Copyright 0 1996 by the Genetics Society of America The Complete Mitochondrial DNA Sequence of the Bichir (Polypterus ornutipinnis), a Basal Ray-Finned Fish: Ancient Establishment of the Consensus Vertebrate Gene Order Katharina Noack, Rafael Zardoya and Axel Meyer Department of Ecology and Evolution, and Program in Genetics, State University of New York, Stony Brook, New York 11794-5245 Manuscript received May 27, 1996 Accepted for publication July 20, 1996 ABSTRACT The evolutionary position of bichirsis disputed, andthey have been variously alignedwith ray-finned fish (Actinopterygii) or lobe-finnedfish (Sarcopterygii),which also include tetrapods. Alternatively, they have been placed into their own group, the Brachiopterygii. The phylogenetic position of bichirs as possibly the most primitive living bony fish (Osteichthyes)made knowledge about their mitochondrial genome of considerable evolutionary interest.We determined the complete nucleotide sequence(16,624 bp) of the mitochondrial genome of a bichir, Polypterus ornutipinnis. Its genome contains 13 protein- coding genes, 22 WAS,two rRNAs and one major noncoding region. The genome’s structure and organization show that this is the most basal vertebrate that conformsto the consensus vertebratemtDNA gene order. Bichir mitochondrial protein-coding and ribosomal RNA genes have greater sequence similarity to ray-finned fish than to either lamprey or lungfish. Phylogenetic analyses suggest the bichir’s placement as the most basal living member of the ray-finned fish and rule outits classification as a lobe- finned fish. Hence, its lobe-fins are probably not a shared-derived trait with those of lobe-finned fish (Sarcopterygii). OW fish (Osteichthyes) are typically divided into Figure 1 summarizes some of the various phyloge- B two major groups,the actinopterygians (ray- netic hypotheses involving bichirs. One of the earliest finned fish) and the sarcopterygians (lobe-finned fish). hypotheses of polrpterid ancestry held that they were With >25,000 known species, the ray-finned fish com- sarcopterygians, and bichirs were placed into thesubor- prise by far the largest class of vertebrates. Ray-finned der Crossopterygii along with fossil rhipidistians, coela- fish date back to the early Devonian (CARROLL 1988), canths and Iungfish (Figure la) (HUXLEY1861; COPE and their diversity has since thenincreased steadily both 1871; reviewed in PATTERSON1980, 1982). Goodrich in terms of morphologcal diversity and in terms of (1928), for the first time, considered placing polypter- numbers of species. By contrast, the once highly suc- ids within the actinopterygians as a taxonomically dis- cessful and diverse lobe-finned fish, now only include tinct lower division.Since then, thealternative two main two extant fish groups, thelungfish and thecoelacanth. hypotheses of polypterine relationships have been that During the last 150 yr, both of these latter groups have polypterids are either classified as a distinct subclass of been variously thought to be the closest living relatives Osteichthyes, the Brachiopterygii (Figure lb),or as of land vertebrates (e.g., ZARDOYA and MEYER 1996a; basal actinopterygians (Figure IC). Themain problems reviewed in MEYER 1995). of assessingpolypterine relationships are that the fossil While the classification of bony fish into either the record for these fish only extends back to the Eocene ray-finned or the lobe-finned fish is without problems, (GREENWOOD1974; CARROLL1988) and that they show the phylogenetic placement of species of the order a large number of primitive ancestral as well asuniquely Polypteriformes into one of these two categories (e.g., derived characteristics that make the determination of PATTERSON1982; LAUDER and LIEM1983, also referred their phylogenetic position, and hence their classifica- to as Cladistia) has been historically difficult, and it tion, difficult. Bichirs have a row of dorsal finlets that still remains somewhat uncertain (Table 1). Unlike any extend down their back, primitive rhombicganoid other groupof living fish (reviewed in PATTERSON 1982; scales, numerous spiracular plates, and lobed pectoral LAUDER and LIEM 1983), polypteriform fish (bichirs, fins. Their pectoral fins have a fleshy external appear- Polypterus, and reedfish, Calamoichthys) have been ance similar to those of lungfish and the coelacanth. a number of widely differing taxonomic placed into Polypterine fish also have a well-vascularized lung that groups during the last century. arises from the ventral foregut and a pair of external gills in juveniles, a rare feature, not seen in any other Correspondingauthor Axel Meyer, Department ofEcoIogy and Evolu- tion, State University of New York, Stony Brook, NY 117945245, fish, except in some species of lungfish. E-mail: [email protected] Many authors still argue that Polypteriformes are Genetics 144: 1165-1180 (November, 1996) 1166 K. Noack, R. Zardoya and A. Meyer TABLE 1 morphologically too distinct from actinopterygians to Systematic position of bichirs be classified as such (JESSEN 1973; NELSON1973; JARVIK 1980; BJERRING 1985).This view, however, does not clar- Superclass: Gnathostomata ify what other group of fish polypterids are most closely Class Chondrichthyes related to. Most recent authorshave, basedon morpho- Subclass: Holocephali (Chimaeras) logical characters such as scale structure and neuro- Subclass: Elasmobranchii (sharks, rays) Class Sarcopterygii cranial ossification, placed polypterids with actinoptery- Subclass: Coelacanthimorpha (Latimeria) gians (Figure IC) (GARDINER1973; SCHAEFFER1973; Subclass: Dipnoi (lungfishes) PATTERSON1980,1982; LAUDER and LIEM1983; GARDI- Subclass: Tetrapoda (amphibians, reptiles, birds, NER and SCHAEFFER 1989). However, the placement of mammals) Polypteriformes within the Actinopterygii is also uncer- Class Actinopterygii Subclass: Chondrostei tain. Often polypterids are placed within the Chondros- Order: Polypteriformes (bichirs, Polypterus, tei, a group to which most authors also add the stur- reedfishes, Calamoichthys) geons and their relatives (Table 1, e.g., GARDINER1967; Order: Acipenseriformes (sturgeons, Acipenser; ANDREWS 1967; CARROLL1988; NELSON1994). Whereas paddlefishes, Polyodont) Subclass: Neopterygii other researchers classify polypterids as a taxon of equal Order: Semionotiformes (gars, Lepisosteus) rank and basal to the Chondrostei (reviewed in PAT- Order: Amiiformes (bowfins, Amia) TERSON 1982; LAUDER and LIEM1983). Division: Teleostei Among other reasons, knowledge ofthe phylogenetic Modified from NELSON(1994). position of the bichir is of importance for the recon- struction of the development and evolution of verte- A Bichirs Lobe-finned fish HUXLEY 1861 COPE 1871 Tetrapods Ray-finned fish B Bichirs I FIGURE1.-Alternative hwotheses of rela- ,I JESSEN 1973 tionships between bichir, ray-finned fish and Ray-finned fish NELSON1973 lobe-finned fish. (A) Polypteriformes are lobe- JARVIK 1980 finned fish. (B) Polypteriformes are a third sub Lobe-finned fish BJEMNG 1985 class of Osteichthyes with an unresolved posi- tion in respect with ray-finned and lobe-finned fish. (C) Polypteriformes are primitive ray- Tetrapods finned fish. C Bichirs GOODRICH 1928 GARDINER 1973 -Ray-finned fish SCHAEFFER19n PAlTERSON 1982 I I Lobe-finned fish L*EM 1982 GARDINER & SCHAEFFER 1989 Tetrapods NELSON 1994 Polypterus Mitochondrial DNA 1167 brate traits such as paired limbs from fins (e.g., SHUBIN 3.1-kb fragment was amplified using specific oligonucleotide primers and then cloned into pGEM-T (Promega). and ALBERCH 1986; SORDINOand DUBOULE1996). How- Plasmid DNA was extracted from each cloneusing a Wizard ever, so far, only two molecular phylogenetic studies miniprep kit (Promega). After ethanol precipitation, purified have incorporated nucleotide sequence datafrom bich- DNA was used as template for Taq Dye Deoxy Terminator irs. NORMARKet al. (1991) used partial DNA sequences cyclesequencing reactions (Applied Biosystems Inc.) follow- of the mitochondrial qtb, COI and COII genes to infer ing manufacturer’s instructions. Sequencing was performed with an automated DNA sequencer (373A Stretch, Applied phylogenetic relationships of Neopterygians, using the Biosystems Inc.). Sequences were obtained using both M13 bichir as outgroup taxon. LE et al. (1993) used portions universal sequencing primers and 37 specificallydesigned oli- of the large nuclear ribosomal gene in an attempt to gonucleotide primers. The sequencesobtained from each infer the phylogeny among major fish lineages. Their clone were -350 bp in length and each sequence overlapped study supported theview that polypterids are aprimitive the next by -100 bp. In no case were differences in sequence ray-finned fish lineage, albeit with low confidence. observed among the overlapping regions. Sequence data were analyzed by use of the GCG program Mitochondrial genes are increasingly used to infer package (DEVEREUXet al. 1984) and alignments were per- phylogenetic relationships both among closely related formed using CLUSTAL W (THOMPSONet al. 1994). The data species as well as distantly related ones (KUMAZAWAand were subjected to maximum parsimony (MP) analyses using NISHIDA1993; CAO et al. 1994; RUSSO et al. 1996; ZAR- PAUP Version 3.1.1 (SWOFFORD1993). MP analyses were per- formed using heuristic searches (TBR branch swapping, DOYA and MEYER1996~). ZARDOYA and MEYER(1996a) MULPARS option in effect, simple stepwise addition of taxa). show that combined sets of all mitochondrial protein Maximum likelihood (ML)