Prolactin receptor ( PRLR ) gene polymorphism in Chios , White Karaman and Awassi sheep breeds

The objective of the present study was to determine the polymorphism in the prolactine receptor (PRLR) gene in Chios, White Karaman and Awassi, which are native sheep breeds in Turkey. By means of PRLR gene sequence homology between sheep and humans, two primer pairs were designed for polymerase chain reaction (PCR) amplification within intron 1 and exon 10 of the PRLR gene in sheep. A total of 160 amplicons (99 for intron 1 and 61 for exon 10) were subjected to DNA sequence analysis. For intron 1, 6 different haplotypes were determined. For exon 10, 7 different haplotypes were obtained. Some variations determined for exon 10 (g.14A>T p.Q14L; g.160G>A p.D160N; g.166G>A p.E166K; g.167A>T p.E167V; g.176A>T p.H176L; g.206G>A p.S206N; g.208G>A p.G208R) led to changes in the amino acids, but no amino acid changes were determined in g.2A>T, g.81A>G, g.138A>G, g.186C>T, g.207T>C. It was noted in particular that White Karaman and Awassi were similar to each other in both PRLR exon 10 and intron 1 haplotypes, whereas the Chios breed had a different variation.


Introduction
Prolactin is an anterior pituitary peptide hormone involved in many different endocrine activities and is essential for reproductive performance.This action is mediated by its receptor.The prolactin receptor (PRLR) has been detected in various tissues including brain, ovary, placenta and uterus in several mammalian species (Shirota et al. 1990, Tzeng & Linzer 1997, Cassy et al. 1998).The prolactin receptor, encoded by the PRLR gene, is a member of the growth hormone/prolactin receptor gene family containing regions of identical sequences (Bole-Feysot et al. 1998).The prolactin and growth hormone receptors are homologous to receptors for members of the cytokine superfamily (Clevenger et al. 1998).Mice homozygous for null mutations in PRLR are sterile due to a failure of embryonic implantation, demonstrate irregular cycles, have reduced fertilization rates and defective embryonic development (Ormandy et al. 1997, Bole-Feysot et al. 1998, Baran et al. 2002, Grosdemouge et al. 2003), and show impaired maternal behaviour (Lucas et al. 1998).These characteristics make PRLR a strong candidate gene for reproductive traits (Kmiec et al. 2001, Kovaks et al. 2010).
The gene coding for ovine PRLR was mapped on chromosome 16 (Jenkins et al. 2000).In ovine, there are two distinct prolactin receptor isoforms: long PRLR and short PRLR produced by alternative splicing mechanism (Bignon et al. 1997) The Chios sheep breed has a high milk yield and an outstanding prolificacy.The average litter size is 2.3 and found in western region in Turkey.The Awassi is principally a milk breed, but meat production from this breed is also important and the twinning rate is 10-20 % and found southeast Anatolia in Turkey.The White Karaman is a breed indigenous to Turkey with a twinning rate of 20-30 % and found in central Anatolia in Turkey.(Akcapinar 2000).
It is well known that Chios sheep are highly prolific in comparison with many other breeds.
The objectives of the present study were investigating polymorphisms in the PRLR gene exon 10 and intron 1 by using DNA sequence analysis.Exon 10 encodes most of the cytoplasmic domain, which is essential for signal transductions (Bole-Feysot et al. 1998).

Animal resources and DNA isolation
Fifty blood samples each from Chios, White Karaman, and Awassi ewes, with a total of 150 samples were used.Jugular blood samples (2 ml per ewe) were collected from each of the animals.These individuals were chosen at randomly to prevent the relationship.Blood samples from sheep representing 10 subpopulations from 3 Turkish native sheep breeds were collected from remote villages belonging to 3 geographic regions in Turkey (Figure 1).Genomic DNA was extracted from the whole blood using the Phenol-chloroform method and then it was dissolved in 10 mM Tris-HCl (pH 8.0) buffer and kept at −20°C.

PCR amplifications and sequence analysis
PCR amplification for exon 10 and intron 1 (267 and 391bp in length) of the PRLR gene was amplified and sequenced using primers and the PCR conditions described by Chu et al. (2007).The PCR products, after being purified with the GeneClean Turbo PCR Purification Kit, were sequenced in CEQ 8 000 capillary electrophoresis system (Beckman Coulter).The sequencing was done at the Ankara University Biotechnology Institute.Fifty blood samples each of breeds were used for sequence analysis, however, successful results were obtained in only 34 (White Karaman), 37 (Awassi) and 29 (Chios) samples for intron 1 and 18 (White Karaman), 15 (Awassi) and 28 (Chios) samples for exon 10.Sequences were analysed using the BIOEDIT software (Hall 2007, http://www.mbio.ncsu.edu./BioEdit/bioedit.html) for sequence alignment.NETWORK 4.5.1.6.(Bandelt et al. 1999, http://www.fluxus-engineering.com) was used to build the network of intron 1 and exon 10 haplotype groups using the median joining algorithm.MEGA version 4.1 (Tamura et al. 2007, http://www.megasoftware.net) was used for the phylogenetic sequence analyses of haplotypes by the Neighbour-Joining method based on Kimura-2P model and the reliability of the inferred tree was assessed by bootstrap (1 000 replicates) (data not shown).
The DnaSP software 5 (Rozas et al. 2003) was used to calculate haplotype diversity (H d ) and nucleotide diversity (π); Watterson's theta estimator for the studied species separately using a haplotype sequence was obtained.Pi is based on the average number of nucleotide differences between the sequence, and theta is based on the total number of segregating sites in the sequence (Iso-Touru et al. 2009).
To estimate the effect of selection, we calculated Tajima's D (Tajima 1989), Fu and Li's D* and F* test (Fu & Li 1993) and Fu's Fs test (Fu 1997) for each group separately.Tajima's D test compares the difference between the number of segregating sites and average number of pairwise (Tajima 1989).Under neutrality Tajima's D value is assumed to be zero; under

PCR amplifications and sequence analysis
PCR amplification for exon 10 and intron 1 (267 and 391bp in length) of the PRLR gene was amplified and sequenced using primers and the PCR conditions described by CHU et al. (2007).The PCR products, after being purified with the GeneClean Turbo PCR Purification Kit, were sequenced in CEQ 8000 capillary electrophoresis system (Beckman Coulter).The sequencing was done at the Ankara University Biotechnology Institute.Fifty blood samples each of breeds were used for sequence analysis, however, successful results were obtained in only 34 (White Karaman), 36 (Awassi) and 29 (Chios) samples for intron 1 and 18 (White Karaman), 15 (Awassi) and 28 (Chios) samples for exon 10.Sequences were analysed using the BIOEDIT software (http://www.mbio.ncsu.edu./BioEdit/bioedit.html ) for sequence alignment.NETWORK 4.5.1.6.(http://www.fluxus-engineering.com) was used to build the network of intron 1 and exon 10 haplotype groups using the median joining algorithm.MEGA version 4.1 (TAMURA et al. 2007) was used for the phylogenetic sequence analyses of haplotypes by the Neighbour-Joining method based on Kimura-2P model and the reliability of the inferred tree was assessed by bootstrap (1000 replicates) (data not shown).
The DnaSP software 5 (ROZAS et al. 2003) was used to calculate haplotype diversity (H d ) and nucleotide diversity (π); Watterson's theta estimator for the studied species separately using a haplotype sequence was obtained.Pi is based on the average number of nucleotide differences positive selection there is an excess of rare polymorphisms and Tajima's D value is negative.Negative D values can also be due to population expansion.If there is balancing selection, intermediate frequency genetic variants are kept and Tajima's D value is positive (Iso-Touru et al. 2009).The statistical analysis package DnaSP 5 (Rozas et al. 2003) was used for the neutrality tests.
The impact of amino acid variants on protein structure via analysis of multiple sequence alignments was done with SIFT (Sorting Intolerant From Tolerant), which uses sequence homology to predict whether an amino acid substitution will affect protein function and hence, potentially alter the phenotype.It gives a normalized probability score value that the amino acid change is tolerated.If the score value is less than 0.05, the amino acid change is predicted to be deleterious.The median conservation value for the diversity of the sequence in the alignment is measured as well, and the default value is 3.0.Higher conservation values can lead to higher false positive error (Pauline & Henikoff 2003).

Results and discussion
In this study, the sequence analysis of the intron 1 and exon 10 of the PRLR gene revealed interesting variations in the studied populations and subpopulations.For exon 10, except for reference sequence (AF041257 haplotype 1 [H1]), seven different haplotypes were obtained (Table 1).The most common haplotype was haplotype 5 (H5) for the Chios breed and haplotype 3 (H3) for White Karaman and Awassi.Haplotypes 4-8 (H4-H8) were not detected in either the White Karaman or Awassi breeds.For intron 1, 6 different haplotypes were determined (Table 2).The most common haplotypes were haplotype 2 (H2) for White Karaman and haplotype 4 (H4) for Awassi.However, haplotype 6 (H6) was the most common for the Chios breed.These haplotypes were determined based on reference sequence AF042358 and called H1 in this study.Twelve variations were determined in exon 10, of which seven were non-synonymous mutations: g.14A>T p.Q14L; g.160G>A p.D160N; g.166G>A p.E166K; g.167A>T p.E167V; g.176A>T p.H176L; g.206G>A p.S206N and g.208G>A p.G208R; while five variations (g.2A>T, g.81A>G, g.138A>G, g.186C>T, g.207T>C) were determined as synonymous mutations (Table 1).The PRLR haplotype sequences from these sheep breeds have been deposited in the GeneBank database under the acc.no.HM437203 -HM437214.
Based on the observed mismatch distributions and the constructed radiation tree (Figure 2), two main groups were determined with each primer (Table 3).Neutrality tests were applied separately to these haplotype groups.
Neutrality tests at PRLR gene, Tajima's D value, Fu and Li's D* and F* values and Fu's Fs values are shown in Table 3. Tajima's D value, Fu and Li's D* and F* values and Fu's Fs values for intron 1 group A are positive and the others negative.Artificially selected populations, like livestock species, do not fulfil the assumptions of random mating and constant population size for the neutrality test, hence positive Tajima's D values are likely due to the demographic histories of these species or breeds rather than true balancing selection (Iso-Touru et al. 2009).Haplotype diversity (Hd), nucleotide diversity (pi) and Watterson's theta estimator were calculated separately for the studied species using the haplotype sequences obtained.Since in all groups nucleotide diversities were low but haplotype diversities were high, recent population growth is suggested.
Table 1 Polymorphic sites (excluding the ambiguous ones) and amino acid changes at the PRLR gene exon 10 for the Turkish sheep breeds.Nucleotides are numbered from 1 to 225, using AF041257 as the reference sequence.
The most common haplotype for exon 10 is haplotype 3.For exon 10 haplotypes, H2 differs from H1 by four nucleotides; H3 differs from H1 by three nucleotides; H5 differs from H1 by seven nucleotides; H4, H6 and H7 differ from H1 by eight nucleotides; H8 differs from H1 by nine nucleotides.Haplotypes 4-8 were observed only in the Chios breed.
We studied the possible impacts of the amino acid changes for the protein structure with the methods implemented in SIFT programs.The substitution p.Q14L and p.E167V was predicted to affect the protein function by SIFT analysis (Table 4).These substitutions were observed only in the Chios breed.diversity (pi) and Watterson's theta estimator were calculated separately for the studied species using the haplotype sequences obtained.Since in all groups nucleotide diversities were low but haplotype diversities were high, recent population growth is suggested.Studies on the PRLR gene have concentrated more on porcine breeds (Kernerova et al. 2009).
A mutation was identified in the porcine PRLR gene (Vincent et al. 1998).An AluI PCR-RFLP polymorphism was identified in the porcine 457 bp-long fragment of the PRLR gene (Vincent et al. 1997).A new HpaI PCR-RFLP polymorphism was identified in the porcine PRLR gene (Putnova et al. 2002).
The prolactin receptor (PRLR) gene was studied as a candidate gene for the prolificacy of Jining Grey goats.Five pairs of primers were designed to detect single nucleotide polymorphisms of exon 10 by PCR-SSCP.Only the products amplified by primers P1, P2, P4 displayed polymorphisms.For primer P1 and primer P2 sequencing revealed two mutations (g.186G>A and g.220T>C; g.52G>A and g.122G>A, respectively ) and for primer P4, sequencing revealed one mutation (g.143A>G) of the PRLR gene (Zhang et al. 2007).
The sheep PRLR gene has been screened for polymorphisms by PCR-SSCP (Mu et al. 2006).Three genotypes (AA, AB and BB) were detected by three primer pairs.Chu et al. (2007) had detected ovine PRLR gene polymorphism by PCR-SSCP.Three primer pairs were designed for PCR amplification within intron 1 and exon 10 of the PRLR gene in sheep.
The mutations we noticed (g.272 C>T) were identical for each breed to those reported by Chu et al. (2007) who found position g.84 T>C in intron 1.These investigators had used two primer pairs (termed primer 1 and primer 2) for intron 1 of the ovine PRLR gene amplification.In the present study, we used one primer pair (391bp) for intron 1 PRLR gene amplification.Iso-Touru et al. (2009) sequenced PRLR gene exon 10 (891bp), coding for the major part of cytoplasmic domain, from 6 different sheep breeds (Romanov breed, Wrzosowka breed, Dagestan local, Andi, sheep from Komi village, Finnsheep and Alandsheep).These researchers were found 6 different haplotypes, which is OVIS_PRLR1-OVIS_PRLR6 (GenBank acc.no.: FJ901296-FJ901301).
As a result, this suggests that population expansion can be based on the negative neutrality test values obtained in intron 1 group B and exon 10 groups A and B. Under positive selection there is an excess of rare polymorphisms and Tajima's D, Fu and Li's D and F values and Fu's Fs values are negative.Both population and subpopulation variations for intron 1 and exon 10 of the PRLR gene polymorphisms were found in Chios, White Karaman and Awassi ewes.It was interesting to note that White Karaman and Awassi sheep were similar to each other in terms of both intron 1 and exon 10 haplotypes, whereas the Chios breed had different variations.
The Chios breed showed greater and different haplotype diversity and different variations in comparison with White Karaman and Awassi.Whether other gene(s) or quantitative trait loci (QTL) regulating ovulation are also present requires further research.We propose that this difference is probably the consequence of different environmental conditions, selection and possibly even QTL or gene(s) regulating ovulation rate.
In the present study, intron 1 and exon 10 of the ovine PRLR gene polymorphisms were screened by the sequencing technique.We have reported here for the first time single nucleotide polymorphisms of the PRLR gene for both intron 1 and exon 10 in Turkish sheep breeds.We concluded that the identified SNPs lend themselves readily for further research regarding physiological impacts such as milk production and reproductive traits in livestock.

Figure 1
Figure 1 Geographic distribution of Turkish sheep samples

Figure 1
Figure 1Geographic distribution of Turkish sheep samples

Figure 2 A
Figure 2 A neighbour-joining phylogenetic tree (radiation style) constructed from exon 10 (A) and intron 1 (B) sequences of PRLR genes from Turkish sheep breeds.

Figure 2 Figure 4
Figure 2A neighbour-joining phylogenetic tree (radiation style) constructed from exon 10 (A) and intron 1 (B) sequences of PRLR genes from Turkish sheep breeds.

Table 2
Polymorphic sites (excluding the ambiguous ones) at the PRLR gene intron 1 for the Turkish sheep breeds.Nucleotides are numbered from 1 to 302, using AF042358 as the reference sequence.

Table 4
Predicted affection status for the amino acid substitutions from PRLR gene exon 10