Evolution of CHR-2 SINEs in cetartiodactyl genomes: possibleevidence for the monophyletic origin of toothed whales

Masato Nikaido,1 Fumio Matsuno,1 Hideaki Abe,1 Mitsuru Shimamura, 1 Healy Hamilton,2 Hisashi Matsubayashi,1

Norihiro Okada 1

1Department of Biological Sciences, Graduate School of Bioscience and Biotechnology, Tokyo Institute of Technology, Nagatsuta-cho, Midori-ku,Yokohama 226-8501, Japan2Museum of Paleontology, University of California, Berkeley, CA 94720, USA

Received: 13 June 2001 / Accepted: 12 July 2001

Abstract. Short interspersed repetitive elements (SINEs) are akind of retroposons dispersed among the eukaryotic genomes. Pre-viously, we isolated and characterized a new SINE family, namedCHR-2, members of which are distributed in the genomes of ce-taceans, hippopotamuses, and ruminants. We analyzed systemati-cally more than a hundred members of the CHR-2 SINEs, whichwere isolated from the genomes of cetaceans and cow, togetherwith the additional data available in the DNA databases, andshowed that these SINEs are divided into at least five distinctsubfamilies that share diagnostic nucleotides and/or deletions. Ahybridization analysis clearly demonstrated that, among these fivesubfamilies, two subfamilies, named CD and CDO, are specific tocetaceans and toothed whales, respectively. We reconstruct theevolutionary history of the CHR-2 SINEs during evolution of ce-tartiodactyl genomes.

Introduction

SINEs are short interspersed repetitive elements that are present atmore than 104 copies per genome in multicellular animals frominvertebrates to mammals, and also in plants (Okada 1991a,1991b; Schmid and Maraia 1992; Shedlock and Okada 2000).SINEs are members of the retroposon family, which includes ret-roviruses, long terminal repeat (LTR) retrotransposons, LINEs(long interspersed repetitive elements), and processed retropseu-dogenes. Retroposons are amplified via cDNA intermediates andare reintegrated into the host genome by retroposition (Rogers1985; Weiner et al. 1986). Recent analyses of genomes such as thehuman genome have revealed the major contribution of retroposi-tion to the building of contemporary genomes, in particular thoseof mammals, during evolution (Smit and Riggs 1995; Kazazian2000; Hattori et al. 2000).

SINEs range from approximately 80 to 400 bp in length, andmost are derived from a tRNA (Okada 1991a, 1991b; Okada andOhshima 1995; Shedlock and Okada 2000). Notable exceptions arethe primate Alu and rodent B1 SINEs, which appear to be derivedfrom 7SL RNA (Ullu and Tschudi 1984). Each tRNA-derivedSINE consists of three regions: a tRNA-derived region, whichcontains internal promoter sequences specific for RNA polymeraseIII; a tRNA-unrelated region; and an AT-rich region. Sometimes,especially in the cases of non-mammalian SINEs, the tRNA-unrelated region can be further divided into two regions: a 58

region of unknown origin and a 38 region that resembles the 38untranslated region of a partner LINE (Okada et al. 1997).

We have demonstrated that analysis of sites of insertion ofSINEs can provide robust phylogenetic information (Okada 1991a;Shedlock and Okada 2000). In our attempt over the past few yearsto clarify the phylogenetic relationships among cetartiodactyls, wehave characterized a very large number of members of CHR-1 andCHR-2 SINEs (cetaceans, hippopotamuses, and ruminants) fromgenomic libraries of cetaceans, hippopotamuses and cows (Shi-mamura et al. 1997; Nikaido et al. 1999). A comprehensive analy-sis of the various CHR-1 and CHR-2 SINEs, together with addi-tional data obtained from DNA databases, revealed the genealogi-cal history of these two families of SINEs, as well as other newlycharacterized families of SINEs, such as the CHRS and CHRS-Sfamilies, whose discovery demonstrated the presence of a hugesuperfamily of tRNAGlu-derived SINEs in cetartiodactyl genomes(Shimamura et al. 1999).

In this study, we analyzed about a hundred members of CHR-2SINEs, which are isolated in the process of inferring a phylogenyof cetartiodactyls, and characterized several CHR-2 subfamiliesnamed FL (full length), MD (middle deletion), DT (deletion type),CD (cetacean deletion), and CDO (cetacean deletion odontoceti),which showed a significant difference of distributions. The analy-sis of diagnostic nucleotides revealed the phylogenic relationshipsamong CHR-2 families and the period in which each subfamilyincreased its copy number explosionary. We suggest that the con-cept of subfamilies and their distributional difference has signifi-cant power for the inference of the phylogenetic tree.

Materials and Methods

Extraction of DNA.Total genomic DNA of each animal was isolated byextraction with phenol–chloroform as described by Blin and Stafford(1976). All samples of extracted DNA were stored at 4°C. The speciesanalyzed in this study are listed in Table 1.

Construction and screening of genomic libraries, subcloning, andsequencing.Genomic libraries were constructed by ligation of genomicDNA from the short-finned pilot whale (Globicephala macrorhynchus), thesperm whale (Physeter macrocephalus), the bottlenosed dolphin (Tursiopstruncatus), the Dall’s porpoise (Phocoenoides dalli), the Baird’s beakedwhale (Beradius bairdii), the Amazon river dolphin (Inia geoffrensis), theminke whale (Balaenoptera acutorostrata), the humpback whale (Mega-ptera novaeangliae), and the cow (Bos taurus) into the plasmid pUC18.Both the plasmid and the genomic DNA were completely digested byHindIII. Clones containing CHR-2 were first isolated with an RNA probe,i.e., labeled in vitro transcripts from the total genomic DNA of spermwhales (Manley et al. 1980). After the characterization of each CHR-2subfamily, the libraries of DNA fragments from each animal were screenedfor CD and CDO with [g−32P]dATP-labeled oligonucleotides specific for

Correspondence to:N. Okada; E-mail: nokada@bio.titech.ac.jp

The nucleotide sequence data newly reported herein have been submittedto GenBank and have been assigned accession numbers as follows:AB071536-AB071595

Mammalian Genome 12, 909–915 (2001).DOI: 10.1007/s0033501-1015-4

© Springer-Verlag New York Inc. 2001

Incorporating Mouse Genome

Fig. 1A. Continued on next page.

M. Nikaido et al.: Evolution of CHR-2 SINEs in whales910

Fig. 1A. Continued on next page.

M. Nikaido et al.: Evolution of CHR-2 SINEs in whales 911

Fig. 1A. Continued on next page.

M. Nikaido et al.: Evolution of CHR-2 SINEs in whales912

these sequences for further understanding about the CHR-2 SINEs andphylogeny of whales. The inserts of positive clones that appeared to con-tain these SINE sequences were sequenced by the dideoxy chain-termination method (Sanger et al. 1977). CHR-2 sequences were alignedby eye using GENETYX-MAC Ver.10.1, and gaps were inserted to allowarrangement of sequences in order. The general consensus sequence wasbased on the alignment of FL, MD, and DT subfamilies (deletions were nottaken into account). The most frequently occurring nucleotide at a certainposition was chosen as the consensus nucleotide at that position. We de-fined the consensus nucleotide as the nucleotide that occurred more than55% of the time at the respective position.

Dot blot analysis.Total genomic DNA from individuals of each specieswe analyzed were dotted onto a GeneScreen Plus membrane (Du Pont-NEN Products, Boston, Mass.) with a dot blotting apparatus (model DP-96,Advantec, Tokyo). The genomic DNA for dot blot hybridization was pre-pared to be progressively decreased from 500 ng to 100 ng in totalamounts, and DNA of linearized plasmids that contained each CHR-2subfamily were dotted as a positive controls from 50 ng to 100 ng in totalamount. All of the DNAs were denaturated in 0.4M NaOH for 10 minbefore they were blotted. To examine the distributions of each CHR-2subfamily in cetartiodactyl genomes, the probes were prepared as follows.FL, MDs, and DT: A fragment of DNA was amplified by polymerase chainreaction (PCR) from cloned DNA, which included a FL sequence [thesequence corresponds to 66–319 of PM(2) 14]. The PCR product waslabeled internally in vitro by using the primer sets that were used in PCRwith [a−32P]dCTP; CD and CDO: the oligonucleotides that were able todistinguish CD from CDO and the other CHR-2 subfamilies were used asprobes. They are labeled at their 58 ends with [g−32P]dATP.

Hybridization was performed at 42°C overnight in a solution of 6 ×SSC, 1% SDS, 2 × Denhardt’s solution, and 100mg herring DNA/ml in afinal volume of 10 ml. Washing was performed at 50°C for 2 h in asolutionof 2 × SSC and 1% SDS.

Results and Discussion

We originally isolated CHR-2 SINEs by screening genomic librar-ies with in vitro labeled transcripts from the total genomic DNA(Manley et al. 1980; Endoh and Okada 1986) of the sperm whaleas probes (Shimamura et al. 1997). A large amount of sequenceinformation was accumulated for these SINEs, as well as forCHR-1 SINEs, during the characterization of orthologous loci with

respect to the presence or absence of a SINE. CHR-2 SINEs werefound to be dispersed in the genomes of cetaceans, hippopota-muses, and ruminants, as were CHR-1 SINEs. This observationimplied the monophyly of this group, contradicting the traditionalgrouping of artiodactyls (Shimamura et al. 1997). We have nowcompleted a comprehensive characterization of CHR-2 SINEs,which has allowed us to divide this family into at least six distinctsubfamilies by reference to internal duplications and diagnosticmutations (Figs. 1A and 1B). Our results, moreover, suggest themonophyly of toothed whales in view of the distribution of thesesubfamilies in the genomes of cetaceans.

Internal duplications account for about half of the total lengthof each CHR-2 sequence, and such duplications vary amongCHR-2 clones. The CHR-2 SINEs in the first group are the long-est, and they are designated the FL (full-length) subfamily; theyinclude, for example, PM(2)25, Minke 14, pilot 1, AB1007202,and BT(2)30. The SINEs in the second and the third groups aredesignated MDI (middle deletion) and MDII, respectively. Eachhas a deletion of a region that differs in terms of sequence from theFL subfamily. MDI and MDII are represented by Minke12 andMacco13, respectively. The SINEs in the fourth group are theshortest ones. Division of the SINEs into these four groups,namely, FL, MDI, MDII, and the group of shortest SINEs, wasbased on the number of internal duplications, and no diagnosticnucleotides were involved in the classification.

During characterization of the shortest CHR-2 SINEs, whichwere isolated from the genomes of the minke whale and the spermwhale, we recognized several SINEs that shared several diagnosticnucleotides that were not found in the FL and MD (MDI plusMDII) SINEs. Two subfamilies of the former SINEs were recog-nized, namely, the CD and CDO subfamilies (the designations ofthese subfamilies are explained below). The shortest SINEs thatdid not contain diagnostic nucleotides found in the CD and CDOsubfamilies of shortest SINEs were collectively designated DT(deletion type) SINEs.

The diagnostic nucleotide in the CD subfamily is G at nucleo-tide (nt) position 226, 252, and 262; for example, where A is foundin the FL, MD, and DT subfamilies. Although we searched exten-sively for members of the CD subfamily in the genomes of non-cetaceans (hippopotamuses and ruminants), we failed to find them,

Fig. 1. Compilation of CHR-2 SINEs(A) and a comparison of consensussequences of subfamilies of CHR-2 SINEs(B). The correlations betweennames of clones and animals are as follows: Hump, humpback whale;Minke, Minke whale; Macco, Sperm and Sp, sperm whale; Amz, Amazonriver dolphin; Isi, Dall’s porpoise; Bado, bottlenosed dolphin; Mago, short-finned pilot whale; Tuti, Baird’s beaked whale. The DNA sequences ofclones PM, GM, and BT were reported previously and are now available inGenBank. The sequences of BTBLGPS, BTAGODNA, BTFCG2REC, andABs were found during homology searches with the BLASTN program

(Altschul et al. 1990) by using CHR-2-related sequences as query se-quences. The nucleotide numbers in the alignments were based on those ofthe general consensus sequence. The consensus sequence of the subfamilywas derived from alignments of each member of the various groups. Thepositions of probes used for the dot hybridization analysis are shown inshaded boxes. Duplicated regions of 22–25 nt are boxed in the generalconsensus sequence, and partially related sequences are underlined. Diag-nostic nucleotides and deletions are highlightened.

M. Nikaido et al.: Evolution of CHR-2 SINEs in whales 913

an observation that suggests that the CD subfamily is specific tocetaceans (cetacean-specificdeletion type).

During an extensive survey of members of the CD subfamily,we found a novel subfamily that shared several diagnostic nucleo-tides that differed from those of the CD subfamily. The membersof this novel subfamily were isolated from pilot whales and spermwhales, and no members of this new subfamily were isolated frombaleen whales. Therefore, we designated this subfamily CDO (ce-taceandeletion specific forodontoceti). This subfamily was ap-parently specific to the genomes of Odontoceti (toothed whales).The SINEs in this subfamily were obviously shorter than SINEs inthe CD subfamily as the result of a long deletion (about 60 nt) inthe 38-end region. Members of the CDO subfamily not only hadthe same deletion but also included diagnostic nucleotides specificto this subfamily. The diagnostic nucleotides of the CDO subfam-ily were GG at nt positions 98 and 99, where AA was found inother subfamilies. Some diagnostic nucleotides were shared by theCDO and CD subfamilies. For example, at nts 103, 111, and 194,we found G, A, and T, respectively. This observation implies asibling relationship between these two subfamilies.

In view of the diagnostic nucleotides shared among subfami-lies, we performed dot-blot hybridization analysis to confirm theputative distribution of each subfamily. The results are shown inFig. 2A. Since no diagnostic nucleotides were recognized for the

FL, MD, and DT subfamilies, we could not design oligonucleotideprobes that would allow us to distinguish among them. Therefore,we first examined the absence or presence of any member of theCHR-2 family as using a probe that corresponded to the 38 regionof FL (see Fig. 6 in Shimamura et al. 1997). This probe hybridizedto the genomes of cetaceans, hippopotamuses, and ruminants, butnot to those of pigs, camels, and other mammals, with the resultsbeing consistent with our previous conclusion about the paraphylyof artiodactyla (Shimamura et al. 1997, 1999).

Oligonucleotides specific to the CD subfamily and the CDOsubfamily, respectively, were designed by reference to the regionin which several diagnostic nucleotides were clustered (see Fig.1B). Using the oligonucleotide specific to CD SINEs as the probe,we detected no signals except in the analysis of cetaceans. Thus,this subfamily of SINEs appeared to be distributed specifically inthe genomes of cetaceans, whereas the CDO subfamily was ap-parently restricted to toothed whales. This distribution of CD andCDO subfamilies, as revealed by hybridization analysis, was con-sistent with our assumptions derived from data accumulated duringthe screening of genomic libraries. From the number of positiveclones isolated during the screening of genomic libraries withprobes specific to the CD subfamily and the CDO subfamily ofSINEs, we estimated that the numbers of copies of SINEs of eachsubfamily in the genome of the sperm whale was approximately

Fig. 2. Distribution of subfamilies of CHR-2 SINEs, as revealed by dot-blot hybridization(A) and integration of results into their phylogenetic context(B). The phylogenetic tree ofcetaceans and artiodactyls was based on our previous conclusions (Nikaido et al. 1999).

M. Nikaido et al.: Evolution of CHR-2 SINEs in whales914

104 in each case. The apparent absence of any significant differ-ence between the numbers of copies of CD and CDO SINEs wasconsistent with the results of dot hybridization analysis (Fig. 2A).

From the patterns of distribution of the various subfamilies ofCHR-2 SINEs, we estimated the timing of their amplification inthe phylogeny of cetartiodactyls, as shown in Fig. 2B. As sug-gested in a previous report, CHR-2 SINEs might have been gen-erated from CHR-1 type III SINEs in a common ancestor of ce-taceans, hippopotamuses, and ruminants (Shimamura et al. 1999).This hypothesis was based on the observation that CHR-2 SINEsspecifically share diagnostic nucleotides that are specific to CHR-1type III SINEs. Since the FL, MDI, MDII, and DT subfamilies areall distributed in the genomes of cetaceans, hippopotamuses, andruminants, one of these subfamilies can be presumed to have beengenerated in a common ancestor of these species from CHR-1 typeIII SINEs. Among these four subfamilies, it is difficult to identifywhich subfamily of CHR-2 SINEs was created first, because twoalternative explanations are equally plausible, namely, generationof subfamilies from FL to DT via MD by deletion or generation ofsubfamilies from DT to FL via MD by duplication. Since the CDand CDO subfamilies share several distinct diagnostic nucleotides,they are obviously closely related and, thus, it is reasonable tospeculate that CDO was generated from CD by deletion in view ofthe fact that CDO is restricted to toothed whales, whereas CD ismore widely distributed among cetaceans. It is also likely that CDwas generated from DT because CD and DT SINEs are equal inlength.

The patterns of distribution of SINEs are sometimes informa-tive in terms of phylogeny (Serdobova and Kramerov 1998; Shi-mamura et al. 1999). The distribution of CDO SINEs suggests thatall toothed whales, including sperm whales, might very likely bemonophyletic. Since 1993, Milinkovitch and his group here pub-lished several papers in which they claimed that toothed whales areparaphyletic, with sperm whales being more closely related tobaleen whales than to the other toothed whales (Milinkovitch et al.1993, 1994, 1996; Milinkovitch 1995). Our data are inconsistentwith such a hypothesis, and on the basis of information of thesesubfamilies, we recently succeeded in isolating and characterizingparticular SINEs inserted at a particular locus in a common an-cestor of toothed whales (Nikaido et al. 2001). Our results tend tosupport the traditional morphological classification (e.g., Fordyceand Barnes 1994) and are not consistent with hypotheses proposedby Smith et al. (1996) and Arnason and Gullberg (1994), as wellas Milinkovitch et al. (1993).

Acknowledgment.This work was supported by research grants from theMinistry of Education, Science, Sports and Culture of Japan to N. Okada.

References

Arnason U, Gullberg A (1994) Relationship of baleen whales establishedby cytochrome b gene sequence comparison. Nature 367, 726–728

Deininger PL, Batzer MA (1993) Evolution of retroposons. Evol Biol 27,157–196

Endoh H, Okada N (1986) Total DNA transcription in vitro: a procedure todetect highly repetitive and transcribable sequences with tRNA-likestructures. Proc Natl Acad Sci USA 83, 251–255

Fordyce RE, Barnes LG (1994) The evolutionary history of whales anddolphins. Annu Rev Earth Planet Sci 22, 419–455

Hattori M et al. (62 co-authors) (2000) The DNA sequence of humanchromosome 21. Nature 405, 311–320

Kazazian HH Jr. (2000) Genetics. L1 retrotransposons shape the mamma-lian genome. Science 289, 1152–1153

Manley JL, Fire A, Cano A, Sharp PA, Gefer ML (1980) DNA-dependenttranscription of adenovirus genes in a soluble whole-cell extract. ProcNat Acad Sci USA 77, 3855–3859

Milinkovitch MC (1995) Molecular phylogeny of cetaceans prompts revi-sion of morphological transformations. Trends Ecol Evol 10, 328–334

Milinkovitch MC, Orti G, Meyer A (1993) Revised phylogeny of whalessuggested by mitochondrial ribosomal DNA sequences. Nature 361,346–348

Milinkovitch MC, Meyer A, Powell JR (1994) Phylogeny of all majorgroups of cetaceans based on DNA sequences from three mitochondrialgenes. Mol Biol Evol 11, 939–948

Milinkovitch MC, LeDuc RG, Adachi J, Farnir F, Georges M, HasegawaM (1996) Effects of character weighting and species sampling on phy-logeny reconstruction: a case study based on DNA sequence data incetaceans. Genetics 144, 1813–1817

Nikaido M, Rooney AP, Okada N (1999) Phylogenetic relationshipsamong cetartiodactyls based on insertions of short and long interspersedelements: hippopotamuses are the closest extant relatives of whales. ProcNatl Acad Sci USA 96, 10261–10266

Nikaido M, Matsuno F, Hamilton H, Brownell Jr. LR, Cao Y et al. (2001)Retroposon analysis of major cetacean lineages: the monophyly oftoothed whales and the paraphyly of river dolphins. Proc Natl Acad SciUSA, 98, 7384–7389

Okada N (1991a) SINEs. Curr Opin Genet Dev 1, 498–504Okada N (1991b) SINEs: short interspersed repeated elements of the eu-

karyotic genome. Trends Ecol Evol 6, 358–361Okada N, Ohshima K (1995) Evolution of tRNA-derived SINEs. In Maraia

RJ, ed.The Impact of Short Interspersed Repetitive Elements (SINEs) onthe Host Genome.(Austin, Tex.: R.G. Landes Company), pp. 61–79

Okada N, Hamada M, Ogiwara I, Ohshima K (1997) SINEs and LINEsshare common 38 sequences: a review. Gene 205, 229–243

Rogers J (1985) Origins of repeated DNA. Nature 317, 765–766Schmid C, Maraia R (1992) Transcriptional regulation and transpositional

selection of active SINE sequences. Curr Opin Genet Dev 2, 874–882Serdobova IM, Kramerov DA (1998) Short retroposons of the B2 super-

family: evolution and application for the study of rodent phylogeny. JMol Evol 46, 202–214

Shedlock AM, Okada A (2000) SINE insertions: powerful tools for mo-lecular systematics. BioEssays 22, 148–160

Shimamura M, Yasue H, Ohshima K, Abe H, Kato H et al. (1997) Mo-lecular evidence from retroposons that whales form a clade within even-toed ungulates. Nature 388, 666–670

Shimamura M, Abe H, Nikaido M, Ohshima K, Okada N (1999) Geneal-ogy of families of SINEs in cetaceans and artiodactyls: the presence ofa huge superfamily of tRNAGlu-derived families of SINEs. Mol BiolEvol 16, 1046–1060

Smit AFA, Riggs AD (1995) MIRs are classic, tRNA-derived SINEs thatamplified before the mammalian radiation. Nucleic Acids Res 23, 98–102

Smith MR, Shivji MS, Waddell VG, Stanhope MJ (1996) Phylogeneticevidence from the IRBP gene for the paraphyly of toothed whales, withmixed support for Cetacea as a suborder of Artiodactyla. Mol Biol Evol13, 918–922

Ullu E, Tschudi C (1984) Alu sequences are processed 7SL RNA genes.Nature 312, 171–172

Weiner AM, Deininger PL, Efstratiadis A (1986) Nonviral retroposons:genes, pseudogenes, and transposable elements generated by the reverseflow of genetic information. Annu Rev Biochem 55, 631–661

M. Nikaido et al.: Evolution of CHR-2 SINEs in whales 915