Identification of Novel Placentally Expressed Aspartic Proteinase in Humans

This study presents pioneering data concerning the human pregnancy-associated glycoprotein-Like family, identified in the genome, of the term placental transcriptome and proteome. RNA-seq allowed the identification of 1364 bp hPAG-L/pep cDNA with at least 56.5% homology with other aspartic proteinases (APs). In silico analyses revealed 388 amino acids (aa) of full-length hPAG-L polypeptide precursor, with 15 aa-signal peptide, 47 aa-blocking peptide and 326 aa-mature protein, and two Asp residues (D), specific for a catalytic cleft of the APs (VVFDTGSSNLWV91-102 and AIVDTGTSLLTG274-285). Capillary sequencing identified 9330 bp of the hPAG-L gene (Gen Bank Acc. No. KX533473), composed of nine exons and eight introns. Heterologous Western blotting revealed the presence of one dominant 60 kDa isoform of the hPAG-L amongst cellular placental proteins. Detection with anti-pPAG-P and anti-Rec pPAG2 polyclonals allowed identification of the hPAG-L proteins located within regions of chorionic villi, especially within the syncytiotrophoblast of term singleton placentas. Our novel data extend the present knowledge about the human genome, as well as placental transcriptome and proteome during term pregnancy. Presumably, this may contribute to establishing a new diagnostic tool for examination of some disturbances during human pregnancy, as well as growing interest from both scientific and clinical perspectives.


Introduction
Pregnancy-associated glycoproteins (PAGs) belong to a superfamily of aspartic proteinases (AP), which also include mammalian pepsins (A, C and F), cathepsins (D and E), renin and numerous other enzymes such as parasite plasmepsins and retroviral enzymes [1,2]. All AP members possess a two-bilobe structure with a cleft capable of short peptide binding and are classified into two subfamilies: catalytically active or potentially inactive due to several amino acid (aa) substitutions within two domains creating the binding cleft [3,4]. Among APs, pepsins fulfil digestive functions outside the cells, whereas cathepsin D and E are typical intracellular enzymes generally localized in the lysosomal compartment that provides the acidic environment necessary to accomplish their catalytic functions [5,6]. On the other hand, PAG-Like (PAG-L) family products revealed properties as various In addition, cDNA evaluation by capillary sequencing firmly confirmed the nucleotide sequence originating from the RNA-seq. Five pairs of homological PAG-L primers applied for PCRs allowed obtaining the entire cDNA sequence (1364 bp), named hPAG-L, and deposited in GenBank database (Acc. No. KX856064). Among the 109 electrophoresed, gel-out purified and sequenced cDNA amplicons, 80 high quality chromatograms (HQ range: 50-93.9%) were applied to estimate the coding (CDS) and non-coding untranslated regions (UTR). Among the identified 1364 bp of cDNA sequence, 1167 bp were determined as CDS, 20 bp as 5 UTR and 177 bp as 3 UTR (Figure 1).

Identification of cDNA Sequence Originating from Term Placental Transcriptome
The hPAG-L/pep polypeptide precursor retains two highly conserved domains (NH 2 and COOH), specific to other members of AP superfamily. Geneious ® 8.1.7 allowed identification of 15 aa-signal peptide (SP), 47 aa-blocking peptide and 326 aa-mature hPAG-L/pep precursor ( Figure 1).
The identified SP aa sequence of the hPAG-L/pep shared the highest similarity with SP of the human pep A and it varied with the other members of the AP family in different species (Table 1). Multiple alignments of the various APs enabled the prediction of 47 aa-blocking peptide (16-62 aa) of the hPAG-L/pep precursor that shared the highest homology with human pep A, whereas identity is equal/similar with peps C and F as well as other PAGs (Table 2).
A putative cleavage position was predicted between PTL60-62 of the blocking pro-piece and VDE63-65 of the mature hPAG-L/pep precursor ( Figure 1). Two Asp residues (D), specific for the catalytic cleft of AP were predicted at positions 94 aa in the NH 2 -terminus (VVFDTGSSNLWV91-102) and 277 aa in the COOH-terminus (AIVDTGTSLLTG274-285 Figure 1) of the hPAG-L precursor. The sequences of the NH 2 -and COOH-terminal domain of the hPAG-L/pep are identical to human pep A and very homologous to many other APs (Table 3). Surprisingly, no potential N-glycosylation site was predicted in the hPAG-L/pep precursor. In addition, in silico analyses permitted the identification of the molecular mass of the hPAG-L/pep polypeptide precursor (41.993 kDa) and its electrostatic property (pI 3.93).                pep A and very homologous to many other APs (Table 3). Surprisingly, no potential N-glycosylation site was predicted in the hPAG-L/pep precursor. In addition, in silico analyses permitted the identification of the molecular mass of the hPAG-L/pep polypeptide precursor (41.993 kDa) and its electrostatic property (pI 3.93).

Sequence Length (bp)
Gene Segment hPAG-L/pep bPAG1 bPAG2 pPAG2 CfPAG-L Two coded D residues within both domains creating a catalytic cleft (a feature of the AP members) were localized within exons 3 and 7 of the hPAG-L/pep ( Figure 2). All exon-intron junctions with 5 donor and 3 acceptor sites were identified ( Table 5). The sequences between each of the intron-exon junctions were determined and conformed to the standard gt-ag rule for 5 donor and 3 acceptor sites. The de novo identified hPAG-L gene is composed of: 24.7% A; 27.4% C; 25.6% G; 22.3% T and 53.0% G+C (Geneious ® 8.1.7.). Table 5. Characteristics of exon-intron junctions within the hPAG-L/pep gene.

Identification of Cellular hPAG-L/pep Localization
Heterologous dF-IHC with anti-pPAG-P ( Figure 4A-H) and anti-Rec pPAG2 polyclonals ( Figure  5A-H) allowed localization of the hPAG-L/pep proteins within term placental cells. Generally, stronger immuno-positive signals of the hPAG-L/pep (green) were identified with anti-Rec pPAG2 than anti-pPAG-P polyclonals. The strongest immune-positive hPAG-L/pep signals were observed within the analyzed regions of chorionic villi (CV) and villous core (VC), especially within the syncytiotrophoblast nuclei (red) covering the surface of the terminal villous tree (arrowheads in   Figures 4E and 5G) within VC surrounding fetal capillaries (FC) close to the intervillous space (IS). No hPAG-L/pep signals were immuno-detected within placental septa (PS), which were infiltrated by decidual cells (Figure 5D) or within the maternally-originated stratum basale (SB; Figure 5B). A negative control did not generate any signal (Figures 4H and 5H).  H and 5E-H). The hPAG-L/pep signals were also detectable ( Figures 4E and 5G) within VC surrounding fetal capillaries (FC) close to the intervillous space (IS). No hPAG-L/pep signals were immuno-detected within placental septa (PS), which were infiltrated by decidual cells (Figure 5D) or within the maternally-originated stratum basale (SB; Figure 5B). A negative control did not generate any signal (Figures 4H and 5H).

Discussion
This study presents pioneering data concerning identification of the human placental PAG-L/pep cDNA (Acc. No. KX856064) and protein (60 kDa). In addition, the exonic-intronic structure of entire hPAG-L/pep gene has been identified (Acc. No. KX533473). Direct comparison of our data is impossible because similar data are not available. Therefore, our data can only be compared to animal species in which the PAGs have already been identified.

Discussion
This study presents pioneering data concerning identification of the human placental PAG-L/pep cDNA (Acc. No. KX856064) and protein (60 kDa). In addition, the exonic-intronic structure of entire hPAG-L/pep gene has been identified (Acc. No. KX533473). Direct comparison of our data is impossible because similar data are not available. Therefore, our data can only be compared to animal species in which the PAGs have already been identified.

Identification of hPAG-L Transcript
The novel 1364 bp hPAG-L/pep cDNA, identified with term placental mRNA, allowed identification of nucleotide homology (at least 56.5%) with other AP members. Previously, nucleotide sequences of the PAG cDNAs have only been identified in cattle, sheep, pig, goat, horse, zebra, white-tailed deer, water buffalo, American bison, wapiti, giraffe and the Eurasian beaver [1]. Such a limited number of cloned cDNAs resulted from difficulties during the wild eutherian placenta harvesting required for high-quality RNA and cDNA library. The numbers of the PAG-L cDNAs vary between species and are multiple in cattle, sheep, goats and pigs, while a single PAG-L cDNA has been identified in the horse, zebra, mouse, cat and beaver [1,2,20]. A possible explanation of this fact might be various placenta types, different requirements for the development of the fetus and special environmental needs in different species.
The identified cDNA allowed an encoded 388 aa hPAG-L/pep polypeptide precursor (Figure 1; Tables 1-3) to be characterized, which contains 15 aa-signal peptide, 47 aa-blocking peptide and 326 aa-mature polypeptide, making it similar to other PAG precursors. Among the identified PAGs, the length of the 15 aa-signal peptides, as well as the 33-38 aa-blocking pro-pieces are very conservative in various species [21,[32][33][34]. Different PAG precursors [1] vary in their entire length (375-389 aa), molecular mass (30-90 kDa) and electrostatic properties (4.0-9.08 pI). Our in silico analyses of the hPAG-L/pep precursor (41.977 kDa; Ip = 3.93 pH) also contributed to the enlargement of the diversity and confirmed membership in the PAG family. The identified hPAG-L/pep precursor was also similar to peps that are composed of 15-16 aa signal peptides, 42-46 aa activation segments and 321-332 aa of mature proteins [5,35].
We are aware that the identified placental hPAG-L/pep transcript is identical to another human AP (pep A); thus, our hPAG-L should be classified as an catalytic active form. Presently, we can expect that the hPAG-L/pep and pep A are similarly activated by degradation of placental or gastric polypeptide precursors due to an identical blocking peptide sequence ( Table 2). Such expectation may confirm equal/similar aa homology (56.4-67.6%) of hPAG-L/pep with peps C and F as well as other catalytically active PAGs (fPAG, pPAG2, CfPAG, ePAG and bPAG2).
High N-glycodiversity is very common in the PAGs [1], but in pepsinogens it occurs occasionally and no more than two N-glycosylation sites are generally present [36,37]. Surprisingly, within the aa sequence of the hPAG-L/pep precursor, no potential sites of N-glycosylation were predicted. Thus, it confirms that the hPAG-L/pep precursor is different from pep A.
Due to the conserved sequences of two aspartic acids (D) located within two domains (NH 2 -and COOH-terminal), creating the substrate binding cleft, the hPAG-L/pep precursor was classified as a catalytically active AP member, similar to human pep A ( Figure 1; Table 3). The PAG-Ls identified in the mouse [38], horse, zebra, cat [21,39] and beaver are also classified as active APs [20]. Within the diversified PAG family in species with multiple PAG members, either potentially active as well as potentially inactive forms exist [1,2,32,40,41]. Multiple aa substitutions within both domains contribute to a loss of catalytic activity of many PAGs [1,3,33,42].
Presumably, some similarities of the PAG-L family in the human and some Rodentia species (beaver or mouse) resembled discoid placenta type and potentially comparable requirements of developing fetuses in those taxa.
Most PAG/PAG-L cDNAs share relatively a higher nucleotide homology with each other than to pepsinogens [2]. Interestingly, the hPAG-L/pep shares higher homology with pepsinogens than other PAGs, similar to CfPAG-L and ePAG [20,21]. Pepsins were initially considered to be restricted to the stomach of many vertebrates [5]; however, in lower vertebrates, progastricsin (also known as pep C) was also found in the esophageal mucosa of the frog [43] as well as larval pepsinogen cDNA in whole bodies of the pufferfish [44]. Phylogenetically, peps F and PAGs belong to the same cluster [2,5]. The high (99%) homology of the identified hPAG-L/pep cDNA to the human pep A indicates that both genes are very similar but are two related AP genes with completely different expression.

Identification of hPAG-L Exonic-Intronic Structure
Because data concerning PAGs in human genome are not available, the presently obtained results can only be compared to studies performed in some animal species. The entire identified hPAG-L/pep gene sequence (9330 bp; Acc. No. KX533473) comprises a structure of nine exons and eight introns ( Figure 2). The location of the two D residues, within exons 3 and 7 of hPAG-L/pep, is similar to other PAGs, which is specific for the catalytic cleft of all APs. To date, the entire structure of the PAG genes has only been identified in three species (cattle, pig and beaver). The length of the hPAG-L/pep exons (1-9) is greatly similar to exon lengths of bPAG1, bPAG2, pPAG2 and CfPAG (Table 4) or even the same (especially exons: 3,4,8), whereas other exons vary. The gDNA alignments of the hPAG-L/pep exons with bPAG1, pPAG2 and CfPAG-L revealed high homology in the range of 52.1-78.6% (Table 6). However, the length of the hPAG-L/pep introns (A-H) varied from previously discovered PAGs, except in the length of intron G for hPAG/pep and bPAG1 or bPAG2, as well as their total lengths (Table 4). Furthermore, the pairwise sequence alignment of intronic regions in the aforementioned PAGs revealed generally lower homology (25.4-58.5%; Table 6).
Previously, Southern hybridization of gDNA (with selected restrictases) revealed a diversified number of the PAG-L genes in some eutherian species, e.g., the elk, yak, wildebeest, impala, several other antelopes [33], the pig, goat, horse, cow, sheep, deer and wild boar and bisons [1]. Southern blot of amplicons also revealed the PAG-L family in the alpaca, the dromedary and the Bactrian [45]. Sequence identification and comparison of cDNA and gDNA enabled defining the exonic-intronic boundaries for only four PAGs. So far, multiple bovine PAG cDNAs [33,42] have allowed the identification of the bPAG1 gene (8095 bp) as the first representative with an identified exon-intron structure, with an intron length ranging from 87 bp to 1.8 kbp [17]. Identification of the pPAG1 and pPAG2 cDNAs [32] has also led to identification of the pPAG2 gene structure [19]. The pPAG2 belongs to the pPAG2-L subfamily together with other members: pPAG4, 6, 8, 10 and they constitute catalytically active APs. However, potentially inactive members of the pPAG1-L subfamily, pPAG3 and 5, have also been identified [40]. The pPAG2 structure [19] encompasses nine exons (99-200 bp) and eight introns (A-H; 85-1.8 kbp). The entire length of the pPAG2 with a promoter region is equal to 8755 bp [19,46]. Recently, CfPAG-L (Acc. No. KX377932) was discovered in the Eurasian beaver (7657 bp) as an AP member containing nine exons and eight introns [20]. The lengths of the hPAG-L/pep (56-200 bp), as well as CfPAG-L (59-200 bp) exons, are similar to the other known bPAG1, bPAG2 and pPAG2, although the length of the introns differ from previously identified PAG.
Since the results obtained in this study are consistent with the exon-intron structures (length and homology alignment) of four previously described PAGs and other APs, the identified hPAG-L/pep was assuredly classified as a new AP member. However, the high homology of the hPAG-L/pep to the pep A family in various species is also a novel finding. The multigenic AP family is widely distributed in various taxa and emerged from duplication or fusion of the paralogous progene [5,6]. In mammals, the major AP members are well-known pepsinogen genes classified as A, B, C and fetal forms known as pepsinogen F [35,47]. Complete gene structures have been determined, e.g., for human pep A [48], C [49] and prochymosin [50]. The structure is conserved among APs, including PAGs, pepsinogens, cathepsins D, E, and renin, suggesting evolution from a common ancestral gene [51].

Identification of hPAG-L Proteins
Western blotting (Figure 3), with anti-pPAG-P and anti-Rec pPAG2 polyclonals, identified a uniform cellular protein profile of native hPAG-L/pep isoform (60 kDa) in term singleton placentas. Similar data are unfeasible in humans. In animals, multiple heterogeneous secretory PAG isoforms, 43-70 and 45-85 kDa released by placental explants, have been found in the pig and cattle, respectively [34,52]. In the pig, gestation-stage dependent diversity of glycosylated forms of the pPAG proteins occurs, which contain an average of 9.66% of N-linked oligosaccharides [53]. In the bPAG family, oligosaccharide heterogeneity is caused by diversified tetra-antennary glycans [54]. In addition, three purified PAG isoforms (72, 74 and 76 kDa) secreted by the placenta of the American bison have been sequenced [55]. However, in the European bison, among two major groups (43-45 and 67-69 kDa) of immuno-detected secretory PAG isoforms [52], eleven novel diversified pregnancy-stage dependent (45-129 day post coitum-dpc) EbPAGs (50-71 kDa) have been sequenced [56]. Various PAG isoforms also exist in species of the Cervidae order: 33-55 kDa in the white-tailed deer [41], 39-62 kDa in the fallow deer [57], and dominant 55 kDa fraction-specific isoforms for different pregnancy stages (50-200 dpc) in the European moose [58]. It seems that such diversity of multiple PAGs originated from gene duplication during the evolution of different species.

Identification of Cellular hPAG-L Localization
Double-labeling heterologous immuno-detection revealed the strongest positive hPAG-L signals within the chorionic villi, localized especially within the syncytiotrophoblast cells (Figures 4 and 5). Similar data are unachievable in humans. Localization of the hPAG-L expression may be directly compared only with results for the beaver (a discoid-placenta type species) in which CfPAG-L signals are found either in regular or giant trophectodermal cells [59].
In other animals, cellular expression of the PAGs was previously mostly localized in embryo-originated chorionic cells in some species of the Artiodactyla order, with cotyledonary (bovine, bison, white-tailed deer, moose) or the diffuse (porcine, alpaca, camels) placenta types, as described below. In the pig, pPAG expression is restricted to diversified chorionic cell layers throughout (16-61 dpc) placenta development [60]. In ruminants, multi-nucleated, enlarged and multi-granulated trophectodermal cells expressing PAGs have been observed in the white-tail deer [1], while in placentomes of the European bison, the EbPAGs were localized in apical regions of cotyledonary villi folds [61]. Similarly, in camelids during alpaca pregnancy (150-347 dpc), Lama pacos-LpPAGs are present in the trophectoderm cell layer and within very rare giant cells [62]. In the term placenta of both camels, CdPAG (Camelus dromedarius-dromader) and CfPAG (Camelus ferrus-Bactrian) are present in the cytoplasm of the outer folded layer of the mononuclear trophectodermal cells, mostly at the apex of the placental folds [63]. In the moose (Alces alces), AaPAG-L signals are related to placentome growth (50-200 dpc) and are localized in the trophectodermal cells, especially within secretory granules [58]. Despite the morphological and developmental divergences of various placenta types, the localization of the hPAG-L/pep resembled chorionic expression previously determined in other mammals.
This study describes pioneering identification of novel aspartic proteinase named hPAG-L/pep in the genome (Acc. No. KX533473; 9330 bp), placental transcriptome (Acc. No. KX856064; 1364 nt mRNA) and proteome (60 kDa) of the human. The expression of the hPAG-L/pep in chorionic cells can influence the regulation of placental development. The identified placental glycoprotein presumably can be used as a novel biomarker for prenatal pregnancy diagnosis of embryo/fetus well-being by a noninvasive test based on concentration measurement in peripheral maternal blood, similar to β-hCG test in the human as well as various PAG tests in the ruminants. In addition, the identified cDNA (ORF) and 9-exonic and 8-intronic gDNA sequences provide a major pattern of SNPs/InDels required for a novel marker preparation profitable for genotyping and detection of various genomic disorders in embryo/fetus and mother, similar to our report on SNPs/InDels for the pig [46,64] and the European moose [58]

Ethics Statement and Collection of Samples
All clinical samples (placentas and blood) were collected at the Clinical Ward for Gynecology, Obstetrics and Oncological Gynecology at the Regional Specialist Hospital in Olsztyn, following informed written consent from the parturient women. The study protocol was approved by the Bioethics Committee of the Warmia-Mazury Medical Chamber (OIL.164/15/Bioet; 2 April 2015) in Olsztyn, Poland. Only healthy mothers, after uncomplicated single pregnancy and without diagnosed medical conditions were selected for this study. Term placentas (n = 2) were collected from the women who underwent scheduled Caesarean section. Whole blood samples from both men (n = 3) and women (n = 3) were also collected from jugular veins. Placental tissues were immediately preserved in liquid nitrogen and blood samples were placed on ice and transported directly to the laboratory. Samples of blood were centrifuged (3500× g) for 30 min at 4 • C, plasma was discarded and the buffy coat of the white cells, as well as placental tissues were stored at −70 • C until further analyses.

Total RNA Extraction
Total RNA was isolated from term placental fragments, using a QiagenRNeasy Kit in conjunction with the QiagenRNase-Free DNase Set (Qiagen, Hilden, Germany), according to the manufacturer's recommendations. RNA quality was evaluated via microfluidic electrophoresis (2100 Bioanalyzer; Agilent Technologies, Santa Clara, CA, USA). Only a high RNA integrity number (RIN > 8.0) of each sample was accepted for high throughput mRNA sequencing (RNA-seq).

High Throughput mRNA and Bioinformatics
Complementary DNA (cDNA) libraries were constructed using the protocol of TruSeq Stranded mRNA LT Sample Prep Kit (Illumina, San Diego, CA, USA) involving the following steps: RNA purification and fragmentation, synthesis of the first and the second strand of cDNA, 3 adenylation and adaptor ligation. After amplification and quantification (KAPA Library Quantification Kit, Illumina), the cDNA libraries were indexed, diluted and pooled in equimolar ratios.

Capillary Sequencing
Capillary sequencing was performed to confirm the coding sequence of the placental AP, identified by RNA-Seq. Briefly, total RNA (from the same samples that were used for RNA-Seq) was transcribed to cDNA in two-step RT-PCR using an Enhanced Avian HS RT-PCR Kit (Sigma-Aldrich, USA). The first strand cDNA was synthesized with dNTPs, and random hexamers were used as primers. PCR amplification of target cDNA templates to obtain hPAG-L amplicons was performed with specific primers (Table 7) designed by applying Geneious ® 8.1.7 software (Biomatters Ltd., Auckland, New Zealand) and Oligo Calc: Oligonucleotide Properties Calculator (http://www.basic.northwestern. edu/biotools/oligocalc.html), basing on PAG-L sequence identified using RNA-Seq. The obtained amplicons of examined hPAG-L, parallel to porcine PAG10 (pPAG10) cDNA-used as a positive control and negative control (without templates)-were separated in 1.5% agarose gels, along with a marker (100-3000 bp; Thermo Fisher Scientific, Waltham, MA, USA), UV-visualized using Midori Green Nucleic Acid Staining Solution (NIPPON Genetics Europe GmbH, Dueren, Germany) and archived (G:Box, SynGen, Sacramento, CA, USA). Gel-out purified hPAG-L amplicons were used as templates for capillary sequencing (3130 Genetic Analyzer, Applied Biosystems, Foster City, CA, USA) in both sense and antisense directions. Amplicon labeling was performed with the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems), under the following conditions: initial denaturation (at 96 • C for 1 min) and 30 cycles of amplification (96 • C/10 s, 50 • C/5 s, 60 • C/4 min). Each labeling mix (20 µL) was composed of 12 µL (5-10 ng) of amplicon template, 1.2 µL Ready Reaction Mix, 4 µL BigDye Terminator v1.1/3.1 Sequencing buffer (5×), 2 µL of each primer and 0.8 µL H 2 O. The labeled hPAG-L amplicons were purified with the BigDye X Terminator Purification Kit (Applied Biosystems) and separated in capillaries filled with POP-7™ polymer (Applied Biosystems). The obtained hPAG-L sequences were analyzed by Geneious ® 8.1.7. In addition, in silico analyses of the cDNAs were performed applying the following online tools: http://www.cbs.dtu.dk/services/SignalP/; http: //prosite.expasy.org; http://www.cbs.dtu.dk/services/NetNGlyc/.

Genomic Identification of the hAP/PAG-L Sequence
Genomic DNA (gDNA) templates (n = 6) were isolated from the leukocytes with the use of a commercially available kit (Sherlock AX, A&A Biotechnology, Gdynia, Poland). Only high quality gDNA templates (700 ng) were used for PCR amplifications of the hPAG-L gene fragments. In order to identify either initial or partial nucleotide sequence of the hAP/hPAG-L, the gDNA amplicons were produced with 19 pairs of specific homologous primers (Table 8), designed on the basis of the hPAG-L cDNA sequence originating from the aforementioned RNA-seq. JumpStart™ Taq ReadyMix™ (Sigma-Aldrich, St. Louis, MO, USA) was used for efficient PCR amplification, under the following conditions: initial activation (95 • C/2 min), followed by 40 following cycles: 95 • C/1 min for the denaturation of gDNA templates, 60 • C for primer annealing (1 min) and 72 • C/4.5 min for amplicon synthesis. The obtained hPAG-L gDNA amplicons were electrophoresed, gel-out purified, subjected to capillary sequencing and analyzed as described above (see Capillary sequencing). Identified cDNA and gDNA sequences of hPAGL-L have been deposited in GenBank (Accession nos: KX856064 and KX533473, respectively

Identification of the Exon-Intron Organization of the hPAG-L
To estimate the length of the introns and exons in the hPAG-L, the identified sequences were analyzed by NetGene2 v. 2.4 software (www.cbs.dtu.dk/services/NetGene2/) to predict a structure, based on multiple alignments (Geneious ® 8.1.7, www.geneious.com and BLAST, https://blast.ncbi. nlm.nih.gov/Blast.cgi) of the entire hPAG-L genomic sequence with the identified cDNA (see above).

Cellular Placental Protein Extraction
Cellular proteins were isolated as previously described for other species [58,65]. Briefly, the frozen human placental tissues (n = 2) were homogenized on ice and lysed by alkaline buffer (Total Protein Extraction Kit, Genoplast, Rokocin, Poland). The obtained protein homogenates of each placental sample were concentrated (3000 rpm/4 • C) by ultra-filtration in Centriprep-10 cartridges (>MWCO 10 kDa; Amicon, Billerica, MA, USA) until a 0.5 mL of final volume was received. Total protein concentration was determined by the standard Bradford procedure. Concentrated placental proteins (10 µg/line) were separated by denaturing polyacrylamide electrophoresis (SDS-PAGE, 12.5% gels), parallel to porcine placental proteins (positive control for Western blotting), endometrial proteins of cyclic pigs (negative control) and a molecular marker (10-250 kDa; Fermentas, Waltham, MA, USA). Electrophoresed proteins were stained with Coomassie Brilliant Blue (CBB) to identify total human placental protein profiles.

Conclusions
Our discerning and comprehensive studies provide novel data identifying the placental hPAG-L/pep transcript, gene structure and chorionic protein in humans. Our pioneering data extend the present knowledge of the human genome, placental transcriptome and proteome, which may contribute to establishing a new diagnostic tool for examination of various disturbances during human pregnancy, with growing interest from both scientific and clinical perspectives.