Genetic analysis of the Turkish gray wolf ( Canis lupus ) based on partial mitochondrial DNA sequences

In this study, we focused on determining the genetic variability of Turkish gray wolves and also to reveal phylogenetic relationships of the Eurasian wolf populations using mitochondrial DNA sequences. Partial mitochondrial DNA sequences (440 bp) were obtained from 12 Turkish gray wolves, including D-loop region (332 bp). We found seven D-loop haplotypes (332 bp) among the 12 Turkish gray wolves. The D-loop sequences of the Turkish gray wolf were compared to sequences registered in GenBank from Eurasia under the name Canis lupus. Five Turkish haplotypes were shared with conspecific sequences from other regions of Eurasia in the species’ range. Two haplotypes were unique for the Turkish wolves. The existence of shared haplotypes suggests that the gray wolves in Turkey and other regions might be originated from the same gene pool. The present study reports significant data for determining the genetic variability and revealing the phylogenetic relationships of Turkish gray wolves; it is suggested that the genetic variability of Turkish gray wolves is relatively high.

The gray wolf has been considered the first mammal species domesticated by humankind (Clutton-Brock 1995), with multiple origins (Pang et al. 2009, von Holdt et al. 2010).It is also known to be the most important wildlife species (Sillero-Zubiri et al. 2004).
The Turkish wolves inhabit all biomes, which include prey.The population size of the Turkish gray wolf has been estimated at around 5000 -7000 individuals (Can 2004, Salvatori & Linnell 2005).Due to the decrease of prey animals such as sheep, gazelle, red deer, roe deer, wild boar and brown hare, habitat fragmentation and illegal hunting etc., the Turkish gray wolves have shown a decline in the last century; they occur in small subpopulations (Can & Togan 2004, Salvatori & Linnell 2005, Şekercİoğlu et al. 2011).Therefore, measures for conservation management have been taken by the Ministry of Forestry and Water Affairs in Turkey (see www. ormansu.gov.tr).
By presenting preliminary results of an analysis of partial mitochondrial sequences, we aim to contribute to the knowledge of the genetic variability of Turkish gray wolves and the relationships with other conspecific populations throughout Eurasia.

Sampling
The tissue samples (ear, skin, tail, etc.) were collected from the 12 road-killed gray wolves in the Asian part of Turkey (Table 1, Fig. 1).

DNA extraction and PCR conditions
Tissue samples stored in 99% ethyl alcohol were used to obtain genomic DNA, which were extracted by using a commercial extraction kit (The DNeasy Blood and Tissue Kit, Qiagen) according to the manufacturer's instructions.A short fragment of mitochondrial DNA including the D-loop (control region) was amplified by using DLH (Forward: 5'-CCTGAAGTAAGAACCAGATG-3') and LF15926F (Reverse: 5'-ATATAAAATACTTTGGTCTT GTAAACC-3') primers (Kirschning et al. 2007).

Sequencing, Aligning and Analyzing
The Macherey-Nagel Nucleospin Gel and PCR Cleanup kit were used to purify the PCR products, and sequencing of which was performed in forward and reverse directions with PCR primers; DLH and LF15926F (Kirschning et al. 2007), by using a sequencer (ABI 3100 Genetic Analyzer).We used Geneious v.R6.1.6(http://www.geneious.com), in which MAFFT v7.017 was used for the multiple sequence alignment with default parameters (Katoh et al. 2002), to align the mitochondrial DNA sequences, and DnaSP v.5.10.01 (Librado & Rozas 2009) to determine mitochondrial haplotypes and to estimate haplotype and nucleotide diversities of gray wolf.MEGA 6.0 (Tamura et al. 2013) was used to calculate genetic distances among the Turkish haplotypes of gray wolf, based on the K2P (Kimura 2-parameter) model of DNA substitution (Kimura 1980).
Phylogenetic and network analyses were performed with mitochondrial D-loop sequences of 317 bp from Turkish gray wolves and GenBank (NCBI: The National Center for Biotechnology Information), including Dloop sequences registered under the name Canis lupus from Eurasia (Table 1).The golden jackal (Canis aureus) (AY289997: Aggarwal et al. 2007) was used as outgroup in phylogenetic analysis.
The HKY (Hasegawa-Kishino-Yano) + I (Invariant) + G (Gamma) was the most suitable model of nucleotide substitution according to the corrected Akaike In formation Criterion (AICc) and the Bayesian In for mation Criterion (BIC) using jModeltest2 (Darriba et al. 2012), and this model was used in phylogenetic analysis to reconstruct Bayesian (Bayesian Inference) tree by means of MrBayes v.3.2 (Ronquist et al. 2012).In Bayesian analysis (Ronquist et al. 2012), the MCMC (Markov Chain Monte Carlo) technique was used to calculate the Bayesian posterior probabilities for 2.6 million generations with tree sampled every 100 generations, and to discard the first 25% of samples as burn-in (The Average Standard Deviation of split Frequencies < 0.01: 0.009400).The remaining samples after discarding burnin, were retained and to estimate posterior probability and 95% Bayesian credibility interval, and to create the consensus tree (50% majority rule).FigTree v1.3.1 soft-ware (Rambaut 2009) was used to picture the Bayesian phylogenetic tree.

Results
A fragment of the 440 bp of mtDNA (tRNA Thr + tRNA Pro + D-loop) was amplified successfully for the 12 Turkish gray wolves, including 332 bp for D-loop (control region).A comparison of the 440 bp fragments revealed the presence of seven haplotypes (Tr.Cl.D1 -Tr.Cl.D7) with 12 segregating sites and 10 parsimony informative sites.The haplotype Tr.Cl.D1 was the most common, and it was found in four out of the 12 samples.Dis tribution of Turkish haplotypes was given in Table 1 and Figure 1.Seven haplotypes of the Turkish gray wolf (Tr.Cl.D1 -Tr.Cl.D7) are deposited in the GenBank database (Accession numbers: KY039989 -KY039995).The haplotype (Hd) and nucleotide (π) diversities were 0.8788 and 0.01088, respectively.The Turkish wolves were polymorphic.Mean sequence divergence of the seven Turkish wolf haplotypes was 0.012, ranging from 0.002 to 0.021, based on K2P.When sequences including D-loop region (332 bp) of the Turkish samples were analyzed, seven D-loop haplotypes were found (Table 1).The haplotype (Hd) and nucleotide (Pi) diversities for the D-loop region of Turkish wolf were 0.8788 and 0.01392, respectively.
Phylogenetic relationships of the mitochondrial Dloop haplotypes (317 bp) are shown in the Bayesian tree (Fig. 2).Branch nodes of the Bayesian tree (Fig. 2) were supported with relatively high posterior probabilities (0.5 -1).In the Bayesian tree (Fig. 2 In the median joining haplotype network (Fig. 3), the Turkish haplotypes of this study grouped together with  conspecific haplotypes in the inner parts; this grouping was not entirely compatible with the Bayesian tree (Fig. 2).

Discussion
The gray wolf, Canis lupus, is a widespread canid species in Turkey (Kryštufek & Vohralik 2009, Ambarli et al. 2016).Although there are several genetic studies including Turkish gray wolves, the genetic characterization of the Turkish gray wolves is still scarce.To date, the samples of Turkish gray wolf have only been included in six studies that used mitochondrial DNA sequences (Vilá et al. 1997, 1999, Randi et al. 2000, Pilot et al. 2010, 2014, Aksöyek et al. 2016).
In the present study, genetic variability and phylogenetic relationships of Turkish gray wolves were researched by using partial sequences of mitochondrial DNA (tRNA Thr + tRNA Pro + D-loop: 440 bp).The analysis revealed seven haplotypes, two of them (Tr.Cl.D3 and Tr.Cl.D4) were unique and new for the Anatolian part of Turkey, when compared to the other geographical regions Eurasia.One postulates that this may be an indication of a relatively high genetic diversity of the Turkish gray wolf.
The mutation rate in the mitochondrial D-loop region of wolves is relatively high (Savolainen et al. 2000), which might explain a high genetic variability, as had been previously reported in other two Turkish canids, Canis aureus (İbİş et al. 2015a) and Vulpes vulpes (İbİş et al. 2014).Similarly, a high genetic diversity in gray wolves was also reported in some studies (Jędrzejewski et al. 2005, Pilot et al. 2010, Asadi-Aghbolaghi et al. 2014, Bray et al. 2014, Djan et al. 2014, Pilot et al. 2014), including packs in Białowieża Primeval Forest on the border between Poland and Belarus, and in the modern European, Iranian, Saudi Arabian Dinaric-Balkan and Caucasus populations.
In the dataset including 317 bp fragments composed of sequences of the present study and those available in the GenBank from Eurasia under the name Canis lupus (Table 1), we found two new haplotypes (Tr.Cl.D3 and Tr.Cl.D4) in the 12 Turkish gray wolves (Table 1).The five remaining haplotypes (Tr.Cl.D1, Tr.Cl.D2, Tr.Cl.D5, Tr.Cl.D6 and Tr.Cl.D7) were shared with sequences derived from the GenBank (Table 1).Four D-loop sequences of Turkish gray wolves analyzed in previous studies (Vilá et al. 1997, 1999, Randi et al. 2000, Pilot et al. 2006, 2010, 2014) are of different lengths (257 bp, 542 bp, 658 bp and 659 bp).Of these sequences, three sequences (AF115693: Randi et al. 2000, FJ978020 and FJ978033: Pilot et al. 2010, 2014) were included in the analyses of present study.On the other hand, the sequence AF005296 (257 bp) (Vilà et al. 1997, 1999, Pilot et al. 2006) was not used in the current analyses due to the shortage of the mentioned sequence.
Our analyses indicate that the Turkish haplotypes in this study were relatively distant from each other and clustered into distinct sublineages in the Bayesian tree (Fig. 2).In Figure 2, the previous Turkish sequence, FJ978033 (Pilot et al. 2010(Pilot et al. , 2014)), was grouped with the  haplotypes Tr.Cl.D1 and Tr.Cl.D2 of this study, whereas the remaining previous Turkish sequences, AF115693 (Randi et al. 2000) and FJ978020 (Pilot et al. 2010(Pilot et al. , 2014)), were clustered relatively distant from the haplotypes of this study.The phylogenetic resolution in the Bayesian tree (Fig. 2) was relatively low, which might be due to the high mutation rate in mitochondrial D-loop region of wolves.Similarly, a low phylogenetic resolution was also reported by Lebarbenchon et al. (2010) and Rodrigues et al. (2016) for least weasels (Mustela nivalis), based on the D-loop (control region) sequences.
The Turkish wolf haplotypes were grouped together with conspecific haplotypes in distinct subgroups in the inner parts of the haplotype network (Fig. 3), which is considered to show the intraspecific phylogenetic relationships better than phylogenetic trees (Crandall et al. 2000).The Bayesian tree (Fig. 2.) and the haplotype network (Fig. 3) were partly in agreement with each other, and both of them (Figs. 2, 3) have shown that the gray wolf haplotypes were divided into two main haplogroups.However, a clear geographical pattern was not observed for the distribution of gray wolf haplotypes in the Bayesian tree (Fig. 2) and the haplotype network (Fig. 3).In this context, the result of present study was compatible with that of Pilot et al. (2006).The existence of shared D-loop haplotypes found in this study suggested ongoing and/or past gene flow within the gray wolves in Turkey and other regions of Eurasia and that they originated from the same gene pool.Therefore, this study supported the view of Pilot et al. (2014) suggesting the ongoing and/or past gene flow between the Caucasian and Eastern European wolf populations.
To confirm the high genetic diversity and to elucidate the phylogeography of Turkish gray wolves, additional analysis is needed to obtain more data from the Turkish gray wolves in a multilocus study carried out by using two molecular markers, mitochondrial and nuclear DNA.

Fig. 3 .
Fig. 3. Median-joining network constructed from the mitochondrial D-loop haplotypes (317 bp) of gray wolves.The bold circles indicate the Turkish gray wolf haplotypes (seeTable 1 for haplotype codes).