Complete mitogenome of asiatic lion resolves phylogenetic status within Panthera

Base composition of mitochondrial genome of Panthera leo persica

Structure of mitogenome of Panthera leo persica

Location of the mitochondrial genes were deduced, using online annotation server MITOS [14]. Complete mitogenome contained 37 genes which include 13 protein coding genes, 2 rRNA genes, 22 tRNA genes and the control region (Figure 1). Gene order and origin of reading frame of all protein coding genes were identical to Carnivora. Major difference in P. leo persica mitogenome was observed in size of the control region (1,603 bp).

Proteins and codons

The studied mitogenome comprise of 13 protein coding genes (11,364 bp) and share 66.62% of total genome. These genes encode 3788 amino acids (Table 2). The longest gene was ND5 (1815 bp) and the shortest gene was ATP8 (198 bp). The start and termination codons appeared universal among species. ATG was commonly used as start codon for all the genes except ND2, ND5, COX1, ND6 and ND3, which used ATA, ATT and ATC, respectively. TAA was commonly found as stop codon while Cyt b ended with AGA. Termination of Cyt b gene with AGA, is typical for mammalian species [15–17] which is distinct from others such as amphibians [18], reptiles [19, 20], Aves [21] and Nematoda [22]. ND3, ND4, ND2 and COX3 have incomplete stop codons. These incomplete termination codons are common in metazoan mitogenome and can be converted into complete ones by poly A after transcription (TAA) [23].

tRNA and rRNA

Total 22 tRNA genes were found in mitogenome of P. leo persica, amongst these tRNA^ser and tRNA^leu had two termination codons (Table 2). Length of the tRNA genes in vertebrates is reported from 59 to 75 bp [24]. In this study, the shortest tRNA is tRNA^ser (59 bp) and longest is tRNA^leu (75 bp). Cloverleaf secondary structure of tRNAs of this genome was predicted, out of which representative structures of tRNA^phe and tRNA^ser are shown (Figure 2). Out of all tRNAs, tRNA^ser lacked the “DHU” arm [25]. The location of 12S rRNA (964 bp) was between tRNA^phe and tRNA^val and 16S rRNA (1576 bp) was between tRNA^val and tRNA^leu.

Control region

The control region of the mitogenome of P. leo persica is located between tRNA^Pro and tRNA^Phe. It contains only promoters and regulatory sequences for replication and transcription, but no structural genes (Figure 3). CR is divided into three parts: the left domain (16,321-16,808bp) which includes the hyper variable segment HVS-1, the central conserved region (CCR) of 529 bp, and the right domain of (280–867 bp), which include HVS-2. Further the repetitive sequence RS-2 is in the left domain located at the 5’ end. It consists of long repetitive motif of about 80 bp which is repeated 4.7 times (3’-CCCCATGAATATTAAGCATGTACAGTAGTTTATATATATTACATAAGGCATACTATGTATATCGTGCATTAACTGCTTGT–5’). RS-3 is located in right domain at 3’ end. It consists of short repeat units of 14 bp which is repeated 20 times (3’-ACACGTACACACGT-5’). RS-3 is highly variable in the arrangement of specific motifs and is heteroplasmic. It is varied in terms of repeat numbers and motif among Felidae species (Table 3) [26]. When there are repeats in the genome longer than the read length, it is important to confirm NGS data with capillary sequencing. Here, RS-2 and RS-3 repeats have longer region than read length of Ion Torrent PGM. Hence RS-2 and RS-3 were validated with capillary sequencing and capillary data (RS2 and RS3 repeats) was replaced in original genome obtained from NGS. Conserved sequence blocks (CSB-2 and CSB-3) are in HVS-2. CSB-1 is in the CCR, which is located between RS-2 and RS-3. CR of P. leo persica comprises 1,603 bp and it contains origin of heavy chain replication. The capillary data of RS2 and RS3 repeats was submitted to TRA (Trace archive Id: TR-16480).

Phylogenetic analyses based on gene cluster

Phylogenetic analyses was carried out on the basis of partial [27] and full gene sequences of 12S rRNA, 16S rRNA, ND2, ND4, ND5, ATP8 and Cyt b using ML, MP methods (Figure 4). The monophyly of the genus Panthera, including P. leo persica, Uncia uncia, Panthera pardus, Panthera tigris amoyensis, Panthera tigris, and Neofelis nebulosa was clearly depicted. Results demonstrate 100% bootstrap value in ML method and 99% in MP method, using full gene sequences of P. leo persica (Figure 4A). Previously sequenced Panthera leo partial genes formed node with Uncia uncia with 96% bootstrap value in ML method and 88% in MP method (Figure 4B). This study strongly supports that P. leo persica is a sister species of P. pardus, than of U. uncia, as earlier reported [27]. N. nebulosa was forming distinct branch. P. tigris and P. tigris amoyensis being same species, formed different branch in the Panthera node while U. uncia was placed in the lineage between tiger and lion-leopard node.

Estimates of divergence times

The ML and Bayesian analyses based on 11 complete mitochondrial genomes, excluding control region yielded the identical tree topology supported by high bootstrap value (above 90%, Figure 5) and high posterior probabilities (above 0.97, Figure 5). Uncorrelated lognormal relaxed clock (Figure 5) and strict molecular clock model (Additional file 1: Figure S1) were studied to estimate divergence times. Data of uncorrelated lognormal relaxed clock were considered for lineage specific rate heterogeneity. Divergence time was estimated using single calibration point at 10.78 mya for Felidae individuals, assuming common ancestor between genus Panthera and other Felinae species [28, 29]. BEAST analysis suggested divergence time of Felidae individual was 9.62 mya (Figure 5) from the common ancestor. The divergence of P. leo persica was estimated 2.96 mya from the closely related species P. pardus (ML, 98%; Bayesian, 1.00). Divergence of other individuals; N. nebulosa (7.09 mya), P. tigris (5.2 mya) and U. uncia (3.77 mya). Estimation of divergence times of genus Panthera in the present study reconfirms earlier reports [30, 31].

Sample collection and DNA extraction

Tissue sample of a freshly dead lion cub (female, 1.5 yrs) with chip ID 00-06B7-22E3 was collected from Sakkarbag Zoo, Junagadh, Gujarat, India (N 21.540848, E 70.468481). The samples were immediately transferred to laboratory and stored in Allprotect Tissue Reagent (Qiagen) at −80°C. Tissue sample was homogenized in liquid nitrogen and DNA extraction was carried out using BioVison Mitochondrial DNA Isolation Kit (BioVision Research Products, CA, USA) following manufacturer’s protocol.

Sequencing

DNA quality such as ratio of absorbance at 260/280 nm and 260/230 nm was measured using Nanophotometer (Imlpen). Qubit® 2.0 Fluorometer was used to obtain an accurate quantitation of DNA. Library was prepared using Ion Express Plus Fragment library kit (Life Technologies). Mitochondrial DNA was sheared into blunt ended fragments by enzymatic lysis using Ion Shear Plus Reagents (Life Technologies). The fragmented DNA was ligated to Ion-compatible adapters, followed by nick repair to complete the linkage between adapters and DNA inserts. Sequencing was performed using Ion Express Template 300 chemistry (Life Technologies) on 316 chip following manufacturer’s protocol.

Capillary Sequencing of D-loop region was performed to validate Ion Torrent data. PCR of D-loop was performed using forward (5’TCAAGGAAGAAGCAATAGCC 3’) and reverse (5’GGATTGTTGGGCGTGTAAA 3’) primers. Further sequencing was carried out using PCR primers and internal primer (5’TATTCTCTATGCGGGGGTTC 3’) using Big Dye v3.1 Chemistry (Applied Biosystems) on 3500 Genetic Analyzer (Applied Biosystems) following manufacturer’s protocol.

Data analysis

Reads longer than 100 bp were mapped to the P. pardus reference mitogenome [GenBank: NC_010641], using CLC Genomic Workbench 5 [32]. Assembled mitogenome was annotated, using MITOS online mitochondrial genome annotation server [14]. Control region repeats RS-2 and RS-3 were identified using Tandem Repeats Finder [33].

Molecular phylogenetic analysis

Phylogeny of P. leo was constructed based on partial and full mitochondrial gene cluster (12S rRNA+16S rRNA+ND2+ND4+ND5+ATP8+Cyt b). Partial P. leo genes having accession number A79300, AF006457, AY170043, AY634398, AF006458, S79302 and DQ899945 [27] were retrieved from NCBI. Full sequences of above same genes of P. leo persica were compared with gene sequences of P. uncia [NC_010638], P. pardus [NC_010641], P. tigris [NC_010642], P. tigris amoyensis [NC_014770], N. nebulosa [NC_008450], A. jubatus [NC_005212], P. concolor [NC_016470], L. rufus [NC_014456], F. catus [NC_001700] and P. bengalensis euptilurus [NC_016189] obtained from NCBI organelle genome resource (Table 1). The data was subjected to Maximum likelihood (ML) and Maximum parsimony (MP) methods using MEGA v5.05 software package [34] to establish phylogenetic relationship.

These seven genes were concatenated and multiple alignment was carried out using ClustalW built-in MEGA. The number of bootstrap replicates was set to 1000 and phylogeny was constructed using Hasegawa-Kishino-Yano model, determined by MEGA 5.05 as the best fitting nucleotide substitution model.

Divergence time estimation

Multiple sequence alignment of genomes was carried out and phylogenetic tree was constructed using HKY+G+T substitution model at 1000 bootstrap replicates. Divergence times were estimated between species using Bayesian evolutionary analysis by sampling trees (BEAST v1.7.1) software [35]. Markov Chain Monte Carlo (MCMC) procedure was used within a Bayesian analysis framework to estimate posterior distributions of evolutionary rates and divergence times. These analyses were performed using DNA sequence alignment of complete mitogenomes, excluding control region. Divergence times were estimated using an uncorrelated lognormal relaxed clock to account for lineage-specific rate heterogeneity as well as using strict molecular clock model (Additional file 1: Figure S1). Bayesian MCMC analyses in BEAST were performed using HKY model of evolution and gamma+ invariant rate heterogeneity models. Divergences were estimated under a Yule speciation process that is generally more appropriate when considering sequences from different species [36]. Monophyletic constraints were imposed for the node to calibrate evolutionary rates. Normal priors were used for the times to the most recent common ancestor (TMRCA) of Pantherinae and Felinae is (mean 10.78 ±1.87 mya), based on the posterior distributions obtained [29].

Simultaneous Markov chains were run for 10,000,000 generations, sampling every 1000 steps. Thus the “.trees” file will contain 10,000 trees and after removing 10% burnin, tree was annotated using TreeAnnotator. Posterior probability limit was set to zero to annotate all nodes. Maximum clade credibility tree was selected for target tree to find highest product of the posterior probability for all nodes. The phylogenetic tree graphic was generated using FigTree v1.3.1 [37].

GenBank accession number

The complete mitogenome sequence with gene prediction and functional annotation was submitted to GenBank with accession no. KC834784. The raw sequence data generated from Ion Torrent was submitted to SRA with accession no SRR821548. Mitochondrial D-loop sequence was submitted to GenBank with accession no. KC917264.