Introduction

Pigs have become an important animal species for biomedical research and agriculture. Recent advances in genome editing technologies have facilitated the generation of pigs resistant to disease as well as models for studying human diseases and for xenotransplantation (Redel and Prather 2015; Ryu et al. 2018; Yang and Wu 2018). Sex selection of genetically modified embryos or cloneable cells is critical to the efficiency of every project. For projects that involve sex-specific phenomena (lactation, spermatogenesis, etc.), the sex selection of founders reduces the number of animals required to determine the utility of individual genetic events or to determine the value of additional efforts. Similarly, since males can produce offspring at a much higher rate than females, if is often valuable to limit founder production to males only. Lastly, for projects that require founders of both sexes, there is value in being able to balance or bias the sex of founder animals.

The ability to distinguish males from females also has a substantial relevance for wildlife studies to understand population characteristics and to improve their management. To date, PCR-based methods for sex determination have been described for more than one hundred mammalian species (Hrovatin and Kunej 2018; Strah and Kunej 2019). Most of these assays were designed based on sex-specific insertions and deletions (SSIndels) to target the zinc finger protein (ZFX and ZFY) or the amelogenin (AMELX and AMELY) gene variants on X and Y chromosomes (Strah and Kunej 2019). In pigs, PCR-based methods for sex determination involve a duplex PCR-amplification of a Y chromosome specific sequence and simultaneous amplification of a non-sex discriminative sex region as endogenous control (Pomp et al. 1995; Sathasivam et al. 1995); or the amplification of X–Y homologous genes, more specifically the amelogenin gene variants AMELX and AMELY by using a single pair of primers (Fontanesi et al. 2008; Sembon et al. 2008). In humans and cattle, mutations at the amelogenin genes in the Y and X chromosomes have been reported (Grzybowski et al. 2006; Dash et al. 2020) showing inconsistencies in the expected amplicons size. This can result in incorrect sex identification, thus questioning the use of the amelogenin gene for sex determination. In pigs, the AMEL-PCR method was not sensitive enough to dependably amplify the expected products from low cell number samples–oocytes, and early embryos (Sembon et al. 2008). Therefore, the development of a new, rapid, sensitive, and inexpensive method for sex determination in pigs is desired.

Studies aiming to characterize the mammalian sex chromosomes have helped to identify the genes involved on sexual differentiation. Both Y and X chromosomes have been undergoing structure and content differentiation for more than 170–180 million years (Skinner et al. 2016). The mammalian X chromosome has been shown to be very conserved between different mammalian species retaining much of its ancestrally. On the other hand, the gradual degeneration of the Y chromosome due to evolutionary events with the loss of 95% of the ancestral genes limits its characterization (Marshall Graves 1998; Liu et al. 2019). The presence of highly repetitive regions makes difficult to generate complete Y chromosomes sequence data, resulting in few species with a complete and well assembled Y chromosome sequence. However, after sequencing the Y chromosome from primates; mice, cats, dogs and cattle three genes (USP9Y, DDX3Y and UTY) were discovered to be conserved throughout suggesting a potential role on sexual development (Skinner et al. 2016; Liu et al. 2019).

DDX3X and DDX3Y, also known as DBX and DBY respectively, are members of the DEAD box protein family and are present on both sex chromosomes. In humans, DDX3Y gene is located in a Y chromosomal region required for spermatogenesis and its expression is restricted to the male germ line. DDX3X is present in a X chromosome region and it was found to be expressed ubiquitously in tissues (Ditton et al. 2004). Recently, an assay based on the DDX3X/Y gene for sex identification in bats was able to correctly confirm the sex in 100% of samples of known sex (22 males and 17 females) (Zarzoso-Lacoste et al. 2018). In bovine, it was shown that the DEAD box protein (DDX3X/Y) gene was able to determinate by PCR the sex origin of beef products. The assay showed to be reliable and resulted in 100% of success rate when meat samples (14 males and 14 females) of known sex were tested (Gokulakrishnan et al. 2012). Similarly, a sex determination assay using DDX3Y and DDX3X genes was developed for three Japanese Mustelid species using tissue samples which sex was previously recorded. The assay produced the expected results for both female and male samples (Sekiguchi et al. 2010).

The possibility of sex-specific selection by using the DDX3X/Y gene emerges as an important factor in the production efficiency of porcine models for biomedical and agriculture purpose. Based on the need to identify sex in embryos and early fetuses, we developed a novel method by using the DDX3X and DDX3Y genes to identify the genetic sex in pigs by using a unique primer set to amplify either DDX3X or DDX3Y. We also successfully applied the DDX3X/Y method to sex confirmation and identification of porcine oocytes, embryo biopsies, Day-6 blastocysts and Day-35 fetuses, and also evaluate the application of this assay across different mammalian species.

Methods

Tissue samples and DNA extraction

Genomic DNA was extracted from tails of 87 pigs representing a variety of different lines (Table 1). Additionally, we collected pig hair samples and performed sex identification of 11 Day-35 fetuses. Female and male genomic DNA samples from Bison bison, liver samples from Bos taurus as well as endometrium and testicles samples from Ovis aries were kindly donated from Dr. Robert Schnabel, Msc. Emma Stephenson, Msc. Eleonore O’Neil and Dr. Jon Green, respectively, from the University of Missouri.

Table 1 Porcine lines utilized for tail DNA extraction
Full size table

For DNA isolation (tail, liver, endometrium and testicle), tissue was resuspended in 500 µL of lysis buffer I (LBI) (40 mM Tris, pH 8.9; 0.9% Triton X-100; 0.9% Nonidet P-40; 0.4 mg/mL proteinase K) and incubated at 65 °C for 15 min and then heated to 95 °C for 10 min. The DNA was purified by organic extractions of phenol and chloroform, and then precipitated with ethanol. Single hair follicles and a small sample of tissue from Day-35 fetuses were taken and suspended in 10 µL of LBI; however, no organic extractions of phenol and chloroform were performed for these samples. Genomic DNA (0.5 ng) from tissue samples and cell lysate sample from fetuses and hair samples (1 µL) were used to determine sex by PCR.

In vitro embryo production

Ovaries were collected from slaughtered prepubertal gilts according to the approved protocol and standard operating procedures by the Institutional Animal Care and Use Committee of the University of Missouri. Cumulus–oocyte complexes were aspirated from follicles 3–6 mm in diameter with a sterile 18-gauge needle attached to disposable syringe. Only cumulus–oocyte complexes with uniform cytoplasm and at least three layers of surrounding cumulus cells were subjected to in vitro maturation for 42–44 h at 38.5 °C in 5% CO2/humidified air, as described previously (Zhang et al. 2010; Chen et al. 2018). After maturation, the cumulus cells were removed from oocytes by vortexing for 3 min in 0.1% (wt/vol) hyaluronidase and thirteen Metaphase II (MII)-stage oocytes, identified by the presence of the first polar body, were collected for DNA extraction. The remaining matured oocytes were washed and placed into 50 μL droplets of IVF medium (modified Tris-buffered medium containing 2 mg/mL bovine serum albumin (BSA) and 2 mM caffeine) and incubated at 38.5 °C until sperm was added. For IVF, a 0.1-mL frozen sperm pellet was thawed in 3 mL of sperm washing medium (Dulbecco’s phosphate-buffered saline (Gibco, Gaithersburg, MD) supplemented with 0.1% FAF-BSA and 10 μg/mL gentamicin). The sperm was washed by centrifugation in 45% Percoll solution and then in modified Tris-buffered medium. After, the semen pellet was resuspended in IVF medium and the concentration was adjusted to 0.5 × 106 cells/mL.

The fertilization drops containing the oocytes received 50 μl of the sperm suspension (final concentration of 0.25 × 106cells/mL) and gametes were co-incubated at 38.5 °C in 5% CO2/humidified air for 4 h. Presumptive zygotes were washed and cultured in groups in 500 μL of porcine zygote medium 3 plus 1.69 mM arginine and 5 μM PS48 (MU2) in a four-well dish at 38.5 °C in 5% CO2/humidified air. After 28–30 h of cultured the cleaved embryos were moved to a humidified atmosphere of 5% CO2, 5% O2, and 90% N2 at 38.5 °C until day 6 post-fertilization. Day-6 blastocysts were washed in TL-Hepes-buffered saline and transferred to a physiological saline solution at pH 1.79 until the zona pellucida (ZP) and any sperm attached were removed. The embryos were washed in TL-Hepes buffered saline and collected in individual tubes for DNA isolation.

Parthenogenetically activation (PA) of porcine oocytes

Metaphase II (MII)-stage oocytes were selected for TPEN activation (200 μM) for 30 min in TL-Hepes-buffered saline (Lee et al. 2015). Activated oocytes were cultured in MU2 medium supplemented with cytochalasin B (7 μg/uL) for 4 h, washed three times and then cultured in MU2 in a four-well dish at 38.5 °C in 5% CO2/humidified air. After 28–30 h of cultured the cleaved embryos were moved to a humidified atmosphere of 5% CO2, 5% O2, and 90% N2 at 38.5 °C until day 7 post-activation.

Oocytes and embryos DNA extraction

Fifty-seven Day-6 in vitro-produced (IVP) porcine embryos and thirteen metaphase II (MII)-stage oocytes were collected for DNA isolation. Single porcine blastocyst and MII oocytes were suspended in 5 µL of lysis buffer I (LBI) (40 mM Tris, pH 8.9; 0.9% Triton X-100; 0.9% Nonidet P-40; 0.4 mg/mL proteinase K) and incubated at 65 °C for 15 min and then heated to 95 °C for 10 min. Cell lysate from each blastocyst stage embryo (2 µL) and each MII oocyte (2 µL) were used to determine sex by PCR.

Parthenogenetically activated (PA) and in vitro fertilized (IVF) embryos biopsy

A total of twelve Day-7 in vitro-produced embryos and twelve parthenogenetically embryos (all female) were placed in drops of approximately 100 µl of TL-Hepes-buffered saline for biopsy. Then, a portion of the trophectoderm (TE) was manually sectioned by using a needle blade (Fine Sciences Tools, Foster City, CA) (Bredbacka et al. 1995). The biopsies were collected in individual tubes and suspended in 3 µL of lysis buffer I (LBI) (40 mM Tris, pH 8.9; 0.9% Triton X-100; 0.9% Nonidet P-40; 0.4 mg/mL proteinase K) and incubated at 65 °C for 15 min and then heated to 95 °C for 10 min. Cell lysate from each embryo biopsy (1 µL) was used to determine sex by PCR.

Polymerase chain reaction

Primers DDX3-R(GCCACTAGAATTGGGCTTTTTCCT) and DDX3-F (TGCTTGCTCGTGATTTCTT GGA) were designed in regions of same identity between the X and Y porcine chromosomes and were expected to yield a PCR-fragment of 485-bp for DDX3X target sequence and 545-bp for DDX3Y. PCR reactions were performed by using different sources of DNA polymerase as OneTaq Hot Start (NEB,USA) DNA Polymerase or LA Taq Polymerase (TaKaRa Bio). Two different melting temperatures (64 °C for Sus scrofa; 58 °C for Bos Taurus, Bison Bison and Ovis aries) and two different extension temperatures (68 °C for OneTaq protocol or 72 °C for LaTaq protocol) were defined according to the species and DNA polymerase applied. OneTaq PCR mix was prepared by adding the equivalent to 0.1 µL of DNA polymerase, 2.5 µL of 1 × One Taq Standard Buffer, 0.5 µL of 10 mM dNTP mix solution and 1 µL of each forward and reverse primers (10 µM) adjusting the final reaction volume with water to 12.5 µL. For the LaTaq mix we added 0.2 µL of DNA polymerase, 8.3 µL of 2 × GC Buffer I, 2.7 µL of 2.5 mM dNTP mix solution, 1 µL of each forward and reverse primers (10 µM), and water to complete a final volume of 16.7 µL. Amplification conditions were initial denaturation step at 94 °C for 30 s followed by 35 cycles of denaturation at 94 °C for 15 s, annealing for 30 s, extension for 1 min with an addition of 1 s per cycle, followed by a final extension for 2 min by using a Mastercycler EP Gradient machine (Eppendorf Hauppauge, NY). PCR products were separated by electrophoresis on a 2% agarose gel. Representative images from our sex determination assay are shown in Fig. 1. The sex determination assay developed by Hao et al. (2006) was applied to analyze samples for which the gender was not previously available. However, this assay was not successful for oocytes, biopsies and blastocysts samples because of their limited amounts of genomic DNA present.

Fig. 1
figure 1

Sex determination based on the amplification of porcine DDX3X and DDX3Y genes. a PCR-amplification of DDX3X and DDX3Y fragments from different pig breeds: \(7/8\) Minnesota Mini \(1/8\) Domestic (lanes 1 to 6), 1/2 Domestic 1/2 NIH c/c mini (lanes 7 to 16), 1/2 Domestic 1/2 Minnesota Mini (lanes 17 to 21), 1/4 Minnesota mini 3/4 Domestic (lanes 22 and 23) b Sex determination of Day-35 fetuses (lanes 1 to 11) c Sex determination of Day-6 in vitro-produced embryos (lanes 1 to 11) and GV-stage oocytes (lanes 12 to 16). In each lane the presence of two bands (545 and 485 bp) indicates male and one band (485 bp) female. L: pBR322/MspI ladder (New England Biolabs, Beverly, MA), CM: male positive control (male genomic DNA), CF: female positive control (female genomic DNA), N: Non-template control. Sex indicated in which line ( : male; : female)

Full size image

Isolation and sequencing of DDX3 fragments

DDX3 PCR products from one boar and one sow (1/2 Domestic 1/2 NIH c/c mini) were cloned into a vector for sequencing (pCR4- Topo, Invitrogen, Grand Island, NY) by the manufacturer’s protocol. The TOPO cloning reactions were transformed into High Efficiency DH5α chemically competent Escherichia coli cells (NEB) and plated onto Luria–Bertani/ChromoMax IPTG X-Gal (Thermo Fisher Scientific)/Kanamycin (30 mg/mL) agar plates overnight. For each pig, 5 clones were selected and submitted for DNA isolation by using a PureLink Quick Plasmid Miniprep kit (Invitrogen). The clones were subject to traditional bisulfite (Sanger) sequencing at the University of Missouri-Columbia DNA core facility and then, the sequences obtained were submitted to a global alignment based on Needleman-Wunsch algorithm by using the Pairwise Sequence Comparison (PASC) tool at the National Center of Biotechnology Information (NCBI) webpage.

In silico analysis of DD3X length polymorphism across different mammalian species

A sequence database was designed by first blasting (Blastn) our DDX3 X/Y-linked pig amplicons (including the primer binding sites) on all mammalian (sub)species (taxid: 40674) reference genomic sequences (refseq_genomic) available. As only few species have a high-quality Y chromosome assembly, just the species that had a query coverage of 100% for the DDX3Y pig amplicon and not more than 3 mismatches at the primer binding sites were selected. Then, the corresponding chromosome, gene and sex information were obtained for each sequence analyzed and the size difference between the DDX3Y and DDX3Y amplicons were predicted.

Statistical analysis

Statistical differences were analyzed by Fisher exact test (GraphPad Prism®). Significant differences were defined when p < 0.05.

Results

Porcine sex determination by DDX3X/Y Amplification among different lines and samples

Representative images from DDX3X/Y PCR amplification using genomic DNA or cell lysate samples are shown in Fig. 1. Analysis of cell lysate samples from single hair follicles and genomic DNA extracted from 87 pigs among different breeds coincided with the phenotypic sex (data partially shown, Fig. 1). Similarly, the results from Day-35 fetuses samples previously submitted to a different sex determination protocol (Hao et al. 2006) demonstrated 100% of concordance (Fig. 1 and Table 2). When genomic DNA from porcine blastocysts was used, the PCR analysis was able to identify the sex in 96.5% of the embryos with the ratio of male and female being 49.1 and 47.4%, respectively (Table 2). For GV-stage oocytes (all female) and embryo biopsies (Fig. 2 and Table 2) the identification rate was lower when compared to tail genomic DNA samples (84.6% and 83.3% vs. 100%), however there was no significant difference when compared to the other samples tested, P > 0.05 (Table 2).

Table 2 Identification of the DDX3X/Y bands in different porcine samples by using a PCR-based method
Full size table
Fig. 2
figure 2

Sex determination of porcine embryo biopsies based on the amplification of porcine DDX3X and DDX3Y genes. a Sex determination of biopsy samples from Day-7 in vitro-produced embryos b Sex determination of biopsy samples from Day-7 parthenogenetically activated (PA) embryos. L: pBR322/MspI ladder (New England Biolabs, Beverly, MA), N: Non-template control. Sex indicated in which line ( : male; : female)

Full size image

Sequence analysis of the porcine DDX3X/Y gene amplified fragment

The PCR-fragments showed the expected size of 485 and 545-bp and matched the nucleotide sequence for the targeted DDX3X and DDX3Y regions, respectively (data not shown). Comparing the DDX3Y and DDX3X fragments showed the presence of regions of homology which resulted in an identity of 76% between the Y and X chromosomes. In addition, several gaps resulting in size variation between the DDX3Y and DDX3Y amplicons were found (Fig. 3).

Fig. 3
figure 3

Alignment of the DDX3Y and DDX3X fragments of 1/2 Domestic 1/2 NIH c/c mini pigs. Underlined sequences indicate primers positions. -, alignment gaps; *, base substitution

Full size image

Sex determination by DDX3X/Y amplification across different mammalian species

In silico analysis confirmed the reliability of the DDX3X/Y gene for sex determination across different mammalian species. A total of 13 species demonstrated a X–Y amplicon length polymorphism for the DDX3 gene (Table 3). When we tested female and male DNA samples from Bison bison, we were able to confirm the sex previously recorded with PCR- fragments showing the expected in silico length differences of approximately 47 bp. Also, the assay was able to identify the sex in samples from Bos taurus and Ovis aries which were not included in the in silico analysis (Fig. 4). The number of species with query coverage greater than 95% for the X and Y chromosomes was 43 and 13, respectively. These discrepant values between X and Y assemblies restricted the number to 13 species analyzed.

Table 3 In silico validation of the porcine DDX3X/Y primer set across different mammalian species
Full size table
Fig. 4
figure 4

Sex determination based on the amplification of DDX3X and DDX3Y genes across different mammalian species. Sus Scrofa hair (lanes 1-4), Bos taurus (lanes 5-6), Ovis aries (lanes 7-8) and Bison bison samples (lanes 9-12). Sex previously recorded indicated in which line ( : male; : female). For all samples the expected results for both female and male samples were confirmed

Full size image

Discussion

In this study, we developed a new simple and robust PCR-based assay for sex determination in pigs by using a unique primer set to amplify the DDX3X/Y gene. As expected, DNA from females produced one band and DNA from males generated two bands. This method proved to be applicable among different samples and across different mammalian species. This assay becomes a new alternative to improve the efficiency of projects involving genetically modified embryos or cloneable cells, and to wildlife studies. The creation of pig models to decipher the mechanism of human diseases, to enhance porcine disease resistance and to test the effects of gene therapy and stem cells treatments have proven a significant progress in biomedical and agriculture research (Madeja et al. 2019).

Our DDX3X/Y approach was able to identify the sex in pigs from six different lines and diagnosis the sex from samples containing small amounts of genomic DNA. This assay also could be applied to samples that were collected non-invasively such as (hair), which often contain highly degraded DNA. The sex was 100% assigned when tail genomic DNA, Day-35 fetuses and hair samples were used. For biopsies samples, blastocysts and oocytes the PCR detection efficiency was lower (83.3%; 96.5% and 84.6%, respectively), however no detection errors were observed. We believe that the unidentified samples were due to PCR failure and the different diagnostic rates between embryos, biopsies samples and oocytes is likely due to limited amounts of genomic DNA present at GV stage oocyte (1 cell, 4C) and the embryo biopsies in comparison to blastocysts (~ 40 cells). In a previous study performed by Sembon et al. (2008) where they used an AMEL-PCR method to sex identification in pigs, they could not detect any bands in several oocytes and EA and IVF embryos samples. They attributed it to human errors when extracting the DNA genomic or setting up the PCR reactions.

Alignment of the porcine amplicons for DDX3Y and DDX3X genes showed a difference of 60 bp between DDX3Y and DDX3X fragments. However, both DDX3X and DDX3Y demonstrated a 100% of identity with the DDX3X/Y nucleotide sequence for Sus scrofa deposited at NCBI. Although we used pigs from a specific line, we believe that the DDX3 gene is very conserved when comparing different pig lines. Skinner et al. (2016) in an attempt to better understand the structure, the sequence, and evolution of the sex chromosomes observed that the USP9Y-DDX3Y-UTY is the only ancestral group of genes that have retained this pattern of rearrangement in pig, primates, mouse, cat, and dog. In cattle, a recent studied published by Liu et al. (2019) also demonstrated that the DDX3Y, USP9Y and UTY are still conserved in the X—degenerate region of the Y chromosome. DDX3Y gene has been shown an important role during human spermatogenesis suggesting its strong influence in sexual development. However, it may not be applicable for all primates and mice (Tyler-Smith 2008; Matsumura et al. 2019).

In vitro and in silico analysis confirmed the ability of the DDX3X/Y approach for sex diagnosis across different mammalian species. To our knowledge, this is the first study to confirm the application of the DEAD box protein gene for sex determination in pigs. Also, we were able to show that the same primers that were used for pig DDX3X/Y amplification could be successfully used for domestic sheep (Ovis aries), American bison (Bison bison) and European cattle (Bos taurus) samples. Our in silico analysis confirmed X–Y amplicon length polymorphisms for the DDX3 gene in other mammals as human, bat, dolphin, whale, deer, domestic goat, donkey and wild horse. Sex identification in some marine mammals are difficult to genotype based on their phenotype due to their poor sexual dimorphism and internal genitalia (Morin et al. 2005). We believe our DDX3X/Y PCR-based method can provide an easy and new alternative for sex determination in marine mammal specimens. Additionally, the ability to sex differentiate animals originated from Bos taurus and Ovis aries makes the DDX3X/Y PCR assay useful for reproductive and biomedical studies involving these important livestock animals. Lastly, another important application of this assay would be for the management of endangered species as Camelus ferus, Equus przewalskii and Neophocaena asiaeorientalis by improving the propagation rate of those species. We believe that with the continued advances in the genome sequencing tools, it will allow us to expand the use of this sex assay on additional species.

In conclusion, the ability to rapidly determine the sex of porcine embryos and cloneable cells, without the need of restriction endonuclease digestion or simultaneous amplification of an internal control, makes the DDX3X/Y an invaluable tool for transgenesis. Furthermore, the use of a unique primer for cross-amplification in different mammalian species can simplifying sex-linked studies. Our DDX3X/Y assay showed to be simple, specific, and highly reproducible showing a wide range of applications for researchers and breeding programs.