白吻动胸龟和剃刀麝香龟线粒体全序列比较及动胸龟科在龟鳖目中的系统发生位置（已处理）

白吻动胸龟和剃刀麝香龟线粒体全序列比较及动胸龟科在龟鳖目中的系统发生位置（已处理） The complete mitochondrial genomes of Kinosternon leucostomum and Sternotherus carinatus and the phylogenetic position of Kinosternidae within Testudines 5 10 15 20 25 30 35 40 Jing Wanxing12 Zhang Yanyun12 Nie Liuwang12 Dai Xueting12 Xia Xingquan12 Huang Zhengfeng12 Wang Ling12 Jiang Yuan12 Liu Luo12 1 Life Science College Anhui Normal University Anhui Wuhu 241000 2 the Provincial Key Lab of the Conservation and Exploitation Research of Biological Resources in Anhui Wuhu 241000 China Abstract We first determined two complete mitochondrial genomes of Kinosternidae Kinosternon leucostomum 16 559bp and Sternotherus carinatus 16 554bp by primer walking and long-PCR technique which were deposited in the GenBank database under the accession number FJ915117 and HQ114563 respectively In the two sequences we identified three distinctive features 1 a 31bp nucleotides intergenic regions located between the COX1 and tRNASer UCN genes but found so far generally overlapped by 9bp or 5bp in other turtle families 2 a 10bp spacer between tRNACys and tRNATyr while most other turtle families had no spacers 3 a single nucleotide insertion in the ND4 gene of S carinatus which was rarely observed in other turtles The phylogenetic analyses based on the 12 mtDNA protein genes indicated that the ML and BI trees all showed K leucostomum and S carinatus assembled together to form a clade of the Kinosternidae which was sister group to the Chelydridae and the assemblage of Kinosternidae Chelydridae had a sister-taxon relationship to the Cheloniidae Testudinoidea clade Although the Asian box turtle and the kinosternids all had the kinetic plastron with the hinge it was found that the Asian box turtle was not close to the Kinosternidae indicating that the hinge structure might not be derived from one common ancestor Key words Kinosternidae mitochondrial genomes phylogenetic the kinetic plastron 1 Instruction The diversity of turtles in the world that have existed in modern times since 1 500 AD and currently generally recognized as distinct consists 328 species Of these 56 are polytypic representing 124 additional recognized subspecies TTWG 2010 They were divided into 14 families and one of which Kinosternidae consists of four genera involving 25 species The challenge of resolving phylogenetic position of Kinosternidae within Testudines has led to several studies that utilize both morphological and molecular data and some hypotheses have been represented to show many debates Fig 1 Shaffer et al 1997 consider Kinosternoidea and Chelonioidea as one clade based on 115 morphological characters combined with the mitochondrial cytochrome b Cyt b gene 12s rDNA data Fig 1 A but the result of Brinkman and Wu 1999 refutes this hypothesis using morphological data which places Kinosternoidea close to Trionychina Fig 1B Based on the U17 small nucleolar RNA snoRNA sequences it is showed that the Kinosternoidea has a close affinity with the Chelydridae while the Chelonioidea is sister group to them Fig 1C Cervelli et al 2003 Contrary to this hypothesis Krenz et al 2005 indicate the Kinosternoidea has a close relationship to the Clade of Chelydridae Chelonioidea using combined data of RAG-1 cytb 12s rDNA Fig 1D Referring to partial 基金项目This research was supported by the National Natural Science Foundation of China NSFC No 30970351 The Specialized Research Fund for the Doctoral Program of Higher Education of China grant number 20080370001 and the Key Laboratory of Biotic Environment and Ecological Safety of Anhui province 作者简介景万星男硕士研究生 通信联系人聂刘旺1962 年生男教授博士生导师主要研究方向为分子生态 学和细胞遗传学 E-mail lwniecomcn -1- mtDNA Parham et al 2006 point out the Kinosternoidea has a sister-taxon relationship to Chelydridae Chelonioidea Testudinoidea Emydidae Platysternidae Geoemydidae 45 50 55 60 65 70 Testudinidae Fig 1E Using 14 nuDNA genes Barley et al 2010 reconstruct the phylogenetic tree of turtles which suggests the Chelonioidea is sister group to the Kinosternidae Chelydridae in common with the result of Cervelli et al 2003 Fig 1 F In the former phylogenetic studies the different data sets are used to infer the evolutionary relationships of the organisms Increasing evidence demonstrated that the phylogenetic performance may vary based on different genes which may provide a misleading inference of the true inter-specific and inter-generic evolutionary relationships Zardoya and Meyer 1996 Miya and Nishida 2000 Roques et al 2006 Vertebrate mitochondrial DNA mtDNA is a closed circular molecule The genome organization is highly compact and the genome is approximately 16 kb in length consisting of 13 protein coding genes 22 transfer RNA genes two ribosomal RNA genes a non-coding control region and some intergenic spacers Roques et al 2006 The arrangement of these genes varies among the different animal lineages and provides the basis for the comparison of mitochondrial genomes to study evolutionary events Gissi et al 2008 Meganathan et al 2011 Moreover the abundance of mitochondria in cells lack of recombination absence of intron higher evolutionary rates of mitochondrial genes and uniparental inheritance makes these molecules a reliable tool to infer phylogenetic relationships at various taxonomic levels Gissi et al 2008 Thus the generation of mitochondrial genome sequences is useful for phylogenetic analyses and also for informing conservation efforts Previous studies demonstrate that mitochondrial genomes can recover robust phylogenies with high statistical support for many taxa and thus may resolve questions of higher-level relationships of turtles Miya and Nishida 2000 Zhang and Nie 2007 There are about 40 complete mitochondrial genomes of living turtles within the GenBank but no the complete mitochondrial genomes of Kinosternidae Therefore the complete mitochondrial genomes for this family are needed In this work in order to re-examine the phylogenetic position of Kinosternidae within Testudines we first sequenced the complete mitochondrial genomes of the K leucostomum and S carinatus which were described to increase the amount of the molecular data about the Kinosternidae Based on the ML and BI methods the complete mitochondrial genomes of 30 species were used to perform molecular phylogenetic analyses -2- Fig 1 Recent hypotheses of the relationships of the major turtle lineages Clade support values when available in 75 the original study are provided A Values represent MP bootstrap proportions nodes without values represent nodes with less than 50 support D Values represent ML bootstrap proportions and BI posterior probabilities E Values represent ML bootstrap proportions and BI posterior probabilities F Values represent BI posterior probabilities and ML bootstrap proportions 2 Materials and methods 80 21 Turtle sample and DNA extraction The specimens of K leucostomum and S carinatus were stored in the specimen storeroom of College of Life Sciences in Anhui Normal University All the samples were preserved at -20?C before DNA extraction Total genomic DNA was extracted from the muscle tissue of the tail with the standard proteinase K method Sambrook and Russell 2001 and kept at -20 ?C until needed 85 for polymerase chain reaction PCR amplification 22 Amplification and sequencing The overlapped fragments were amplified to obtain the complete mtDNA molecule with the same sequencing strategy for the both species The PCR was conducted with primers listed in Table 1 and carried out in a total volume of 50 μL containing 335 μL sterile distilled water 3 μL -3- 90 of sample genomic DNA 5 μL of 10×Buffer TaKaRa Japan 4 μL of MgCl2 25 mol L-1 3 μL of dNTP 25 mM 10 μL of each primer 5 μmol L-1 and 05 μL Taq DNA polymerase 5 UμL TaKaRa using Bio-RAD iCyclers The thermal cycles were 95 ?C pre-denaturing for 2 min followed by 34 cycles of 94 ?C for 40 s 51 ?C to 58 ?C for 45 s and 72 ?C for 1 min plus a final extension at 72 ?C for 10 min The annealing temperature was optimized when necessary 95 Table 1 Oligo nucleotide primers used for K leucostomum and S carinatus Upper primer sequence 5?3 Lower primer sequence 5?3 F1 F2 F3 F4 F5 F6 F7 Fa8 Fa9 Fa10 Fa11 Fa12 Fa13 Fa14 Fa15 Fa16 Fb8 Fb9 Fb10 Fb11 Fb12 Fb13 Fb14 Fb15 AAAGCATTCAGCTTACACCTGA GTCTCTTACAAATAATCAGTGA ACCTGACAAAAACTAGCCCCA AGGATAGAAGTAATCCAATGG AACCACCGTTGTATTCAACTA GCTATCCCCAACAGGAGTAAAAG GCCGCTACCTACAAGAAAAC CGGCGTAAAATGTGACTAA AGCCATCTTACCTGTGTTTT TACGGACAATGCTCAGAAA GTTGTCTGACTATTCCTT ATCGGACAAATAGCCTCAA CCTCAAACTCAAAATACGCT GAACCAGTTACACGAAAACG CATACACGCMTTCTTYAAAGC AGCCTCCATCTTATACTTCA AAAGCACGGCACTGAAGATGC THTTCTCYACYAACCAYAAAG ACGCAGTTCCAGGACGAT AGTACAAATGACTTCCAATCA CATGTTCCGACCATTCAC ATTTGACTTCCACTCCCT AGCAGCCTCCATCCTWTACTT GCCTTCTCAACAGCAAGC R1 R2 R3 R4 R5 R6 R7 Ra8 Ra9 Ra10 Ra11 Ra12 Ra13 Ra14 Ra15 Ra16 Rb8 Rb9 Rb10 Rb11 Rb12 Rb13 Rb14 Rb15 AAGTTCCACAGGGTCTTCTCG AGATTAGGTATATTGGTTCTTG ACTATTCCTGCTCAGGCHCCG TATCTTTCGRATGTCTTGTTC CAATCTTTGGTTTACAAGACC GCTATCCTGTTTAGCTTCTATAG GAARAATCGAATTGAGAATGG GAGGTACAAGGGTTAATCT TTGCTCAAGTTTGGTGTAGT GTATAGAGTTGTGATGGA GCTCTACTTGATTACCTC GCAATAACTAACAGCAAG GATAGGGTAATGATTGTGGA GCTGTTTTTACGGCTGTTTTTG CTAATAGTGATCCGAAGTTTCAT TTTTAGTCACATTTTACGCC TTTCATCTTTCCTTGCGGTAC AAATCCTGCTATRATRGCGAA GCAAGACTTGAATGGTAG TTTGRTTWCCTCATCGTGTG AGGGATTTCTCACTGGTG GGTAGTCCTGCTAGTGAT CAGTCTCATTGAGTYGGCAG CCGTTGGCATGGAGGTTT F and R refer to light and heavy strand primers respectively Y CT R A g W A T M AC H AC T The primers F1-7 and R1-7 were shared by K leucostomum and S carinatus The primers Fa8 – Fa16 and Ra8 – Ra16 were applied to K leucostomum The primers Fb8 – Fb15 and Rb8 – Rb15 were applied to S carinatus 100 The PCR products were then recovered and concentrated using the Gel Extract Purification Kit TaKaRa Japan according to the manufacturers protocol The sequencing was carried out bi-directionally on an ABI 3730 DNA automated sequencer Invitrogen Biotechnology All adjacently segmental sequences overlapped approximately 80 - 200bp 105 The existence of mitochondrial genes in the nuclear genome in eukaryotes is evident from previous studies Hurst and Jiggins 2005 These sequences may confound the results of complete mitochondrial genome sequences To overcome these difculties the Long and Accurate PCR LA-PCR method was used to generate the complete mitochondrial genome of K leucostomum and S carinatus 110 23 Sequence analyses DNA sequences were aligned using ClustalX ver 18 Thompson et al 1997 and analyzed using DNAstar software BLAST searched at National Center for Biotechnology Information NCBI were used to verify the similarity between our sequences and the mitochondrial genomes of other turtleZheng et al 2000 The locations of protein-coding rRNA and tRNA genes were -4- 115 120 identified by tRNA Scan-SE com httplowcomtRNA Scan-SE and SQUEIN ver 1000 The base compositional frequencies were estimated using MEGA ver 4 Tamura et al 2007 Synonymous and non-synonymous substitution rates were calculated using the kaks calculator Zhang et al 2006 The stem-loop secondary structures of L-strand replication origin OL were identified by the Vienna RNA package ver com Hofacker et al 1994 Gruber et al 2008 24 Phylogenetic analysis Table 2 The thirty representative species used for phylogenetic analyses Suborder Eusuchia Pleurodira Cryptodira Family Crocodylidae Pelomedusidae Chelidae Trionychidae Kinosternidae Chelydridae Cheloniidae Testudinidae Geoemydinae Platysternidae Emydidae Species Crocodylus siamensis Caiman crocodilus Pelomedusa subrufa Chelus fimbriata Palea steindachneri Pelodiscus sinensis Apalone ferox Kinosternon leucostomum Sternotherus carinatus Chelydra serpentina Macroclemys temminckii Lepidochelys olivacea Chelonia mydas Indotestudo elongata Testudo marginata Geochelone pardalis Testudo kleinmanni Indotestudo forstenii Mauremys mutica Mauremys reevesi Cuora mouhotii Cuora aurocapitata Cuora flavomarginata Cuora amboinensis Sacalia quadriocellata Cyclemys atripons Platysternon megacephalum Platysternon megacephalum Trachemys scripta Chrysemys picta Accession No DQ353946 NC_002744 AF039066 HQ172156 FJ541030 AY687385 FJ890514 FJ915117 HQ114563 EF122793 EF071948 AM258984 AB012104 DQ656607 DQ080047 DQ080041 DQ080048 DQ080044 DQ453753 AY676201 DQ659152 AY874540 EU708434 FJ763736 GU320209 EF067858 DQ016387 DQ256377 FJ392294 AF069423 Author submission time Ji et al 2006 Janke et al 2000 Zardoya et al 1997 Nie et al 2004 Nie et al 2008 Nie et al 2004 Nie et al 2009 Nie et al 2009 Nie et al 2010 Nie et al 2006 Nie et al 2006 Tandon et al 2006 Kumazawa et al 1998 Nie et al 2006 Parham et al 2005 Parham et al 2005 Parham et al 2005 Parham et al 2005 Nie et al 2006 Nie et al 2004 Nie et al 2006 Nie et al 2005 Nie et al 2008 Nie et al 2009 Nie et al 2009 Nie et al 2006 Nie et al 2006 Parham et al 2005 Beckenbach et al 2008 Sorenson et al 1998 These mitochondrial genomes were all submitted to the GenBank by our research group This study 125 The names of the 28 turtles were referred to TTWG 2010 To analyze phylogenetic position of Kinosternidae within Testudines we chose 30 complete mitochondrial genomes from GenBank in which 28 sequences belong to 10 families of turtles Pelomedusidae Chelidae Trionychidae Chelydridae Cheloniidae Testudinidae Geoemydidae -5- 130 135 140 145 150 Platysternidae and Emydidae Table 2 Crocodylus siamensis DQ353946 and Caiman crocodilus NC_002744 were used as outgroup based on the alliance between turtles and crocodilians These sequences were aligned using ClustalX ver 18 Thompson et al 1997 After removing the intergenic regions VNTR sequences within the control regions ambiguous positions and gaps a 15 713 bp alignment was subjected to phylogenetic analyses The best-t model of sequence evolution was determined using the Akaike Information Criterion AIC implemented in jModelTest ver com Posada 2008 Guindon and Gascuel 2003 The best-fit model was GTRIG with Likelihood settings base 03621 03147 00886 02346 nst 6 rmat 10345 55384 10953 04855 105508 10000 rates gamma shape 09000 ncat 4 and pinvar 02750 Phylogenetic trees were reconstructed using imum Likelihood ML and Bayesian inference BI method Approximate models of the best-fit model were implemented in the ML analysis and the BI analysis The online PhyML ver 30 httpcomphyml Guindon and Gascuel 2003 was exploited for the ML analysis which was initiated with BioNJ using GTR model with the proportion of invariable sites fixed to 0275 the rates of evolution from site to site using a discrete gamma distribution and the shape parameter of the gamma distribution fixed to 090 based on the best-fit model Bayesian BI analyses were performed using MrBayes ver 312 Huelsenbeck and Ronquist 2001 In Bayesian analysis four Markov chains were calculated simultaneously initiated with random starting trees and run for 1 000 000 generations with chains sampled every 100 generations and genetic code selected vertebrate mitochondrial DNA vertmt Robustness of the tree was estimated using 100 bootstrap replicates for ML analysis while support for nodes in Bayesian analysis was assessed using posterior probabilities 3 Results 31 Genome organization and features The map of the complete mitochondrial genomes of K leucostomum and S carinatus was 155 160 165 presented in Fig 2 and showed the mitochondrial genomes of K leucostomum and S carinatus were respectively 16 559bp and 16 554bp in length They all consisted of 2 rRNAs 22 tRNAs and 13 protein-coding genes with a control region Table 3 The ND6 gene and eight tRNA genes were encoded on the L-strand and other genes were encoded on the H-strand All 13 protein-coding open reading frames ORFs generally found in other vertebrates were also present in the two mitochondrial genomes with the same organization and similar length Table 3 However the length of ATP8 180bp in K leucostomum and 177bp in S carinatus was longer than the generally length 168bp of other turtles and a single nucleotide insertionAin the ND4 gene of S carinatus was found at the position 1012bp The length differences of overlapped sequences between the two turtles emerged within four regions ND2-tRNATrp ATP8-ATP6 ND4L-ND4 and ND5-ND6 Table 3 The longest intergenic regions consisting of 31bp were located between the COX1 and tRNASer UCN genes and 10bp spacers were found between tRNACys and tRNATyr in K leucostomum and S carinatus -6- Fig 2 Organization of the mitochondrial genomes of K leucostomum and S carinatus ND1-6 and ND4L 170 175 180 185 190 195 200 subunits 1-6 and 4L of nicotine amid adenine dinucleotide dehydrogenase ATP6 and ATP8 subunits 6 and 8 of adenine triphosphatase COX1–3 cytochrome coxidase subunits 1–3 12S and 16S 12S and 16S rRNA Each tRNA gene is identified by the 3-letter amino acid code D-loop Control region OL represent the replication origin of L-strand The base composition of the complete sequences were A 360 G 120 C 238 T 282 and AT 642 in the K leucostomum and A 358 G 120 C 249 T 273 and AT 631 in the S carinatus L-strand replication origin The L-strand replication origin OL was located in a cluster of 5 tRNA genes known as the WANCY region between the tRNAAsn and tRNACys genes In K leucostomum and S carinatus the OL was 28 nucleotides which had the potential to fold in a stem-loop secondary structure with a stem comprised of 10 paired nucleotides and a loop of 8 nucleotides fig6 24 25 The conserved motif 5-GCGGG-3 Hurst et al 1999 was also found at the base of the OL stem In addition the stem region of OL had CGG which is considered the initiation site for L-strand replication and the trinucleotide sequence is identical in several turtles Brennicke and Clayton 1981 Zhang and Nie 2007 Control region The control regions CR of K leucostomum and S carinatus which spaned 1 006 bp and 1 003 bp respectively were located between tRNAPro and tRNAPhe consistent with its location in other tutles Table 3 The contents of A T G and C were 338 328 109 225 in K leucostomum and 330 332 113 225 in S carinatus The control region comprises three domains termination associated sequence TAS domain central conserved domain CD and conserved sequence block CSB domain Sbisa et al 1997 Within the control regions of K leucostomum and S carinatus we identified the putative termination associated sequences TAS and the conserved sequence block CSB 1 2 3 and F according to other turtles and other vertebrates Doda et al 1981 Walberg and Clayton 1981 Yan et al 2008 Zhang et al 2009 and found the cored sequences ACAT and TGTA of the TAS domains which were thought to act as recognition sites for the termination of H-strand replication by forming stable hairpin-loop structures Zhang et al 2009 Only one type of VNTR sequence was identified at the 3 end of CRs possessing the same TA motifs with copy numbers 27 and 17 for K leucostomum and S carinatus respectively -7- Table 3 Features of the K leucostomum and S carinatus mitochondrial genomes Region KLb Position KL Size bp SC Strand 3intergenic spacesa KL SC Phe 12S rRNA tRNAVal 16S rRNA tRNALeu UUR ND1 Ile tRNAGln tRNAMet ND2 tRNATrp Ala tRNAAsn tRNACys tRNATyr COX1 tRNASer UCN Asp COX2 Lys ATP8 ATP6 COX3 tRNAGly ND3 tRNAArg ND4L 1-69 70-1029 1030-1100 1101-2697 2698-2771 2772-3740 3740-3809 3809-3879 3879-3947 3949-4992 4988-5056 5058-5126 5128-5200 5227-5292 5303-5372 5374-6922 6954-7024 7027-7094 7095-7775 7777-7849 7851-8030 8021-8704 8704-9487 9488-9556 9557-9907 9911-9980 9981-10277 1-70 71-1034 1035-1105 1106-2703 2704-2777 2778-3746 3746-3817 3817-3887 3887-3955 3957-4995 4996-5065 5067-5135 5137-5209 5236-5301 5312-5381 5383-6931 6963-7033 7036-7104 7105-7785 7787-7859 7861-8037 8028-8708 8708-9491 9492-9559 9560-9910 9914-9983 9984-10280 69 960 71 1597 74 969 70 71 69 1044 69 69 73 66 70 1549 71 68 681 73 180 684 784 69 351 70 297 70 964 71 1598 74 969 72 71 69 1039 70 69 73 66 70 1549 71 69 681 73 177 681 784 68 351 70 297 H H H H H H H L H H H L L L L H L H H H H H H H H H H 0 0 0 0 0 -1 -1 -1 1 -5 1 1 26 10 1 31 2 0 1 1 -10 -1 0 0 3 0 -4 0 0 0 0 0 -1 -1 -1 1 0 1 1 26 10 1 31 2 0 1 1 -8 -1 0 0 3 0 -5 ND4 His tRNASer UCN Leu CUN ND5 ND6 Glu Cyt b Thr tRNAPro 10274-11641 10274-11645 11653-11722 11652-11721 11723-11789 11722-11788 11789-11860 11788-11859 11862-13670 11866-13668 13667-14191 13665-14189 14192-14261 14190-14259 14269-15409 14266-15406 15410-15483 15407-15481 15484-15553 15482-15551 1368 70 67 72 1809 525 70 1141 74 70 1372 70 67 72 1802 525 70 1141 75 70 H H H H H L L H H L 11 0 -1 1 -4 0 7 0 0 0 6 0 -1 6 -2 0 6 0 0 0 Control region 15554-16559 15552-16554 1006 1003 T incomplete stop codon Negative numbers indicate overlapping nucleotides a Numbers correspond to the nucleotides separating adjacent 3 genes underneath b KL and SC represent K leucostomum and S carinatus 205 32 Conserved nucleotides of every mitochondrial component The two mitochondrial genomes had high proportion of conserved nucleotide from tRNAPhe to control region and it got up to 911 all over the complete mitochondrial genomes Fig 3 In the two mitochondrial genomes of K leucostomum and S carinatus the tRNAHis and tRNALeu CUN -8- 210 215 had the same nucleotide and the lowest proportions of conserved nucleotides were found in the CR region and tRNAThr which was about 85 There were all above 86 conserved nucleotides in the 13 protein genes and the average proportion was 902 In the 22 tRNAs the proportions of conserved nucleotides all surpass 85 and the average proportion was944 The two rRNAs were so conserved that the conserved nucleotides were up to 939 957 in 16S rRNA and 12SrRNA respectively 105 100 95 90 85 80 75 Fig 3 The proportion of conserved nucleotide sites among every gene between the two mitochondrial genomes CR Control region 220 33 Synonymous and non-synonymous substitutions To gain insight into the selection pressure of protein coding genes in the two species analyzed we estimated the rates of synonymous KS silent mutation and non-synonymous KA amino-acid altering mutation substitutions for all 13 protein coding genes of the two mitochondrial genomes of K leucostomum and S carinatus and performed the KAKS test for positive selection of each protein coding Our analyses of 13 different protein coding genes 225 showed that the KAKS ratio was less than one KAKS ratio obtained in between 0008 and 0262 fig4 The highest KAKS ratio was obtained for the gene ATP8 0262 which was followed by ATP6 0156 The lowest KAKS ratio was obtained for the gene COX2 which was just 0008 030 025 020 015 010 005 000 ND1 ND2 COX1 COX2 ATP8 ATP6 COX3 ND3 ND4L ND4 ND5 ND6 Cytb Fig 4 The synonymous and non-synonymous substitution rates kaks estimated for K leucostomum and S carinatus 230 34 Phylogenetic position of the Kinosternidae within Testudines In this study the topologies of the ML and BI trees based on the 30 complete mitochondrial genomes were the same to each other Fig 5 Robustness of the trees was estimated using bootstrap proportions BPs in ML analysis and posterior probabilities PPs in Bayesian analysis 235 BI respectively The nodes all over the two trees all gained high support values From the -9- resultant trees the P subrufa Pelomedusidae and C fimbriata Chelidae comprised the Pleurodira clade which formed a sister group to the other assemblage of 26 cryptodires BP 100 PP 100 The Trionychidae group had a close relationship to all remaining cryptodires used in this study BP 100 PP 100 The Kinosternidae and Chelydridae clustered together to generate a 240 245 250 clade BP 71 and PP 100 The assemblage of Kinosternidae Chelydridae had a sister-taxon relationship to the Cheloniidae Testudinoidea clade Emydidae Platysternidae Geoemydidae Testudinidae with the BP 75 and PP 098 The Testudinoidea composed of two sister clades BP 78 and 099 one was Emydidae Platysternidae assemblage BP 100 and 100 the other was Geoemydidae Testudinidae assemblage BP 100 and 100 Fig 5 imum likelihood and BI analysis trees of the 28 turtles of 10 families with 2 crocodiles used as outgroup inferred from the 30 complete mitochondrial genomes The five nodes with black points are supported by the different BP values and PP values 1 BP 99 and PP 100 2 BP 78 and 099 3 BP 75 and PP 098 4 BP 71 and PP 100 5 BP 80 and PP 098 and other nods are all BP 100 and PP 100 4 Discussion 41 Genome features The gene arrangement and the base composition of the two mtDNA were very similar to each other which conformed to that of the consensus vertebrate type The two mitochondrial genomes all had high AT bias which is consistent with the high proportion of AT found in other turtles 255 and other vertebrates Sbisa et al 1997 Zhang et al 2009 One hypothesis that attempts to - 10 - explain this bias is that the DNA polymerase could use those bases in a more efficient way during mtDNA replication Clary and Wolstenholme 1985 The lower energetic cost to break the A-T links during mtDNA replication and transcription would generate AT bias on organisms Xia 1996 260 265 270 The light strand replication origin OL in many other vertebrates was also present in the mitochondrial genomes of K leucostomum and S carinatus Except the side-necked turtle Pelomedusidae Chelidae the nucleotides of OL sequences are highly conservative and their secondary structures are similar in all reported Cryptodiran turtle species Zardoya and Meyer 1998 Zhang and Nie 2007 All the OL structures have near the same stems and some different base variations in the loops table 4 fig 6 The common absence of OL may be an ancestral character of Pleurodiran turtles Wang et al 2011 Similar features have been reported in some other reptiles and chicken Macey 1997 Desjardins and Morais 1990 Seligmann and Krishnan 2006 Maybe this character could be used as a classified criterion between Pleurodira and Cryptodira Table 4 The comparion of the OL sequence regions of all the species chosen in this study Family Crocodylidae Pelomedusidae Chelidae Trionychidae Chelydridae Kinosternidae Cheloniidae Testudinidae Platysternidae Emydidae Geoemydinae Species Alligator mississippiensis Caiman crocodiles Pelomedusa subrufa Chelus fimbriata Palea steindachneri Pelodiscus sinensis Apalone ferox Macroclemys temminckii Chelydra serpentine Kinosternon leucostomum Sternotherus carinatus Chelonia mydas Lepidochelys olivacea Geochelone pardalis Testudo marginata Testudo kleinmanni Indotestudo elongata Indotestudo forstenii Platysternon megacephalum Chrysemys picta Trachemys scripta Sacalia quadriocellata Cyclemys atripons Mauremys mutica Mauremys reevesi Cuora amboinensis Cuora aurocapitata Cuora mouhotii Cuora flavomarginata Regions of the OL sequences TTTATCCC-----AAA TCTA-CCC----TAA--G--------------- TATTT A----CC-AC---AAA------ATAACTTATTA AATTTCCCGCTATAAAA--AAAAGCGGGAAAAC AATTTCCCGCTTTAAAAGTAAAAGCGGGAAAAC TTTTTCCCGCTATAAAA--AAAAGCGGGAAAAC CTTTTCCCGCTCTCTAA---AAAGCGGGAAAAC CTTTTCCCGCTCTCTAA---AAAGCGGGAAAAC CTTTTCCCGCTTTTTA-G--AGAGCGGGAAAAC CTTTTCCCGCTCT-TAA--AAAAGCGGGAAAAC CCTTTCCCGCTCTATAA---AAAGCGGGAAAAC TTTTTCCCGCTCTATAA--AGAAGCGGGAAAAC TTTTTCCCGCTTTA-A---CAGAGCGGGAAAAC CTTTTCCCGCTCTATAA---AAAGCGGGAAAAC CTTTTCCCGCTCTGTAA---AAAGCGGGAAAAC CTTTTCCCGCTCTATAA---AAAGCGGGAAAAC TCTTTCCCGCTCTATAA---AAAGCGGGAAAAC ATTTTCCCGCTCTCGGA--AAAAGCGGGAAAAC CCTTTCCCGCTCTCTAA---AAAGCGGGAAACC CCTTTCCCGCTCTCTAA---AAAGCGGGAAACC ATTTTCCCGCTCTGTAA---AAAGCGGGAAAAC CTTTTCCCGCTCTCTA-TAAAAAGCGGGAAAAC CTTTTCCCGCTCTCTAA---AAAGCGGGAAAAC CTTTTCCCGCTCTCTAA--AAAAGCGGGAAAAC TTTTTCCCGCTCTTTAA---AAAGCGGGAAAAC CTTTTCCCGCTCTCTAA--AAAAGCGGGAAAAC CTTTTCCCGCTCTCTAA---AAAGCGGGAAAAC TTTTTCCCGCTCTCTAA---AAAGCGGGAAAAC Note the sequences with gray underpainting are highly conservative sequences of the OL stems The 31bp intergenic regions of K leucostomum and S carinatus located between the COX1 and tRNASer UCN genes strongly resembled that of K flavescens partial mtDNA Parham et al 275 2006 but the regions of other turtles found so far overlapped by about 9 bp in the majority There are 5 bp overlapped sequence in this regions of the three trionychids Caiman crocodiles - 11 - Pelomedusa subrufa and Trachemys scripta analyzed this study however Crocodylus siamensis and Chelus fimbriata have 3 and 2 bp spacers between the COX1 and tRNASer UCN genes respectively table 5 We further found that K leucostomum and S carinatus all had the 10bp 280 spacer between tRNACys and tRNATyr 11 bp in K flavescens but no spacers appear in Cheloniidae Testudinidae Geoemydinae and Emydidae table 5 Chelydra serpentine and Macroclemys temminckii have 7 and 4 bp spacer respectively If these sequences were the common characters of Kinosternidae they could be used as marks to identify the kinosternidae So to make sure whether it is true more mitochondrial genomes of Kinosternidae are needed to gain 285 Fig 6 The stem-loop OL secondary structure of Pleurodiran turtles chosen in this study 1 Apalone ferox 2 Palea steindachneri 3 Pelodiscus sinensis 4 Cuora amboinensis 5 Cyclemys atripons 6 Mauremys mutica 7 Cuora flavomarginata 8 Cuora aurocapitata 9 Cuora mouhotii 10 290 295 300 Sacalia quadriocellata 11 Mauremys reevesi 12 Macroclemys temminckii 13 Chelydra serpentine 14 Trachemys scripta 15 Chrysemys picta 16 Lepidochelys olivacea 17 Chelonia mydas 18 Testudo kleinmanni 19 Geochelone pardalis 20 Indotestudo elongata 21 Indotestudo forstenii 22 Testudo marginata 23 Platysternon megacephalum 24 Kinosternon leucostomum 25 Sternotherus carinatus The previous researches show that there extensively exists a single extra base A at about position 174 of ND3 gene in many turtle taxa This extra base is found to be transcribed and is present in mature mRNA and the unique features of the sequence at that site induces a compensatory frameshift during translation which shows certain mitochondrial translation systems have the ability to tolerate frameshift insertions using programmed translational frameshifting Russell and Beckenbach 2008 We noted that K leucostomum and S carinatus lacked this extra nucleotide as well as absent in P sinensis K flavescens C aurocapitata and Cyclemys atripons Peng et al 2005 Parham et al 2006 Zhang et al 2008a 2008b However - 12 - we found a single nucleotide insertionA at position 1012bpin the ND4 gene of S carinatus The frameshift insertion of one base in animal mitochondrial genomes are widespread anomalies 305 that are evidently tolerated by the mitochondrial translation systems of some animal species and all require a wobble pairing in the codon at the insertion sites Russell and Beckenbach 2008 Table 5 The base numbers of intergenic regions between tRNACys and tRNATyr genes and between COX1 and tRNASer UCN of all species chosen in this study Family Species The intergenic regions between Cys Tyr The intergenic regions between COX1 and tRNASer UCN Crocodylidae Pelomedusidae Chelidae Trionychidae Kinosternidae Chelydridae Cheloniidae Testudinidae Geoemydinae Platysternidae Emydidae Crocodylus siamensis Caiman crocodilus Pelomedusa subrufa Chelus fimbriata Palea steindachneri Pelodiscus sinensis Apalone ferox Kinosternon leucostomum Sternotherus carinatus Chelydra serpentina Macroclemys temminckii Lepidochelys olivacea Chelonia mydas Indotestudo elongata Testudo marginata Geochelone pardalis Testudo kleinmanni Indotestudo forstenii Mauremys mutica Mauremys reevesi Cuora mouhotii Cuora aurocapitata Cuora flavomarginata Cuora amboinensis Sacalia quadriocellata Cyclemys atripons Platysternon megacephalum Trachemys scripta Chrysemys picta 0 3 0 -5 1 -5 1 2 0 -5 2 -5 0 -5 10 31 10 31 7 -9 4 -9 0 -10 0 -9 0 -9 0 -9 0 -9 0 -9 0 -9 0 -9 0 -9 0 -9 0 -9 0 -9 0 -9 0 -9 0 -9 0 -9 0 -5 0 -9 310 315 320 Note the positive numbers mean there are base spacers between the two genes and the negative numbers mean there are overlapped bases between the two genes Analyses of synonymous and non-synonymous substitution rates help to understand the level of evolutionary pressures acting on different genes which in turn helps in studies concerning the mechanisms of genome evolution Yang and Nielsen 2000 The kaks ratio is a simple measure of selective pressures acting on gene that indicates neutral mutation KAKS 1 negative or purifying selection KAKS 1 and positive or diversifying selection kaks 1 Theoretically a KAKS 1 indicates that the rate of evolution is higher than the neutral rate Conversely a gene with KAKS 1 has a rate of evolution less than the neutral rate Yang and Nielsen 2000 Our result was indicative of a strong purifying selection which removes deleterious mutations on all the protein coding genes 42 Phylogenetic relationships In our analyses the BI and ML trees had the same topologies but there were many differences comparing with those mentioned above Fig 1 Fig 5 In the ML and BI analyses the Kinosternidae and Chelydridae formed a sister group to assemble an independent clade BP 71 325 and PP 100 which was similar to the results of Cervelli et al 2003 and Barley et al 2010 but - 13 - the clade of Kinosternidae Chelydridae was sister group to the Cheloniidae Testudinoidea with well-supported values BP 100 and PP 100 in our analyses However the Kinosternidae Chelydridae clade and Cheloniidae clade cluster together to be a monophyly in the results of Cervelli et al 2003 and Barley et al 2010 So it was indicated that the Kinosternidae and 330 335 340 345 350 355 360 365 370 375 Chelydridae probably had one common ancestor and were not close to the Cheloniidae relatively according to the BI and ML analyses this study whereas it was just opposite to the result of Krenz et al 2005 which implicated Cheloniidae and Chelydridae made up of an independent clade and had close relationship to Kinosternidae In our analyses the Cheloniidae had a close relationship to the Testudinoidea with the strong support of BP 75 and PP 098 and they formed a monophyly of Cheloniidae Testudinoidea which supported the hypothesis of Parham et al 2006 Based on the ML and BI trees the Testudinoidea composed of two sister clades BP 78 and 099 one was Emydidae Platysternidae assemblage BP 100 and 100 the other was Geoemydidae Testudinidae assemblage BP 100 and 100 which were identical to the analysis of Parham et al 2006 and Barley et al 2010 Above all our result was highly different from the hypotheses of Shaffer et al 1997 and Brinkman et al 1999 which all used the data of morphology The Asian box turtles of Geoemydidae all have a single hinge between the pectoral and abdominal scute which allows the plastron to close the shell completely or only slightly which is different from that the mud turtles Kinosternidae Kinosternon have two transverse hinges Zhang 2008a Harding 1996 Based on the phylogenetic analyses the Cuora group a group of Asian box turtles had a sister-taxon relationship to the Mauremys group BP 100 PP 100 but was not close to the Kinosternidae So it was obvious that the Cuora group didnt have close ties of consanguinity with the Kinosternidae indicating that the hinges might not be derived from one common ancestor Acknowledgments This research was supported by the National Natural Science Foundation of China NSFC No 30970351 The Specialized Research Fund for the Doctoral Program of Higher Education of China grant number 20080370001 and the Key Laboratory of Biotic Environment and Ecological Safety of Anhui province References [1] Barley AJ Spinks PQ Thomson RC Shaffer HB Fourteen nuclear genes provide phylogenetic resolution for difcult nodes in the turtle tree of life[J] Mol Phylogenet Evol2010 55 1189-1194 [2] Brennicke A and Clayton DA Nucleotide assignment of alkali-sensitive sites in mouse mitochondrial DNA Biol Chem 1981 256 10613-10617 [3] Brinkman DB and Wu XC The skull of Ordosemys an Early Cretaceous turtle from Inner Mongolia Peoples Republic of China and the interrela