A Complete Sequence of the Mitochondrial Genome of the Western Lowland Gorilla

Xiufeng Xu and U&u- Arnason Division of Evolutionary Molecular Systematics, University of Lund, Sweden

The complete mitochondrial DNA (mtDNA) molecule of the gorilla was sequenced. The entire sequence, 16,412 nucleotides, was determined by analysis of natural (not polymerase chain reaction) restriction fragments covering the whole molecule. The sequence was established from one individual and thus nonchimeric. After comparison with the CO11 gene of gorilla specimens with known geographical origin, the sequence was identified as charac- teristic of the Western lowland gorilla, Gorilla gorilla gorilla. With the exception of the NADH2 gene, all genes have a methionine start codon. The inferred start codon of NADH2 is ATT (isoleucine). The COIII, NADH4, and cytochrome b genes are not terminated by a stop codon triplet, and the CO1 gene is probably terminated by an AAA triplet rather than by a regular stop codon. The great majority of genie sequences (rRNAS, peptide-coding Downloaded from https://academic.oup.com/mbe/article/13/5/691/1083040 by guest on 28 September 2021 genes, tRNAs) of the complete mtDNAs of Gorilla, Pun, and Homo show a greater similarity between Pun and Homo than between either of these genera to Gorilla. The analysis of the peptide-coding genes suggests that relative to comparison between Homo and Pun a certain degree of transition saturation has taken place in codon position 3 in comparisons between Gorilla to either Homo or Pun.

Introduction Recent studies on mitochondrial DNA (mtDNA) of Comparisons among and within closely related spe- the gorilla (Ruvolo et al. 1994) have revealed striking cies are of particular value for the understanding of mo- molecular differences among the three acknowledged lecular evolutionary dynamics of mtDNAs that are still subspecies, the Western and Eastern lowland gorillas unsaturated with respect to nucleotide (nt) substitution, and the mountain gorilla. The molecular differences and the issue of the HomolPanlGoriZZa relationship has have suggested evolutionary divergences of 1.5 million been one of the most contentious in the history of mo- years or more between different subspecies. Relative to lecular systematics. In the present paper we establish, the limited morphological distinctions among the three on the basis of cloned natural (not polymerase chain subspecies these evolutionary divergences are notewor- reaction [PCR]) restriction fragments of mtDNA from thy. The results, like similar findings in the common one individual, the degree of mtDNA difference among chimpanzee (Morin et al. 1994), demonstrate the diffi- the three species and detail these differences with re- culties in distinguishing and dating evolutionary diver- spect to individual mitochondrial genes. We address also gences only on the basis of traditional morphology. the problems associated with the use of uncloned PCR Given the marked subspecies distinctions within products for sequencing heteroplasmic mtDNA regions. both the chimpanzee and the gorilla, one should be aware of the potential complications conjunct with the Materials and Methods use of chimeric sequences in molecular comparisons. As DNA, enriched with respect to mtDNA, was iso- pointed out by Arnason, Xu, and Gullberg (1996) chi- lated from frozen kidney tissue of two gorilla specimens merit sequences have been reported for both Homo and (YN90-225 female, YN90-47 male) following the pro- other hominoids (Anderson et al. 198 1; Horai et al. cedure used by Arnason, Gullberg, and Widegren 1995), and it is evident that the use of such data may (1991). The samples were generously provided by Dr. complicate the picture of not only phylogeny reconstruc- Harold M. McClure, Yerkes Regional Primate Research tion but also population-level studies and estimates of Center, Atlanta, Georgia. divergence times. In the present study we describe the The mtDNA of the female specimen was sequenced complete mtDNA sequence of a hominoid, the Western in its entirety. The sequencing was based on 28 unique lowland gorilla, Gorilla gorilla gorilla, that, like the clones (BcZI, BZnI, SpeI, XbaI), most of which were rep- common chimpanzee reported previously (Arnason, Xu, resented several times in the collection. All regions of and Gullberg 1996) has been characterized at the sub- the molecule were represented by a minimum of two specific level. A complete mtDNA sequence of the go- clones. Sequencing was performed manually applying rilla has been presented by Horai et al. (1995). That the dideoxy termination technique (Sanger 1981) with sequence, however, includes previously published data 35SdATP, using both universal and numerous specific se- of other authors. Therefore, some of the accumulated quencing primers. In the case of the male specimen the data may represent different subspecies. complete control region was sequenced after PCR am- plification and subsequent cloning in M13. The PCR Key words: Gorilla, Homo, Pan, hominoids, mitochondrial DNA, clones were identical except for a variable number of molecular relationships, control region. Cs (L-strand) in two parts of the control region. Address for correspondence and reprints: Ulfur Amason, Division of Evolutionary and Molecular Systematics, University of Lund, Sill- The accession number of the complete mtDNA se- vegatan 29, S-233 62 Lund, Sweden. E-mail: ulfur.arnason@ quence of the gorilla is X93347 and that of the control gen.lu.se. region of the male specimen X93348. Users of the se-

Mol. Biol. Evol. 13(5):691498. 1996 quences are kindly requested to refer to the present pa- 0 1996 by the Society for Molecular Biology and Evolution. ISSN: 0737-4038 per and not only to the accession numbers.

691 692 Xu and Amason

Results basis of these overall distance comparisons we have identified the complete molecule now presented as that The length of the reported gorilla mtDNA molecule of the Western lowland gorilla, Gorilla gorilla gorilla. is 16,412 nt and its nt composition (L-strand) in percent is A = 30.9, C = 30.7, G = 13.1, T = 25.3. The com- Comparison with a Previously Reported Complete position of the control region in percent is A = 29.5, C mtDNA of the Gorilla = 32.7, G = 15.0, T = 22.8. The sequence of the Western lowland gorilla pres- Outside the control region the organization of the ently determined was compared with that described by molecule conforms to other complete mammalian Horai et al. (1995). The two sequences differ at two mtDNAs that have been reported, including Homo (An- positions in the CO11 gene. Based on this value and the derson et al. 198 1; Horai et al. 1995; Arnason, Xu, and data presented by Ruvolo et al. (1994), we have con- Gullberg 1996), mouse (Bibb et al. 1981), cow (Ander- cluded that the sequence reported by Horai et al. (1995) son et al. 1982), rat (Gadaleta et al. 1989), fin whale is also representative of the Western lowland gorilla. It

(Arnason, Gullberg, and Widegren 1991), harbor seal should be observed, however, that the sequence pre-Downloaded from https://academic.oup.com/mbe/article/13/5/691/1083040 by guest on 28 September 2021 (Arnason and Johnsson 1992), grey seal (Arnason et al. sented by Horai et al. (1995) is chimeric. While the larg- 1993), blue whale (Arnason and Gullberg 1993), opos- est part of the molecule has been sequenced by Horai sum (Janke et al. 1994), horse (Xu and Arnason 1994), et al. (1995), the 12s rRNA gene described by Hixson hedgehog (Krettek, Gullberg, and Arnason 1995), chim- and Brown (1986), as well as the control region reported panzee (Horai et al. 1995; Arnason, Xu, and Gullberg by Foran, Hixson, and Brown (1988), have been incor- 1996), and gorilla and orangutan (Horai et al. 1995). porated into that complete sequence. The extents of the different features of the complete The control regions of the two complete mtDNAs mtDNA molecule were determined by analogy with oth- are shown aligned in figure 1. Horai et al. (1995) hy- er complete hominoid mtDNAs sequenced by our group, pothesized that, relative to other hominoid sequences, and in the case of the tRNA genes by comparison with there is a large deletion in the control region of the go- the tRNA survey of Kumazawa and Nishida (1993). rilla mtDNA. The account of Horai et al. (1995) is com- Like the human and chimpanzee sequences (Arnason, plicated by the fact that they have not included a 32-nt Xu, and Gullberg 1996), the gorilla has ATT (isoleucine) portion of the sequence reported by Foran, Hixson, and as start codon of the NADH2 gene. Other peptide-cod- Brown (1988). Although its location was not specified, ing genes have a methionine start codon. The CO1 gene we have located the “deletion” to positions 139-169 of of the gorilla is terminated by an AAA codon contrary the control region. After complementing the control re- to Homo and chimpanzee, which both have a regular gion included in Horai et al. (1995), the two control stop codon, AGA. The AAA identity of the codon in regions differ by 45 transitions, 4 transversions, and 13 the gorilla is consistent with the sequence reported by indels (insertions or deletions). Garner and Ryder Horai et al. (1995). The COIII, NADH4, and cyto- (1992), in a survey of the control region of the Gorilla chrome (Cyt) b genes of the gorilla are not terminated (Western and Eastern lowland gorillas, mountain goril- by a complete stop codon. This is consistent with Homo la), assign the sequence reported by Foran, Hixson, and and the chimpanzee. Brown (1988) to the Westen lowland gorilla. The dis- similarity between the sequences shown aligned in fig- The Assignment of the Gorilla Specimen to ure 1 is surprisingly great considering the fact that the Subspecies Gorilla gorilla gorilla two (female and male) complete control regions se- In a recent phylogenetic study Ruvolo et al. (1994) quenced by us differed by only a single transition. If analyzed the complete CO11 gene of several hominoids both sequences of figure 1 are representative for the including Gorilla. The gorilla specimens, which were Western lowland gorilla, it is evident that the mtDNA known with respect to their geographical origin, repre- control region of this subspecies is highly polymorphic. sented the Western and Eastern lowland gorillas and the Outside the control region there are 49 differences mountain gorilla. Comparison between the CO11 gene of between the two complete mtDNA sequences, of which the complete molecule presently reported and the se- 40 are transitions, 7 transversions, and 2 indels. Seven quences described by Ruvolo et al. (1994) shows that of the differences (five transitions, one transversion, and the present sequence and the Western lowland sequence one indel) occur in the 12s rRNA gene incorporated Ggo4 (Ruvolo et al. 1994) differ by just one substitu- from Hixson and Brown (1986). The indel is in a region tion, viz. a G/C nonsynonymous transversion in position where we identify a run of five Cs (L-strand) as com- 5 17 of the CO11 gene. The substitution is in codon po- pared with four Cs in the sequence presented by Horai sition 1. There are six differences, all transitions, be- et al. (1995). The other indel observed occurs in a run tween the present sequence and the Western lowland of six As in our sequence, as compared with five As in sequence Ggo3 reported by Ruvolo et al. (1994). Two the sequence reported by Horai et al. ( 1995). The run of the differences occur in the second and four in the of As is located in a nongenic region between tRNA- third codon position. In position 5 17, the sequence pres- Ser(UCN) and tRNA-Asp. There are five transitional ently described is identical with Ggo3. There are 23 dif- differences in the 16s rRNA gene of the two specimens. ferences between the presently reported sequence and In the peptide-coding genes the two sequences differ at that of the Eastern lowland gorilla (GgoS), and 21 dif- 36 positions, 19 of which are nonsynonymous nt sub- ferences relative to the mountain gorilla (Ggo6). On the stitutions (table 1). The number of differences in first mtDNA of Western Lowland Gorilla 693

. I ...... 1 TTCTTTCATGGGGAGACATTGGGTACCACCC~GTATTGGCT~CCCATC~T~TTATCATGTATGTCGTGCATTCCTGCCAGACACCATG~T~ 100 11111111111111111111111111111 III IllIll 11111111111 ltlltllllllllllllllItllll IIIIIIIIIIIIIIIIIIII 1 TTCTTTCATGGGGAGACAAATTTGGGTACTACC--AGTATTAGCT~CCCATC-AT~TTATCATGTATGTCGTGCATTACTGCCAGACACCATG~T~ 97 ...... 101 TGCACAGCACCAC-AAATGTCCGATCACCTGTAACACACATAC~CCCCCCCCTTCCCCCCCCCCTCCT-CCACCC~TGG~TATC~CT~TCCATTCC~ 198 II lItI IIII III I II IIIIIIIIIt IIII IIIIIII IIIIIIIIIIIIIIIIII I I IIIII IIIIII IIIIIII ItIt III 98 TGTACAGTACCATAAAACGCCCAATCACCTGTAGCACCTGTAGCACATAC~CCCCCC~CCCCCCCCCCCG~CC~CGG~TACC~CT~CCCATCCCT 197 ...... 199 CAT~GTACATAGCACATAAAGTCATTTATCATTTATCGTACATAGCACATTCTAGTT~TCATCCTTGCCCCCACG~ATGCCCCCCCTCAGATAGGAGTCCC 298 II 111111111111 Illllll IIIIIIIIIt t IIIII IIIII I IIIIII III1 IIt I 111111111111111111111111 III III1 198 CACAAAAAGTACATAACACATAAGATCATCATTTATCGCACATAGCACATCCCAGTT~TCACCCTCGTCCCCACGGATGCCCCCCCTCAGATGGG~TCCC 297 ...... 299 TTGAACACCATCCTCCGTGATCAATATCCCCGCAC~GAGTGCTACTCTCCTCGCTCCGGGCCCAT~CACTTGGGGGTAGCT~TG~CTGTATC 398 IIIIIIIIIIIIIIIIIIIIIIIIIIIlllllllllttIIIIII IIIIIItIIIIIIItIIIttlllllllllltlllllllt III IIIII IIII II 298 TTGAACACCATCCTCCGTGATCAATATCCCCGCAC~GAGTGC-ACTCTCCTCGCTCCGGGCCCAT~CACTTGGGGGTAGCT-~GTG~CTGTATC 395

. . I ...... Downloaded from https://academic.oup.com/mbe/article/13/5/691/1083040 by guest on 28 September 2021 399 CGGCATCTGGTTCCTACTTCAGGGTCATAACACCTAAAGCGCTTCACACGTTCCCCTT~T~GACATCACGATGGATCACAGGTCTATCACCCTATTA 498 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIltlllllllIIItIIIIIIIIIIIIItIIItIIIIIIIIIlllltltllllltllllllllltllllllll 396 CGGCATCTGGTTCCTACTTCAGGGTCATAACACCTAAAGCGCTTCACACGTTCCCCTT~T~GACATCACGATGGATCACAGGTCTATCACCCTATTA 495 ...... 499 ACCACTCACGGGAGCTCTCCATGCATTTGGTATTTTTCGTCGGGGGGTGTGCACGCGATAGCATTGCG~CGCTGG~CCGGAGCACCACATGTCGCAG 598 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIItI II IIII ItIIIIltlllllllIItIIIIIIII 496 ACCACTCACGGGAGCTCTCCATGCATTTGGTATTTTTCGTCGGGGGGTGTGCACGCGATAG -----TGAGACGCTGGAACCGGAGCACCACATGTCGCAG 590 ...... 599 TATCTGTCTTTGATTCCTGCCCCATACCATTATTTATCGCACCTACGTTCAATATTACAGCCGAGCGCAGCATGCTTTATGATGCTAATTAATTCATGCT 698 IIIIIIIIIIIIIIIt II IIIIIII I I llllllltl1ltllllltlIIIItIIIIIIIIIItIII II II III II IIII IIIIIIIIItI 591 TATCTGTCTTTGATTCCTGCCCCATGCTACCATTTATCGCACCTACGTTC~TATTACAGCCGAGCGCACAGTG-TTCATGGTGTT~TT~TTCATGCT 689 ...... 699 TGTTGGACATAAAACAACCAGGTGGACGTGAACACAACCAC AAAAAATCCAACCCCCCCCCCCCCGCACTAGAAAACAG 798 ItIIIIIIIIIIIIIIIIItlllll1lll1lllltlllltllllllltlllllllllIltll ItIIIIIIII IIIIIIIIIIIt IIIttIIIIIIII 690 TGTTGGACATAAAACAACCAGGTGGACGTGAACACAACCA 786 . . I ...... 799 CCCCACCAAACTCCAAATTTCATCTTTTGGCGGTATGCATTTTT~CAGTTACCCCTC~CT~CATAGCACC~CCCCACCAGTAC~-CCCCACCCGC 897 IIIIIIIIIIIIIItIIIIlllllllltllllllllttI IIIIIIIIIt tIIIIItIIItlllltll II IIIIIIIIIIIII IIII ItIll 787 CCCCACCAAACTCCAAATTTCATCTTTTGGCGGTATGCACTTTT~CAGTCACCCCTC~CT~CATAG--TCAG-CCCACCAGTAC~CCCCCGCCCGC 883 ...... 898 CCTAGCAACACACACTGCTGCTGATCCTATACCCCGAATTACA 964 IIIItIIIIIIIIIIIIIIIIIIItIIIIIIIIIIIIIIlllttlllllllllllllIItIIIIIII 884 CCTAGCAACACACACTGCTGCTGATCCTATACCCCGAATTA 950

FIG. l.-Alignment between the control region of the presently described complete sequence (female) of the Western lowland gorilla, above, and that reported by Foran, Hixson, and Brown (1988) and incorporated in Horai et al. (1995), below. The underlined part of the control region sequenced by Foran, Hixson, and Brown (1988) was not included in the complete sequence reported by Horai et al. (1995). Each line has 100 characters. Number 1 has been allocated to the 5’-end nucleotide of the control region (L-strand). Indel (insertion or deletion) differences between the two sequences are shown by (-). The control region of another specimen (male) of the Western lowland gorilla differed from the female sequence by a single C/T transition in position 79. and second codon positions, and hence also the number gence estimates proposed. Comparison including all of amino acid (aa) differences between the two sequenc- peptide-coding genes of the two seals and Homo and es, is unexpectedly large relative to the number of nt Pan shows that the ratios for nt substitution according differences in codon position 3. to codon position are essentially the same for the two pairwise comparisons. The transitionkransversion ratios Comparison Among the mtDNAs of Homo, Pun, and for the third codon position of all peptide-coding genes Gorilla are also similar for the two seals and Homo and Pan. We have reported previously the mtDNAs of three We (Arnason, Xu, and Gullberg 1996) have, therefore, pairs of closely related species, namely the harbor and deduced on the basis of these findings that the mtDNAs grey seals, the fin and blue whales, and Homo and the of Homo and Pan are still largely unsaturated with re- common chimpanzee. The lower limit for the divergence spect to nt substitution. of the three seals is at about 2.7 MYA and they are, Between Gorilla and Homo the difference outside therefore, much more closely related than are Homo and the control region is lOS%, counting each gap as a sin- the common chimpanzee, which according to the same gle mutation irrespective of its length. The correspond- calculations diverged 6.1 MYA (Arnason et al. 1996). ing figure for Gorilla/Pan is lO.O%, and that for Homo/ The molecular difference between the two whales, Pan 8.5%. The following account provides details of a which occasionally produce viable offspring (Arnason comparison among the 12s and 16s rRNA genes, the et al. 1991; Spilliaert et al. 1991), is slightly less than tRNA genes, and the peptide-coding genes of the three that between Homo and Pan. The harbor and grey seals species. The comparisons are based on the three com- are the most closely related species-pair for which the plete mtDNAs sequenced in our laboratory. Details of entire mtDNA molecule has been sequenced and diver- pairwise differences between the 12s and 16s rRNA 694 Xu and Arnason

Table 1 Table 2 Nucleotide Differences According to Codon Position (1, 2, Differences Among the 12s and 16s rRNA Genes and the 3) Between the 13 mtDNA Peptide-Coding Genes of the tRNA Genes of Homo (H), Pan (P), and Gorilla (G) Presently Described Sequence and that Reported by H/G P/G Horai et al. (1995) GENE Ti TV Gaps Ti TV Gaps Ti TV Gaps AMINO ACID 12s rRNA . . 37 3 3 36 4 2 36 5 2 1 2 3 RE- 16SrRNA.. 72 9 2 108 11 3 83 14 2 PLACE- tRNA ...... 59 4 - 84 6 - 82 8 - GENE Ti” Tvb Ti TV Ti TV MENTS Total 168 16 5 228 21 5 201 27 4 --- NADHl . . . 1 3 1 189 (4.7%) 254 (6.3%) 232 (5.8%) NADH2 . . . 1 1 1 2 co1 ...... 1 1 Non.-The length of the 12s rRNA genes is 954 nt in Homo, 956 nt in 1 1 1 co11 ...... Pan, and 950 nt in Gorilla. The corresponding lengths of the 16s rRNA gene

ATPase8 . . 1 are 1,559, 1,558, and 1,558 nt, respectively. The combined length of the tRNADownloaded from https://academic.oup.com/mbe/article/13/5/691/1083040 by guest on 28 September 2021 ATPase6 genes is 1,508 nt in all species. Ti = transitions; TV = transversions. co111 . . . : : NADH3 . . . 1 NADH4L . . 1 I the evolution of individual genes may differ consider- NADH4 . . . 1 1 1 1 2 ably from the means provided by all mitochondrial pep- NADHS . . . 1 1 2 NADH6 . . . 1 1 tide-coding genes. Cyt b . . . . . 4 1 7 1 5 As mentioned above, we have documented great Total 8 7 3 15 2 19 similarity in the pattern of molecular difference between the closely related harbor and grey seals. Setting the seal a Ti = transitions. divergence at 2.7 MYA (lower limit) suggests that h TV = tranversions Homo and Pun diverged 6.1 MYA. Using the same ap- proach for dating, we have proposed that the separation genes and the concatenated tRNA genes of Homo, Pun, between the Gorilla lineage and that leading to Homo/ and Gorilla are shown in table 2. It is notable that the Pun took place 8.4 MYA (Arnason et al. 1996). These 12s rRNA genes of the three species are about equally datings were based on nonsynonymous substitutions. different from each other. In the two other sets of se- When the differences between the peptide-coding genes quences there are considerably greater differences be- of Homo and Pun are compared with those between Go- tween Gorilla and either Homo and Pun than between rilla and either Homo or Pun, the results suggest a cer- the two latter species, consistent with the accepted phy- tain degree of third codon position transition saturation logenetic relationships among the three species based on in the comparisons including Gorilla. This is evident in complete mtDNAs (Horai et al. 1995; Arnason et al. the Ti/Tv ratio for codon position 3, which in the Homo/ 1996). Pun comparison is 14.2. For Homo/Gorilla the corre- Pairwise differences between the peptide-coding sponding ratio is 7.3, and for Pun/Gorilla it is 7.5. In genes of the three species are shown in table 3. For each the Homo/Pun comparison the ratio for conservative nt gene the table shows percent total difference, percent substitutions between third and second codon positions conservative nt changes (Irwin, Kocher, and Wilson is 0.8. The corresponding ratio for HomolGoriZZu is 1.3, 199 1), and percent aa difference. Conservative nt and that for Pun/Gorilla 1.4. The values show that dis- changes include all nonsynonymous substitutions in co- tance values based on total nt differences should be don position 1, all substitutions in codon position 2, and taken with caution even when dating divergences as re- transversions in codon position 3. Conservative nt cent as that between Gorilla and Pun/Homo. changes accumulate in reasonably clocklike manner and findings based on these substitutions have generally Analyses of PCR-Amplified Regions Containing been in good agreement with accepted phylogenies (Ir- Runs of Cs (Gs) win, Kocher, and Wilson 1991). With respect to the The present sequencing of the mtDNA of the go- mean values for both conservative nt changes and aa rilla, as well as our analyses of the common chimpanzee difference, the Gorilla is equidistant from both Homo (Arnason, Xu, and Gullberg 1996), have identified some and Pun and with respect to the values for total nt dif- sequence differences relative to sequences previously ference, the Gorilla/Pun difference (11.4%) is just published. The differences have been particularly pro- slightly lower than that for Gorilla/Homo (11.8%). It nounced in regions, such as control region of hominoid should be observed, however, that although the com- mtDNAs and parts of the 12s rRNA gene, which in- bined mean values for all genes show a consistent pat- clude runs of Cs (L-strand) and which may be difficult tern, individual genes may deviate from this pattern, to resolve when the H-strand is used as a template. showing greater difference between Homo and Pun than In the present study we examined the consistency between either of these species and Gorilla. Nucleotide of the number of Cs in two parts of the control region substitutions between each peptide-coding mtDNA gene of the gorilla. The regions correspond to positions 143- of the three species were examined further for type of 162 and 772-784, respectively, of the female sequence substitution (transition, transversion) and codon position shown in figure 1. The use of the G rich H-strand as (table 4). The results, like those of table 3, show that template does not allow resolution of these regions, mtDNA of Western Lowland Gorilla 695

Table 3 Differences in Percent Among mtDNA Peptide-Coding Genes of Homo (H), Pan (P), and Gorilla (G)

CONSERVATIVE NT TOTAL NT DIFFERENCE DIFFERENCE AMINO ACID DIFFERENCE

GENE LENGTH H/P WG P/G H/P H/G P/G H/P I-I/G P/G

NADHl . . 957 9.2 10.6 11.6 2.0 2.5 2.7 5.0 4.7 5.3 NADH2 . . 1,044 9.6 12.6 11.2 1.8 3.6 3.7 4.0 6.3 6.6 co1 ...... 1,542 8.8 10.7 9.9 0.8 1.7 1.6 1.2 1.8 1.6 co11 . . . . . 684 9.5 12.0 10.7 1.2 2.2 1.6 2.2 3.1 1.8 ATPase8 . . 207 8.7 11.6 11.1 2.4 5.3 6.8 5.8 11.6 14.5 ATPase6 . . 681 8.5 11.5 12.8 2.1 2.9 4.1 4.8 6.2 8.8 co111 . . . . 783 9.6 11.1 11.6 1.4 2.2 2.0 2.7 3.4 3.1 NADH3 . . 345 11.9 12.5 11.3 2.9 2.6 3.2 6.1 7.0 7.8

NADH4L . . 297 7.4 8.8 8.1 1.0 1.4 1.0 1.0 2.0 1.0 Downloaded from https://academic.oup.com/mbe/article/13/5/691/1083040 by guest on 28 September 2021 NADH4 . . 1,377 9.7 11.3 12.1 1.8 2.5 3.1 4.6 5.0 6.5 NADHS . . 1,812 10.6 14.1 12.4 3.4 5.1 4.5 6.6 9.9 9.4 NADH6 . . 525 9.7 9.1 11.2 2.1 1.7 2.3 4.6 3.4 4.0 Cyt b . . . . . 1,140 11.7 12.9 11.8 3.4 4.7 3.4 7.1 8.7 6.6 Total 11,394 9.8 11.8 11.4 2.1 3.1 3.1 4.4 5.7 5.8

NOTE.--Lengths are in nucleotides. Conservative nucleotide differences (Irwin, Kocher, and Wilson 1991) comprise all substitutions in codon position 1 except those involving leucine transitions, all substitutions in codon position 2, and transversions in codon position 3. Values below each column show the mean values for the combined length of all genes, not the arithmetic mean for the 13 genes. which, therefore, were sequenced using the L-strand as values for the difference between Gorilla and Homo and template. In position 143-162 the two natural (not PCR) Gorilla and Pan. This is particularly noteworthy consid- clones of the female had 8C-2T- 1OC. Of six PCR clones ering the deviations that may occur among the values of the male four were identical to the natural (not PCR) for single genes. The findings show that due to potential clones of the female, whereas two of the male clones fluctuations in short sequences the results of phyloge- had 8C-2T-11C. One of the latter clones with 8C-2T- netic analyses or population studies based on limited 1 IC was PCR amplified and subcloned. Of eight clones sequence data should be interpreted with caution. sequenced six had 8C-2T- 1 lC, whereas the remaining The dissection of nucleotide substitutions accord- two had 7C-2T-11C. In position 77 l-783 of the control ing to codon position (table 4) provided details of the region the two natural (female) clones had 13C. Also differences among all peptide-coding genes of Homo, this region of the male specimen was PCR amplified and Pan, and Gorilla. The data, in conjunction with the pair- cloned. Of six clones analyzed, two had 1 lC, three 13C, wise comparisons of the harbor and grey seals and the and one 15C. The clone with 15C was PCR amplified fin and blue whales, have made it possible to establish and subcloned. Of 24 clones sequenced 1 had 9C, 1 the substitution rate according to codon position and IOC, 3 12C, 2 13C, 4 14C, 11 15C, and 2 16C. type of substitution (transition, transversion), in all pep- tide-coding genes of mtDNA. This rate, as expressed by Discussion the ratios for total substitution and conservative nucle- otide substitution, respectively, has been shown to be Phylogenetic analyses of complete hominoid reasonably consistent for all pairwise comparisons car- mtDNA molecules (Horai et al. 1995; Arnason, Xu, and ried out so far. It should be noted that the Ti/Tv ratio Gullberg 1996), have identified a sister-group relation- for codon position 3 is considerably lower for the com- ship between Homo and Pan to the exclusion of Gorilla. parisons GoriZZalPan (7.5) GoriZZalHomo (7.3) The present results are in accord with those findings. and than for Homo/Pan (14.2). The results suggest an increased They are also in accord with a split between Homo/Pan and Gorilla of 8.4 MYA, given that Homo and Pan di- degree of transition saturation in the comparisons in- verged 6.1 MYA (Arnason et al. 1996). As demonstrated volving Gorilla. by the total peptide-coding gene data of table 3, Gorilla Our approach for determining the sequence of the is equidistant both to Homo and Pan, although the in- mtDNA of the Gorilla differs technically from that ap- dividual gene values may deviate from the pattern of all plied by Horai et al. (1995). In our case an enriched genes combined. Apart from the NADH6 gene, which mtDNA fraction was isolated from solid tissue before is located on the opposite strand relative to the remain- restriction digestion and cloning, whereas Horai et al. ing genes, the values for the Cyt b gene frequently used (1995) used PCR amplification. We have made partic- in phylogenetic analyses is an example of distance val- ular efforts to check all differences relative to the se- ues that deviate from the pattern provided by the com- quence reported by Horai et al. (1995), and in the se- bined sequence data. quence reported by us all these positions have been de- Our comparisons based on the combined data of all termined without ambiguity in a minimum of two 13 peptide-coding genes show striking agreement of the clones. While the discrepancies between the peptide- Downloaded from https://academic.oup.com/mbe/article/13/5/691/1083040 by guest on 28 September 2021

Table 4 Nucleotide Substitutions According to Codon Position (1, 2, 3) in Peptide-Coding Genes of Homo (H), Pun (P), and Gorilla (G)

H/G P/G 1 2 3 1 2 3 1 2 3 Ti Ti TV Ti Ti TV Ti Ti TV GENE a b TV Ti TV a b a b a b TV Ti TV a b a b a b TV Ti TV a b a b

NADHl ...... 10 8 3 4 1 7 52 1 2 9 6 2 7 1 16 52 2 6 10 11 1 7 - 16 59 1 6 NADH2 ...... 5 7 - 7 - 16 60 - 5 11 13 3 8 - 15 68 2 12 8 14 3 7 - 11 59 2 13 CO1 ...... 9 5 - 1 - 15 98 - 7 12 6 - 3 - 10 117 3 14 3 5 - 2 - 15 110 3 15 co11 ...... 2 3 - 2 - 7 48 1 2 5 6- l- 8 54 26 5 3- l- 651 34 ATPase8 ...... 1 2 1 1 - 1 12 1 - - 3 2 3 1 1 12 - 2 1 5 3 2 1 - 8 1 2 ATPase6 ...... 4 6 2 4 - 10 30 1 1 12 7 1 4 - 9 37 1 7 10 13 1 6 - 11 38 2 6 co111 ...... 5 4 - 3 - 7 52 2 2 5 2 1 5 1 5 60 2 6 8 2 1 4 1 8 59 2 6 NADH3 ...... 6 2 - 4 1 4 21 2 1 6 5 1 1 1 8 20 1 - 6 5 1 3 - 5 17 1 1 NADH4L ...... 1 1 - - - 1 17 1 1 1 l--- 3 18 2 12---- 2 17 1 2 NADH4 ...... 12 9 6 6 - 15 82 2 2 14 10 2 11 1 23 84 1 10 10 16 4 11 1 20 93 1 10 NADHS ...... 16 25 4 17 2 23 92 5 8 17 34 8 24 4 23 124 10 12 10 33 8 21 2 21 113 7 10 NADH6 ...... 4 4 - 3 1 7 29 - 3 4-- 4 1 5 30 - 4 3 3 - 4 - 9 35 - 5 Cyt b ...... 6 20 1 12 - 18 70 1 5 9 22 3 12 1 14 71 3 12 13 16 2 9 1 13 70 4 7 81 96 17 64 5 131 663 17 39 105 115 23 83 11 140 747 29 92 89 126 24 77 6 137 729 28 87 -- 0 ! -- L , -- , Cons nt sub 113 69 56 138 94 121 150 83 115 --I I J-w ’ 1 P-I‘ I 1 Total nt sub 194 69 850 243 94 1,008 239 83 981 R. tot nt sub 2.8 1 12.3 2.6 1 10.7 2.9 1 11.8 R. cons nt sub 1.6 1 0.8 1.5 1 1.3 1.8 1 1.4

NOTE.-Ti = transitions;TV = transversions;a = substitutionsinvolving leucine in both species;b = othernt substitutions.Cons nt sub: conservative nucleotide substitutions (Irwin, Kocher, and Wilson 1991). R. = ratio. Ratios for total nt substitution and for conservative nt changes are based on the values for codon positions 1 and 3, respectively, divided by the value for codon position 2. mtDNA of Western Lowland Gorilla 697 coding genes of the sequence determined by us and that LITERATURE CITED determined by Horai et al. (1995) are limited and would not affect phylogenetic analyses, it is evident that dif- ANDERSON, S., A. T. BANKIER, B. G. BARRELL, M. H. L. DE ferences of this kind, if artifactual, will have a profound BRUIJN, A. R. COULSON, J. DROUIN, I. C. EPERON, D. I? NIERLICH,B. A. ROE, E SANGER, F? H. SCHREIER, A. J. H. effect in population studies if the complete sequences SMITH, R. STADEN, and I. G. YOUNG. 198 1. Sequence and used are chimeric and represent different subspecies. organisation of the human mitochondrial genome. Nature The high degree of difference between the control 290:457-465. regions of the two complete molecules is unexpectedly ANDERSON, S., M. H. L. DE BRUIJN, A. R. COULSON, I. C. great. Therefore, if both our sequence and that incor- EPERON,E SANGER, and I. G. YOUNG. 1982. Complete se- porated from Foran, Hixson, and Brown (1988) are rep- quence of bovine mitochondrial DNA. J. Mol. Biol. 156: resentative of the Western lowland gorilla, it is evident 683-717. that the polymorphism within this subspecies is distinct- ARNHEIM, N., and M. HUEHN. 1979. The genetic behaviour of ly greater than that of other hominoid subspecies. The a cloned mouse ribosomal DNA segment mimics mouse

ribosomal gene. J. Mol. Biol. 134:743-765. Downloaded from https://academic.oup.com/mbe/article/13/5/691/1083040 by guest on 28 September 2021 difference between the two sequences is actually similar ARNASON, U., and A. GULLBERG. 1993. Comparison between to that recorded between Pan troglodytes troglodytes the complete mtDNA sequences of the blue and the fin and Pan troglodytes verus, the two subspecies of the whale, two species that can hybridize in nature. J. Mol. common chimpanzee that separated evolutionarily = 1.5 Evol. 37:3 12-322. MYA (Morin et al. 1994). The difference between the ARNASON, U., A. GULLBERG, E. JOHNSSON, C. LEDJE. 1993. control regions of the two complete gorilla sequences The nucleotide sequence of the mitochondrial DNA mole- can also be considered in the light of our own finding cule of the grey seal, Halichoerus grypus, and a comparison that another control region of the Western lowland go- with the mitochondrial sequences of other true seals. J. Mol. rilla (male specimen) sequenced by us differed by only Evol. 37:323-330. ARNASON, U., A. GULLBERG, and B. WIDEGREN. 1991. The a single transition in relation to the control region of the complete nucleotide sequence of the mitochondrial DNA of complete (female) sequence. the fin whale, Balaenoptera physalus. J. Mol. Evol. 33:556- The present analyses have demonstrated that PCR 568. products from two particular portions of the control re- ARNASON, U., and E. JOHNSSON. 1992. The complete mito- gion of the gorilla are heterogenous with respect to the chondrial sequence of the harbor seal, Phoca vitulina. J. number of Cs (Gs). Differences of this kind may be due Mol. Evol. 34:493-505. to heteroplasmy (i.e., the occurrence, in the same indi- ARNASON, U., R. SPILLIAERT, A. PALSDOTTIR, and A. ARNA- vidual, of mtDNA molecules with different constitution) SON. 1991. Molecular identification of hybrids between the caused by inexact polymerase reading in PCR or se- two largest whale species, the blue (Balaenoptera musculus) quencing reactions or an effect of phage propagation. and the fin whale (B. physalus). Hereditas 115: 183-l 89. ARNASON, U., X. Xu, and A. GULLBERG. 1996. Comparison Deletion events during phage propagation have been re- between the complete mitochondrial DNA sequences of corded in ribosomal gene inserts (Arnheim and Huehn Homo and the common chimpanzee based on nonchimaeric 1979). It should be observed, however, that in some of sequences. J. Mol. Evol. 42:145-152. the cases examined by us there was an increase rather ARNASON, U., Xu, X., A. GULLBERG, and D. GRAUR. 1996. than a decrease in the number of nucleotides. Irrespec- The “Phoca-standard”: an external molecular reference for tive of the cause of the variable numbers of Cs (or Gs) calibrating recent evolutionary divergencies. J. Mol. Evol. observed, the present findings demonstrate that direct (in press). sequencing of PCR products should not be applied to BIBB, M. J., R. A. VAN ETTEN, C. T. WRIGHT, M. W. WALBERG, regions that have long runs of Gs in the template, or and D. A. CLAYTON. 198 1. Sequence and gene organization regions where heteroplasmy might occur. The reason for of the mouse mitochondrial DNA. Cell 26: 167-180. FORAN, D. R., J. E. HIXSON, and W. M. BROWN. 1988. Com- this is that direct sequencing based on templates of this parison of ape and human sequences that regulate mito- kind will either produce unresolvable sequences or se- chondrial DNA transcription and D-loop DNA synthesis. quences showing artificial differences. Nucleic Acids Res. 16:5841-5861. GADALETA, G., G. PEPE, G. DE CANDIA, C. QUAGLIARIELLO, E. SBISA, and C. SACCONE. 1989. The complete nucleotide Acknowledgments sequence of the Rattus norvegicus mitochondrial genome: cryptic signals revealed by comparative analysis between We express our sincere thanks to Dr. Harold M. Mc- vertebrates. J. Mol. Evol. 28:497-516. Clure, Yerkes Regional Primate Research Center, Atlanta, GARNER, K. J., and 0. A. RYDER. 1992. Some applications of Georgia, for generously providing us with the gorilla PCR to studies in wildlife genetics. Symp. Zool. Sot. Lond. samples and to Drs. Einar and Betty Arnason, Albert Lev- 64:167-181. an, Maryellen Ruvolo, Rodney Honeycutt (associate ed- HIXSON, J. E., and W. M. BROWN. 1986. A comparison of the itor), and an anonymous reviewer for valuable comments small ribosomal RNA genes from themitochondrial DNA of of great apes and humans: sequence, structure, evolution, on the manuscript. The work was supported by the Swed- and phylogenetic implications. Mol. Biol. Evol. 3:1-18. ish Natural Sciences Research Council, The Crafoord HORAI, S., K. HAYASAKA, R. KONDO, K. TSUGANE, and N. Foundation, The Erik Philip-Sorensen Foundation, The TAKAHATA. 1995. Recent African origin of modem humans Carl Tesdorpf Foundation, and by contract ERBCHR revealed by complete sequences of hominoid mitochondrial XCT 930254 from the European Commission. DNAs. Proc. Natl. Acad. Sci. USA 92:532-536. 698 Xu and Arnason

IRWIN, D. M., T. D. KOCHER, and A. C. WILSON. 1991. Evo- RUVOLO, M., D. PAN, S. ZEHR, T. D. GOLDBERG, and T. R. lution of the cytochrome b gene of mammals. J. Mol. Evol. DISOTELL. 1994. Gene trees and hominoid phylogeny. Proc. 32: 128-144. Natl. Acad. Sci. USA 91:8900-8904. JANKE, A., G. FELDMAIER-FUCHS,W. K. THOMAS, A. VON SANGER, E 1981. Determination of nucleotide sequences in HAESLER, and S. PA;~Bo. 1994. The marsupial mitochon- DNA. Science 214: 1205-l 210. drial genome and the evolution of placental mammals. Ge- SPILLIAERT,R., G. VIKINGSSON, U. ARNASON, A. PALSDO~IR, netics 137:243-256. J. SIGUFUONSSON,and A. ARNASON. 1991. Species hybrid- KRE-ITEK, A., A. GULLBERG, and U. ARNASON. 1995. Sequence ization between a female blue whale (Buluenopteru mus- analysis of the complete mitochondrial DNA molecule of culus) and a male fin whale (B. physulus): molecular and the hedgehog, Erinaceus europaeus, and the phylogenetic morphological documentation. J. Hered. 82:269-274. position of the Lipotyphla. J. Mol. Evol. 41:952-957. Xu, X., and U. ARNASON. 1994. The complete mitochondrial KUMAZAWA, Y., and M. NISHIDA. 1993. Sequence evolution of DNA sequence of the horse, Equus cabullus; extensive het- mitochondrial tRNA genes and deep-branch animal phylo- eroplasmy of the control region. Gene 148:357-362. genetics. J. Mol. Evol. 37:380-398.

MORIN, I? A., J. J. MOORE, R. CHAKRABORTY,L. JIN, J. GOOD- Downloaded from https://academic.oup.com/mbe/article/13/5/691/1083040 by guest on 28 September 2021 RODNEY L. HONEYCUTT, reviewing editor ALL, and D. S. WOODRUFF. 1994. Kin selection, social structure, gene flow and the evolution of chimpanzees. Sci- ence 265:1193-1201. Accepted February 6, 1996