Introduction

The ovary is a dynamic organ in respect to rapid and extensive degrees of tissue development and remodeling. Each follicle destined for ovulation must execute the development and differentiation of its components: the innermost oocyte and its nurse-cells: granulosa cells and theca cells [36]. The changes in ovary involving dramatic changes in granulosa cell function are sequential and are dictated by specific, regulated responses to gonadotropins, steroids, and growth factors [21, 39, 42].

Insulin-like growth factor-I (IGF-I) system is an important factor in follicular development. IGFBPs increase IGFs half life and maintain a stable pool of IGFs in many compartments of the organisms. However, the affinity of IGFBPs for IGF-I and IGF-II can be modulated by changes in local synthesis as well as their post translational modifications via proteolysis. Such proteolysis has been described in the serum of female rodents and humans during pregnancy after nutritional fasting, during severe illness or following extensive trauma and in culture medium of osteoblastic cells, granulosa cells, and tumor cells [30]. The protease that degrades IGFBP-3, 4, and 5 found in sheep follicular fluid (FF) has been identified a serine metalloprotease. The IGFBP protease in bovine and equine FF also has characteristics of a serine metalloprotease and was hypothesized to be pregnancy-associated plasma protein-A (PAPP-A). Such decrease in the affinity of IGFBPs for IGFs by increasing bioavailability of IGFs can lead to a potentiation rather than an inhibition of action of the ligands.

PAPP-A was originally isolated in 1974, as one of four proteins of placental origin found circulating at high concentrations (at a level of 50 mg/l) in pregnant women [29]. Lawrence and coworkers [28] isolated an IGF-dependent IGFBP-4 protease activity from human fibroblast conditioned media and identified it as PAPP-A. Subsequently, PAPP-A was also shown to be expressed by osteoblasts, vascular smooth muscle cells, and ovarian granulosa cells, as well as by placental trophoblasts in culture [5, 11, 14, 20]. Clearly, PAPP-A could no longer be considered just “pregnancy associated.” In vitro studies and the characterization of a PAPP-A knock-out mouse model suggested this IGFBP protease as an important regulator of local IGF bioavailability [13]. Placenta is the major site of PAPP-A synthesis; however, relatively high level of PAPP-A was found in ovarian FF as IGFBP-4 protease. Mature PAPP-A is secreted as a homodimer of approximately 400 kDa composed of two 200 kDa disulfide-bound subunits [37]. The observation that FF from healthy follicles had considerably higher levels of PAPP-A activity than did atretic follicles led to the hypothesis that PAPP-A expression may play a role in the dynamics of dominant follicle development. Experimentally, this concept was supported by in situ hybridization studies in which PAPP-A mRNA was identified in granulosa cells of healthy, but not atretic follicles [33]. PAPP-A is present in the FF of the future dominant follicles and the follicular selection may be result of series of changes beginning with the acquisition of PAPP-A, decrease in IGFBP-4 and -5 and increase in free IGF, which synergizes with FSH to increase estradiol production [17].

Amino acid sequences of PAPP-A orthologues show a remarkable conservation, which suggests an essential function of PAPP-A in a variety of mammalians. Amino acid sequence identity between human and mouse or human and rat PAPP-A approaches 91 % [45]. In addition, incomplete sequences from other mammals reveal very few substitutions when compared to human PAPP-A [35]. PAPP-A is also present in non-mammalian vertebrates viz. Fugu rubripes, Tetraodon nigroviridis (60 and 50 % conservation, respectively, with human PAPP-A), Danio rerio (zebrafish), and Xenopus laevis. In contrast, PAPP-A is not found in DNA sequence databases from invertebrates such as Drosophila melanogaster or Caenorhabditis elegans [8]. PAPP-A belongs to the metzincin superfamily of metalloproteinases [7] and now considered as the first member of a new metzincin family, the pappalysins. Mature PAPP-A monomer contains 1,547 amino acids representing the C-terminal part of the proform of PAPP-A [26]. Two structural motifs that define the unifying features in the primary structures of members of the metzincin superfamily of metalloproteinases, which include the matrixins (matrix metalloproteinases, MMPs), serralysins, astacins, and adamalysins (ADAMs), can be identified in PAPP-A [9, 48]. The proteolytic domain contains an elongated zinc-binding consensus sequence H E XXHXXGXXH and a Met-turn motif (M). In addition, three calcium binding Lin12-Notch (LNR) repeats are found in PAPP-A. The molecule of PAPP-A is distinct in several regards from other four members of metzincin superfamily of metalloproteinases. A putative Zn2+ binding site, which has been found in PAPP-A molecule, is very similar to that containing the active site Zn2+ of the matrix metalloproteinases and quite similar to these of other metalloproteinases [26]. It is interesting that in PAPP-A the linear distance between the zinc motif and the conserved Met residue is 63 amino acids, whereas in other metzincins this distance is shorter between 7 and 44 residues [48]. The proteolytic domain of ~350 amino acids is thought to be responsible for the ability of PAPP-A to cleave IGFBPs.

In view of the importance of PAPP-A and no information on expression in buffalo ovary, the present investigation was envisaged.

Materials and Methods

Collection of Ovaries and Isolation of Whole Follicles

Buffalo ovaries were obtained from nearby abattoirs and processed further as per our previous report [50]. Non-atretic follicles were grouped as small (<5 mm diameter), medium (5–9 mm diameter), and large (>9 mm diameter) preovulatory and pooled according to their size representing different stages of follicular development [47]. Postovulatory structures, mid-luteal phase corpora lutea (CL) and corpus albicans (CA) were isolated and classified based on morphological appearance including size, extent of vascularity, presence and absence of central blood, and color and overall appearance of ovary.

Granulosa Cell Isolation and Culture

Granulosa cells were isolated from follicles as described by Malhotra et al. [32]. The granulosa cells in FF were aspirated from small, medium, and large antral follicles using 18 G needle and syringe. The FF collected in 2.0 ml eppendorf tube (for RNA isolation) or 15 ml sterile centrifuge tubes (for cell culture) were centrifuged at 2,000 rpm for 5 min to get pellet of cells for RNA isolation and cell culture.

Trypan blue staining was used for assessment of granulosa cell viability during follicle isolation and culture in vitro. Cell viability was estimated at 30–40 % by counting in a hemacytometer. Approximately 8 × 106 viable granulosa cells were plated in a six-well culture plate (Costar Corporation, Cambridge, USA) using bicarbonate-buffered MEM with antibiotics (Plain media, pH 7.2) and 3 % FBS (Sigma Chemical Co., USA, # F7524) to permit cell anchorage. Cells were allowed to attach in culture dish for 24 h at 37 °C at 5 % CO2. Monolayer was maintained in about 2.0 ml per well serum containing media in a water jacketed CO2 incubator (Nuaire, USA) for about 1 day. After 24 h of cell culturing in FBS, cells were observed under inverted microscope (Olympus CK2) for cell attachment and viability. FBS containing MEM was removed. The cells were washed with plain media. Different concentrations (5 and 100 ng/ml) of ovine FSH (NIDDK-oFSH-17) were prepared in plain media. The cells were treated with FSH in a set of three wells for each concentration. For control, 2.0 ml of plain media was added and the plates were kept for incubation at 5 % CO2 for a minimum period of 24 h. After 24 h of treatment, cells were observed under microscope for growth and viability. Suspended media was removed and cells were further suspended in 2.0 ml of TRI reagent.

Isolation of Total RNA

The total RNA was isolated from granulosa cells obtained from FF as well as culture, using TRI reagent (Molecular Research Centre, Inc # TR-118). Granulosa cells grown in monolayer culture were lysed directly in culture dish with 1 ml of TRI reagent/10 cm2 of culture dish after removal of spent media. Luteal tissues (100 mg) were immersed in 1 ml of TRI reagent and homogenized. Similarly, intact 7–8 small, 3–4 medium, and 1–2 large size follicles were homogenized separately in 1 ml of TRI reagent and processed further for RNA isolation. Total RNA with the ratio (A 260/280) ranging from 1.8 to 2.0 were accepted for subsequent procedures. RNA was stored at −70 °C. Integrity of RNA preparation was evaluated by agarose (1 %) gel electrophoresis, after its denaturation with glyoxal and dimethylsulfoxide [43]. The RNA was treated with DNA-free™ kit (Ambion Inc., USA, # 1906) to remove contaminating DNA.

Primer Design and Selection

Primer sequences were designed online using the software Primer 3 Output Program (http://www.genome.wi.mit.edu). PCR primers were selected in such a way that the 5′- and 3′- primers span different exons. Primer sequences information and their expected amplification products are provided in Table 1. All oligonucleotides were commercially synthesized from Operon Biotechnologies GmbH, Cologne, Germany.

Table 1 Primers used for amplification in PCR
Full size table

Reverse Transcriptase-Polymerase Chain Reaction

Reverse transcriptase-polymerase chain reaction (RT-PCR) was performed using M-MuLV RT PCR Kit (Bangalore Genei Pvt. Ltd., Bangalore, India; # KT74) in two steps viz. first strand cDNA synthesis (RT) reaction and PCR amplification. For first strand cDNA synthesis, 1 μg of RNA sample was taken in a sterile DNA/RNA/DNase/RNase/Pyrogen and PCR inhibitor-free microfuge tube (Labware Scientific Inc., # PCR-02N) and diluted to 9.5 μl with sterile water (sH2O). Then 0.5 μl of random hexamers was added. The solution was incubated at 65 °C for 10 min in a thermocycler (Whatman Biometra, Germany) and reagents were added in the order: 1 μl RNase inhibitor, 1 μl DTT (0.1 M), 4 μl 5× reverse transcriptase buffer, 1 μl dNTP mix (30 mM), 1 μl M-MuLV RT enzyme, and sH2O to bring the reaction volume to 20 μl. Furthermore, the reaction was set up at 37 °C for 1 h, 95 °C for 5 min, and stopped at 4 °C. For amplification of cDNA, the following PCR components were added to a DNA/RNA/DNase/RNase/Pyrogen and PCR inhibitor-free microfuge tube: 2 μl cDNA product, 5 μl 10× assay buffer for Taq polymerase, 1 μl dNTP mix (30 mM), 1 μl upstream primer (10 μM), 1 μl downstream primer (10 μM), 1 μl Taq polymerase enzyme (1U/μl), and sH2O to bring the reaction volume to 50 μl. PCR reactions were performed in the thermocycler with cycle conditions: 94 °C for 2 min, followed by 30 cycles of 94 °C for 1 min, annealing temperature (60 °C) for 1 min and 72 °C for 1 min. A 4-min final extension at 72 °C was performed following last cycle and reaction was stopped at 4 °C.

PCR conditions for the analysis of gene expression by semi-quantitative RT-PCR have been optimized as described earlier [32]. The experiment was repeated thrice with different RNA samples.

Analysis of PCR Products

PCR product (7 μl) mixed with 1.4 μl of 6× gel loading dye was loaded on agarose (2 %) gel and run in mini-submarine gel electrophoresis system with TBE buffer (1×). The gel was stained with ethidium bromide (0.5 μg/ml) and the amplified products were visualized on UV-transilluminator. Intensity of individual bands was measured by densitometry using GelQuant.NET software (from biochemlabsolutions.com).

Gel Extraction and Sequencing of Purified PCR Products

Gel extraction of PCR products was done by Wizard® SV Gel and PCR Clean-Up System (Promega, USA). After the gel extraction, the PCR products were custom sequenced (Sigma Genosys Lab., Bangalore).

Statistical Analysis

Densitometric analysis results were expressed as ratio obtained by dividing the band intensity of target gene (PAPP-A) with intensity of control gene (G3PDH or β-actin). The relative abundance of mRNA expression between different stages of ovarian cycle was done by one way analysis of variance (ANOVA) for three biological replicates and Microsoft Office Excel 2007. The values were compared at 5 and 1 % level of probability. Differences were considered significant at P < 0.05 and P < 0.01.

In Silico Analysis of Nucleotide and Deduced Amino Acid Sequences

Additional mammalian sequences of PAPP-A were retrieved from NCBI GenBank as follows: cattle (XM_613511), sheep (AF421140), horse (XM_001487931), mouse (NM_021362), rat (NM_001107939), human (X68280), rhesus monkey (XM_001095160), chimpanzee (XR_024483), Pig (XM_001926753), rabbit (XM_002720482), and giant panda (XM_002915787). Sequences were edited and analyzed using the BLAST and Clustal W programs. Multiple sequence alignments were produced by CLC Sequence Viewer 6.5.1. Secondary structure was predicted using PSIPREDv3.0 of the PSIPRED protein structure prediction web server at UCL department of computer science [10]. Amino acid or peptide sequences were deduced from their respective cDNA sequences and further multiple sequence alignments were done by MEGA 5.0 program [49]. Molecular phylogenetic analysis was performed from the data provided by the multiple sequence alignment patterns of amino acids following the Maximum-Likelihood method based on 1,000 bootstrap replications and Jones–Taylor–Thornton (JTT) matrix-based model [24] by the MEGA 5.0 program. Phosphorylation prediction was done by NetPhos 2.0 server of CBS, Technical University of Denmark [6]. CSS Palm 2.0 online server was used for finding potential palmitoylation sites [41]. Similarly, to predict glycation of ε amino groups of lysines in the deduced peptide sequence, NetGlycate 1.0 server [23] and to predict the Leucine-rich nuclear export signals (NES), NetNES 1.1 server [27] of CBS, Technical University of Denmark was used. DiANNA 1.1 web server was used for finding cysteines and potential disulfide bonds [16].

Results

Expression of PAPP-A mRNA in Buffalo Ovary

The relative expression pattern of PAPP-A mRNA from whole follicles of different sizes, corpus luteum, and corpus albicans is shown in Fig. 1. Moderate PAPP-A mRNA expression was observed in small and large whole follicles as represented in lanes 2 and 4 of Fig. 1a, respectively. In medium sized follicles, PAPP-A mRNA expression was found to be down regulated (lane 3). The expression pattern of PAPP-A gene was found to be significantly (P < 0.05) high in corpus luteum (lane 5). Corpus albicans, on the other hand, showed less expression of PAPP-A mRNA. Similar expression pattern of PAPP-A mRNA was observed when glyceraldehyde 3-phosphate dehydrogenase was used as control gene along with PAPP-A (300 bp, different set of PAPP-A primers) (Fig. 1b). Figure 1c represents the quantitative analysis of PAPP-A mRNA expression.

Fig. 1
figure 1

A–C Expression of PAPP-A in Bubalus bubalus Ovary: PCR was carried out in 1.5 mM MgCl2 concentration at annealing temperature of 60 °C for 30 cycles without (A) or with (B) control gene (G3PDH) as shown in the figures. Lane 1 represents 100 bp ladder. The control gene (G3PDH, 367 bp) bands showed consistent expression. Moderate PAPP-A expression was observed in small (S/WF Sm) whole follicles (lane 2) and large (L/WF La) whole follicles (lane 4). In medium (M/WF Md) sized follicles, PAPP-A mRNA expression was down regulated (lane 3). Lane 5 represents highly up regulated PAPP-A mRNA expression in corpus luteum (CL). Lane 6 represents low expression of PAPP-A mRNA in corpus albicans (CA)

Full size image

Expression of PAPP-A mRNA in granulosa cells isolated from different sized follicles has been shown in Fig. 2a with PAPP-A primers (first design) and Fig. 2b with PAPP-A primers (second design). In both set of experiment, the expression pattern of PAPP-A gene expression was found to be consistent (Fig. 2c).

Fig. 2
figure 2

A–C Expression of PAPP-A mRNA in granulosa cells isolated from different follicles. PCR was carried out in 1.5 mM MgCl2 concentration at annealing temperature of 60 °C for 30 cycles with control gene (β-actin, 243 bp) as shown in the figures. Lane 1 represents 100 bp DNA ladder. Lanes 2, 3, and 4 represent the moderate expression of PAPP-A mRNA (341 and 300 bp with different set of primers) in granulosa cells of small (S/GC Sm), medium (M/GC Md), and large (L/GC La) follicles, respectively. A gradual decrease, though non-significant, in transcript levels was seen from small to large follicles

Full size image

Effect of FSH on PAPP-A mRNA Expression

To assess whether PAPP-A mRNA are hormonally regulated, primary culture of buffalo granulosa cells were treated with ovine FSH (NIDDK-oFSH-17) at concentrations of 5 and 100 ng/ml for 24 h and RNA were isolated as described earlier. As shown in Fig. 3a, b, PAPP-A mRNA expression was found to be enhanced by both the concentration of FSH (5 ng/ml, lane 3; 100 ng/ml, lane 4), as compared to control (lane 2). The stimulatory effect of FSH on PAPP-A expression has been found to be significant (P < 0.05).

Fig. 3
figure 3

A, B Effect of FSH treatment on PAPP-A gene expression in cultured granulosa cells. A Granulosa cells treated with different concentrations of FSH (5 ng/ml, lane 3; 100 ng/ml, lane 4) for 24 h showed an increased expression of PAPP-A mRNA as compared to control (c) (lane 2). Lane 1 represents 100 bp DNA ladder. B Representative bar diagram showing the effect of FSH treatment on PAPP-A gene expression in cultured granulosa cells. Each value is the mean ± standard error of three experiments

Full size image

Buffalo (Bubalus bubalis) PAPP-A Sequence

The primers set designed could amplify a maximum of 341 bp PAPP-A fragment. Custom sequencing of purified PCR products/amplicons confirmed this. The sequence was submitted to NCBI GenBank with the accession number (DQ079591). A BLAST search for this sequence in NCBI confirmed its relatedness/similarity with other mammalian and vertebrate PAPP-A sequences. Sequences of buffalo partial cDNA of PAPP-A, the deduced amino acid sequence, and potential phosphorylation sites have been shown in Fig. 4. Analysis of the obtained sequence showed that the 341 bp transcript originates partly from exon 4 (position 1–106) to exon 5 (position 107–299) and exon 6 in part (position 300–341). ORF of Buffalo PAPP-A cDNA, starting with number 1 on the direct strand, extends from base 1 to base 339 and encodes 113 amino acid peptide of 12.61 kDa molecular weight corresponding to part of proteolytic domain and central domain in the human PAPP-A protein. A total of 6 cysteines exist with a possibility of cysteines at positions 8, 18, 39, and 53 to form the disulfide bond. Phosphorylation sites predicted are Serine (position 33 and 104- Casein kinase II phosphorylation site, 65-protein kinase C phosphorylation site), Threonine (positions 26 and 81), and Tryptophan (position 28). PAPP-A is most likely not O-glycosylated, as N-acetyl galactosamine was found to be absent from the isolated human PAPP-A subunit [38]. In silico analysis of 113 amino acid bubaline PAPP-A also did not reveal any potential O-glycosylation site. As shown in Fig. 6, 15th amino acid position (NDTC) may be the potential site for N-glycosylation with Asn-Xaa-Ser/Thr pattern. Potential sites for glycation of ε amino groups of lysines are at amino acid positions 4 and 67. Similarly, positions 8 and 109 in the 113 amino acid bubaline PAPP-A sequence were found to be predicted as potential palmitoylation locations. Prediction of secondary structure by PSIPRED has revealed two α-helices and two β-strands within the deduced 113 amino acid residues (Fig. 6). Sequence analysis using NetNES 1.1 server also revealed that buffalo PAPP-A sequence contains a putative leucine-rich NES as MHCYLDL residues with two leucines and well within the characteristic LXXXLXL pattern, where L may be any of L, I, V, F, and M amino acids.

Fig. 4
figure 4

cDNA Nucleotide sequence and corresponding deduced amino acid sequence of Bubalus bubalis PAPP-A: Sequence of amplified partial PAPP-A cDNA (341 bp) (GenBank accession no. DQ079591) with deduced 113 amino acid sequence is shown here. The putative phosphorylation sites are Serine (position 33 and 104- Casein kinase II phosphorylation site, 65-protein kinase C phosphorylation site), Threonine (position 26 and 81), and Tryptophan (position 28). The Serine residues are shown as circles, Threonine as solid rectangles, and tryptophan as dotted rectangles. The position of * indicates the end of open reading frame

Full size image

Comparison with Other Reported PAPP-A Sequences of Different Vertebrates

As shown in Fig. 5, comparison of sequence of amplified partial PAPP-A cDNA (341 bp; GenBank accession no. DQ079591) from the buffalo ovary, with corresponding reported cDNA sequences of the same regions of cattle (XM_613511), sheep (AF421140), and horse (XM_001487931) have revealed a homology of 96, 94, and 93 %, respectively. Pig (XM_001926753), rabbit (XM_002720482), and giant panda (XM_002915787) nucleotide sequences showed 90 % identities. Mouse (NM_021362), rat (NM_001107939), human (X68280), rhesus monkey (XM_001095160), and chimpanzee nucleotide sequences illustrated 87–89 % homologies. Similarly, the deduced amino acid sequence (AAY86495.1) from buffalo PAPP-A partial cDNA sequence revealed a homology of 97, 96, 94, 93, and 88 % with cattle, sheep, human, rat, and horse sequences, respectively (Fig. 6). Pig, rabbit, mouse, chimpanzee, giant panda, rhesus monkey,and white cheeked gabbon peptide sequences showed 95 % homology. The buffalo peptide sequence also revealed 91, 90, 86, 85, 83, 79, 77, 70, and 65 % homology with corresponding sequences of Sumatran orangutan, common marmoset, gray short tailed opossum, chicken, zebra finch, Carolina anole, western clawed frog, Arctic char fish, and Zebra fish, respectively. In comparing the homology at nucleotide level, buffalo PAPP-A sequence had the highest homology of 96 % for bovine sequences and lowest homology of 87 % was for rat and mouse sequences among mammals. At the amino acid sequence level, the buffalo PAPP-A sequence had 97 % homology with cattle (highest) and 86 % with gray short tailed opossum (lowest) among mammals. Zebra fish showed only 65 % homology.

Fig. 5
figure 5

Alignment of nucleotide sequences of the Bubalus bubalis PAPP-A with those of other species: the sequences for alignment were taken from 12 mammalian PAPP-A sequences [cattle (XM_613511), sheep (AF421140), horse (XM_001487931), pig (XM_001926753), rabbit (XM_002720482), giant panda (XM_002915787), mouse (NM_021362), rat (NM_001107939), human (X68280), rhesus monkey (XM_001095160), and chimpanzee (XR_024483)] including Indian water buffalo (DQ079591). Dots indicate the identical nucleotides as Indian water buffalo PAPP-A sequence and dash (–) missing nucleotides/gaps. The shaded nucleotides are the nucleotides different from the consensus sequence

Full size image
Fig. 6
figure 6

Comparison of putative amino acid sequence of the Bubalus bubalis PAPP-A with cattle, sheep, horse, mouse, rat, human, rhesus monkey, chimpanzee, pig, rabbit, and giant panda: The aligned sequences were compared with the 1,547 amino acid human PAPP-A showing the domain structure. Accession numbers of sequences compared is shown in the figure. The consensus sequence illustrates clearly the zinc motif with residues HEXXHXXGXXH as shaded region. Conserved NFM (NXM) amino acid residues of met-turn are also shaded. Other amino acid residues which show difference with consensus sequence are shown as shaded. The predicted secondary structure is also shown. Putative NES signals (LXXXLXL) are also shown in the figure well within the α-helical secondary structure. The position of filled square indicates the potential N-glycosylation site (Asn-Xaa-Ser/Thr), upward arrow indicates putative glycation sites, and filled triangle indicates predicted palmitoylation sites

Full size image

Comparison and alignment of deduced amino acid sequences revealed a consensus sequence characteristic of PAPP-A (pappalysin-1). A unique amino acid sequence, NFM (N X M) was found to be conserved among different species and interestingly, this corresponds to conserved Met-turn of PAPP-A. The linear distance between putative elongated zinc-binding consensus sequence H E XXHXXGXXH of other species’ PAPP-A sequences and Met-turn of bubaline PAPP-A amino acid sequence was found to be 63, a characteristic of pappalysins of metzincin family of metalloproteases. The family-specific residue Asn (N) was also found exactly two residues before the Met (M) of the Met-turn. Cysteine–glutamine repeats at alignment positions 8–9 and 18–19 were found to be conserved across all the mammals compared. A putative leucine-rich NES signal with LXXXLXL pattern is also found to be conserved in all the mammals as shown in Fig. 6. Superimposing the NES signal sequence on the consensus sequence of PAPP-A together with its pattern of hydrophobic residues and charged or polar residues clearly reveals its amphipathic α-helical secondary structure. Putative NLSs in PAPP-A were also mapped using cNLS mapper [25] at lower score of 2/3 and bipartite NLS was predicted between amino acid positions 1 and 33, attributing speculation of shuttling of protein between nucleus and cytoplasm or localized to cytoplasm.

Molecular Phylogenetic Analysis

To visualize the phylogenetic relationship of the deduced amino acid sequence of buffalo PAPP-A gene with those of other sequences known in other mammalian species, we analyzed 15 sequences: 1 of buffalo, 3 of other ruminants (cattle and sheep), 2 of pig, 2 of horses, 1 of giant panda, 2 of rodents (rat and mouse), 3 of primates (human, chimpanzee, and rhesus monkey), and 1 of rabbit. Based on multiple sequence alignment of 113 amino acid buffalo PAPP-A sequence with other PAPP-A sequences of these selected species, the phylogenetic tree was constructed. The result is shown in Fig. 7. Closely related mammalian species were grouped in different clades or taxonomical orders. The phylogenetic analysis showed 4 clades: A, B, C, and D. Clade A includes three embranchments: Indian water buffalo/cattle embranchment, sheep embranchment, and pig embranchment—all being members of cetartiodactyla. The B clade includes the horse (perissodactyla) and giant panda (carnivora) embranchment. The C clade includes members of rodentia and the mouse/rat embranchment. The D clade of primates comprised the human/chimpanzee, rhesus monkey, and rabbit embranchments. Finally, there were distinct clades of cetartiodactyla, perissodactyla, carnivore, rodentia, and primates. These data indicate a greater phylogenetic affinity among Indian water buffalo, cattle, and sheep compared to that found with horse, carnivores, and primates. It is of interest to determine whether either of the clades Euungulata (Perissodactyla + Cetartiodactyla) or Zooamata (Perissodactyla + Carnivora + Pholidota) is valid [51]. The phylogenetic tree here, however, unexpectedly supported for the Zooamata clade comprising of perissodactyla and carnivora. All the eutherian orders make for monophyletic clades typically characterized by shared derived characteristics (synapomorphies).

Fig. 7
figure 7

Phylogenetic relationship of the peptide sequence of Indian water buffalo (Bubalus bubalis) PAPP-A with other 14 reported peptide/protein sequences of different mammals: Molecular phylogenetic analysis was performed from the data provided by the multiple sequence alignment patterns of amino acids. The sources of all the sequence data are represented in the phylogenetic tree. The phylogeny is an un-rooted tree recovered using the maximum-likelihood (ML) method based on 1,000 bootstrap replications and Jones–Taylor–Thornton (JTT) matrix-based model by the MEGA 5.0 program. Only bootstrap value greater than >0 % is shown on the branch. The clades are shown as A, B, C, and D

Full size image

Putative Localization of PAPP-A on Bubaline Chromosome

BLAST searches (CoGe BLAST and NCBI BLAST) using the 341 nucleotide buffalo PAPP-A sequence against all the available genome databases of animals related it to PAPP-A and as located in chromosomes 1 (Sus scrofa), 25 (Equus caballus), 9 (location 9q33.2; Homo sapiens), 4 (location 432.2 cM; Mus musculus), 5 (location 5q24; Rattus norvegicus), 11 (Canis lupus familiaris), 17 (Gallus gallus), 15 (Macaca mulatta), 9 (Pan troglodytes), and 8 (NC_007306.4; Bos taurus). This option was not available for Bubalus bubalis genome in any of the databases available. However, in construction of preliminary and first generation whole genome radiation hybrid map of Bubalus bubalis chromosomes using cattle markers it was revealed that Bos taurus chromosome 8 (BTA8) equals Bubalus bubalis chromosome 3q (BBU3q) [3, 4]. This gives a logical and probable location of PAPP-A gene in the Bubalus bubalis chromosome 3q (BBU3q).

Discussion

In this study, we have identified and characterized the transcript of PAPP-A gene in Indian water buffalo (Bubalus bubalis) ovary for the first time and the sequence analysis revealed that it share the characteristics of pappalysin-1 of metzincin family of metalloproteases. The spatial and temporal expression patterns, localization, and hormonal regulation of PAPP-A, in buffalo ovary was analyzed by sequence analysis and semi-quantitative RT-PCR which is a highly sensitive and specific method useful for the detection of rare transcripts and are still appropriate for many purposes [2, 12, 15, 19, 46]. β-actin [18] and G3PDH were used as internal controls. It is now known that changes in intra-follicular IGF system provide a critical “fine tuning” of follicular regulation and is important for selection of a follicle from a cohort of similar sized follicles. Follicular selection may be the result of a progressive series of changes beginning with acquisition of PAPP-A, with concomitant decrease in IGFBPs and increase in free IGF, which synergizes with FSH to increase estradiol production. Thus, in mammalian ovary, the expression of PAPP-A relates to steroidogenic activity and follicular selection [17].

In this study, we detected expression of PAPP-A in different sized antral follicles, postovulatory structures (corpus luteum and corpus albicans), and granulosa cells isolated from follicles of different sizes. Little or no differences in levels of PAPP-A mRNA expression were found in the small and large sized whole follicles and granulosa cells isolated from different sized follicles. Similar observations were found in granulosa cells isolated from small and large healthy bovine and porcine follicles [34]. A higher transcriptional level of PAPP-A mRNA with distinct differences was detected in ovarian corpus luteum. In fact, the expression of PAPP-A mRNA was found to be highly up regulated (Fig. 1a, b, Lane 5 and Fig. 1c), whereas expression of IGFBP-5 mRNA was found to be down regulated (unpublished) in corpus luteum. Above observations have shown that buffalo corpus luteum has a strong message for IGFBP-5. Furthermore, there was decrease in expression of mRNA encoding PAPP-A in corpus albicans (Fig. 1a, b, Lane 6 and Fig. 1c). The result implied that in buffalo corpus luteum, PAPP-A is highly expressed in early luteal phase, with a tendency to decrease thereafter. It has been found previously that IGF system played an important role especially for the development of the early CL by action on luteinization of granulosa-lutein cells and stimulation of oxytocin and progesterone production [44]. Increased level of IGFBPs as a result of decreased PAPP-A activity may lead to decreased levels of IGFs which, apart from other factors, may lead to decreased blood flow and progesterone level and increased apoptosis causing functional and structural luteolysis. In this study, a high expression of PAPP-A mRNA in the corpus luteum was found as observed for woman ovary by in situ hybridization [22]. Relatively low expression of PAPP-A mRNA has been observed in corpus albicans of buffalo ovary as compared to corpus luteum. These facts imply a role for PAPP-A as a marker of ovarian follicle selection and corpus luteum formation in Indian water buffalo. There could be two possibilities of lower transcriptional level of PAPP-A in medium sized follicles than that of large follicles. First, population of atretic follicles may be more in the pool of medium follicles. However, we have collected the follicles based on the criteria used by other workers detailed in “Materials and methods.” Furthermore, we have also quantified Bcl-2 mRNA (not shown here) expression in pool of granulosa cells isolated from different sized follicles with no change in Bcl-2 mRNA expression. Second, stage-specific PAPP-A expression or other cells like theca cells may be contributing to decreased transcriptional level of PAPP-A.

Regulatory mechanism underlying the synergism exerted by FSH in stimulating PAPP-A gene expression (Fig. 3a, b) in differentiating granulosa cells has been elucidated. Many studies have shown that growth factors like GH, insulin, and IGF-I [1] and hormones like FSH [40] and LH [31] play an important role in intraovarian regulation of follicular development in a species-specific manner. In this study, it was found that treatment of cultured buffalo granulosa cells with FSH caused a marked increase in PAPP-A mRNA expression. FSH-induced PAPP-A message is in accordance with the similar observation found in cultured rat granulosa cells [33]. This study confirmed the stimulatory effect of FSH on PAPP-A mRNA expression in cultured buffalo granulosa cells.

Nucleotide sequence was determined by custom sequencing and amino acid sequences were deduced. Partial Indian water buffalo PAPP-A cDNA contains a coding region of 341 bp, corresponds to 113 amino acid sequence and spans from exon 4 to 6. Buffalo PAPP-A nucleotide sequence was found closer to ruminants (cattle and sheep) compared to the other mammalian species. Cattle and sheep nucleotide sequences showed 96–97 % identity, whereas distant relatives like rat and mouse showed only 87 % identity with that of buffalo sequences. Deduced amino acid sequence from the 341 bp ORF was found to be 113 residues long. Buffalo PAPP-A amino acid sequence had a 97 % identity with cattle, 86 % with opossum, and only 65 % identity with Zebra fish. Several characteristics of PAPP-A (pappalysin) was evident on sequence analysis, comparison and alignment of nucleotide and deduced amino acid sequences. TblastN and blastP searches have unearthed the unique and characteristic NFM (NXM) amino acid residues corresponding to conserved Met-turn of PAPP-A and found to be conserved among different species. The family-specific residue Asn (N) was also found exactly two residues before the Met (M) of the Met-turn. The linear distance between putative elongated zinc-binding consensus sequence H E XXHXXGXXH of other species’ PAPP-A sequences and Met-turn of bubaline PAPP-A amino acid sequence was found to be 63, a characteristic of pappalysins of metzincin family of metalloproteases. Domain database (CDD) search results also implied that bubaline PAPP-A belongs to Zinc-metzincin superfamily (zinc-dependent metalloprotease) and pappalysin-like subfamily. Actual alignment was detected with superfamily member cd04275 (ZnMc-pappalysin like). A total of 6 cysteines were found and conserved Cysteine–glutamine repeats at alignment positions 8–9 and 18–19 may have some functional role. The Buffalo PAPP-A amino acid sequence was predicted to have two α-helices and two β-strands. A putative leucine-rich NES signal was also detected and its amphipathic α-helical secondary structure was revealed. Potential sites for N-glycosylation with Asn-Xaa-Ser/Thr pattern were detected at 6th and 15th amino acid position. Palmitoylation and glycation sites were also located. Positions 26, 33, 65, 81, and 104 are potential phosphorylation sites. The probable location of PAPP-A in buffalo chromosome may be traced to chromosome 3q (BBU3q). Phylogenetic analyses also demonstrated distinct clades of cetartiodactyla, perissodactyla, carnivore, rodentia, and primates. A greater phylogenetic affinity was observed among Indian water buffalo, cattle, and sheep compared to that found with horse, carnivores, and primates. It is of interest to determine whether either of the clades Euungulata (Perissodactyla + Cetartiodactyla) or Zooamata (Perissodactyla + Carnivora + Pholidota) is valid. This was also a supporting information for Zooamata clade.

In conclusion, cDNA encoding the pregnancy-associated plasma protein-A (PAPP-A) was amplified and sequenced from the Indian water buffalo (Bubalus bubalis) ovary. The 341 bp buffalo PAPP-A cDNA encodes for 113 amino acid peptide sequence. Sequence analyses have clearly revealed that bubaline PAPP-A belongs to Zinc-metzincin superfamily (zinc-dependent metalloprotease) and pappalysin-like subfamily. It also confirmed the stimulatory effect of FSH on PAPP-A mRNA expression. Further exploration of expression patterns and better understanding of “follicular microenvironment” may pave ways for strategies concerned with ovarian function, fertility, and productivity in females.