fat. and brain, respectively. Lane 9: negative control. The sizes of RT-PCR products were 683 bp for SLC38A4 and 375 bp for GAPDH. b SNP site of the SLC38A4 gene (an A/G SNP). c Biallelic expression of SLC38A4 in eight detected bovine tissues

Full size image

SNP site was searched in the coding sequence of the bovine SLC38A4 gene in 32 individuals, and an A/G transition (GenBank, rs137028117) was identified in six individuals (Fig. 1b). Three out of six heterozygous individuals were used to analyze the allele expression of SLC38A4. Comparison of the sequencing results of RT-PCR and genomic DNA PCR products indicates that biallelic expression of SLC38A4 occurs in all eight analyzed tissues (Fig. 1c). This suggests that SLC38A4 is not imprinted in bovine adult tissues.

Polymorphic Imprinting of SLC38A4 in Bovine Placenta

The approach used for detecting the SNP and expression of SLC38A4 in adult tissues was also employed to investigate placentas (Fig. 2a). Five heterozygous placentas were identified from 30 samples based on the A/G SNP (GenBank, rs137028117) (Fig. 2b). Interestingly, the expression of SLC38A4 in the five heterozygous placentas (placenta 6, 11, 13, 15, and 22) was different. Monoallelic expression occurred in placenta 6, 15 and 22, while biallelic expression occurred in placenta 11 and 13 (Fig. 2c). To distinguish between paternal and maternal expressions in monoallelic expressed placentas (samples 6, 15, and 22), the parental genotype was analyzed by amplifying the genomic DNA of the corresponding maternal blood and paternal sperms samples. In placenta 6 and 15, the homozygous GG and AA genotypes were identified in the corresponding paternal and maternal genomic DNAs, respectively, thus indicating that SLC38A4 exhibits paternal expression (allele G) in the placenta. For placenta 22, the genotypes of his paternal sperm and maternal blood sample were the homozygous AA and heterozygous AG, respectively, so the A in the offspring cDNA originated from the sire. These results indicate that SLC38A4 is polymorphically imprinted in cattle.

Fig. 2
figure 2

Polymorphic imprinting of SLC38A4 in bovine placentas. a Expression of the bovine SLC38A4 in heterozygous placentas by RT-PCR. M, DL2000 (2000, 1000, 750, 500). Lanes 1–5 are RT-PCR products obtained from heterozygous placentas (placenta 6, 11, 13, 15, and 22), respectively. Lane 6 is negative control. The sizes of RT-PCR products were 683 bp for SLC38A4 and 375 bp for GAPDH. b Sequencing results of SNP (rs137028117) in the SLC38A4 gene. Polymorphic imprinting expression was deduced by comparing genotypes of gDNA, cDNA and parental gDNA in five heterozygous placentas

Full size image

DNA Methylation is not Involved in Regulating Polymorphic Imprinting of SLC38A4 in Placenta

CpG plot analysis (http://www.ebi.ac.uk/Tools/emboss/cpgplot/) shows that there are three CGIs in the region surrounding the bovine SLC38A4 5′ promoter and exon 1. To analyze whether methylation of the SLC38A4 promoter region regulates imprinting expression in bovine placenta, the DNA methylation status of three CGIs were tested in three placentas and their corresponding paternal sperms. SLC38A4 expression exhibited paternal imprinting in two placentas (6 and 15), while biallelic expression was detected in placenta 13. The three CGIs selected for bisulfite sequencing were located on the bovine SLC38A4 gene from −234 to +260 bp, from +339 to +636 bp, and from +645 to +982 bp, and their 21, 24, and 24 CpGs were analyzed, respectively (Fig. 3).

Fig. 3
figure 3

Methylation profiles of three CGIs in bovine SLC38A4 in three placentas and their corresponding paternal sperms. The locations of three CGIs are shown surrounding promoter and exon 1. The SNPs in three CGIs are indicated by red bar. Methylated and unmethylated residues are indicated by filled and open circles, respectively. The percentages of methylated CpG sites of the three heterozygous placentas and their corresponding sperms are shown at the left side of the methylation pictures (Color figure online)

Full size image

There was a SNP site in each studied region: an A/T SNP (GeneBank, rs211032768) in CGI 1; an A/C SNP (GeneBank, rs207735752) in CGI 2; and an A/G SNP (GeneBank, rs209305091) in CGI 3. These sites were used to distinguish two parental allele strands. Extreme hypomethylation (< 3%) was observed at the three CGIs in the three detected placenta and their corresponding paternal sperms. Furthermore, no significant difference was observed between the paternal and maternal strands. These results suggest that DNA methylation of these three regions does not appear to directly or indirectly affect the allelic transcription of bovine SLC38A4.

Discussion

In human, the SLC38A4 gene encodes a neutral amino acid transporter and has a significant impact on the fetus and placental development (Desforges et al. 2006; Li et al. 2012). The slc38a4 gene has different allelic expression patterns in mouse highly expressed placenta and liver, where paternal expression is observed in the placenta and biallelic expression is observed in the liver (Smith et al. 2003). The imprinted expression of Slc38a4 can be disturbed by environmental endocrine disruptors in mice yolk sac (Kang et al. 2011). In this study, we report the biallelic expression of SLC38A4 in adult bovine tissues, which is consistent with the previous observation that SLC38A4 is not imprinted in the fetus (Zaitoun and Khatib 2006). SLC38A4 shows a complex isoform- and tissue-specific maternal expression in swine (Bischoff et al. 2009). In short, the imprinting patterns of SLC38A4 are different between mice, swine, and bovine.

Polymorphic imprinting refers to that the imprinting of a gene is variable between individuals. This epigenetic phenomenon has been observed in the human 5-HT2A and IGF2 genes (Bunzel et al. 1998; Sakatani et al. 2001). These variations in gene expression are considered a cause of phenotypic heterogeneity in human disease (Weinstein 2001). In this study, we obtained evidence for monoallelic (paternal) expression of SLC38A4 in bovine placenta; however, biallelic expression of SLC38A4 was also observed in other placenta, indicating that SLC38A4 imprinting in the bovine population is polymorphic. Although the number of known genes demonstrating polymorphic imprinting is increasing, the molecular mechanisms underlying this phenomenon are largely uncharacterized. Polymorphic imprinting of the mouse Ppp1r9a and Kvlqt1 genes are strain dependent, as differences in genetic background seem to affect the imprinting regulation of the two genes (Jiang et al. 1998). Polymorphic imprinting of the human WT1 gene in placenta samples may be caused by alterations of the key sequence essential for WT1 imprinting, which leads to an alteration of cis-acting factors.

SLC38A4 has a maternally methylated germline DMR in the promoter region of mice (Chotalia et al. 2009). In the cumulus cells-derived cloned mouse, the SLC38A4 DMR showed maternal allele-specific methylation in donor cells that showed paternal expression, but not in brain of cloned mouse with biallelically SLC38A4 expression. In the present study, there was no DMR found in the bovine SLC38A4 promoter region in the placenta with imprinting expression of SLC38A4. These results indicate that polymorphic imprinting of bovine SLC38A4 is not regulated by the methylation status of CGIs at the promoter, and that an epigenetic modification other than DNA methylation might be needed for establishing SLC38A4 imprinting in bovine placenta.