Protein sequences indicate the turtles branched off from the amniote tree after mammals. - 16854y

(1)

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Protein sequences indicate the turtles branched off from the amniote tree after

mammals.

Caspers, G.J.M.; Reinders, G.J.; Leunissen, J.A.M.; Wattel, J.

Publication date

1996

Published in

Journal of molecular evolution

Link to publication

Citation for published version (APA):

Caspers, G. J. M., Reinders, G. J., Leunissen, J. A. M., & Wattel, J. (1996). Protein

sequences indicate the turtles branched off from the amniote tree after mammals. Journal of

molecular evolution, (42), 580-586.

General rights

It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)

and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open

content license (like Creative Commons).

Disclaimer/Complaints regulations

If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please

let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material

inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter

to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You

will be contacted as soon as possible.

(2)

J Mol Evol (1996) 42:580-586

lou..ALoFMOLECULAR

IEVOLUTION

Protein Sequences Indicate That Turtles Branched Off from the Amniote

Tree After Mammals

Gert-Jan Caspers, 1 Geert-Jan Reinders, 1 Jack A.M. Leunissen, 2 Jan Wattel, 3 Wilfried W. de Jong 1'3

1Department of Biochemistry, University of Nijmegen, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands

2CAOS/CAMM Center, University of Nijmegen, P.O. Box 9010, 6500 GL Nijmegen, The Netherlands

3Institute for Systematics and Population Biology, University of Amsterdam, P.O. Box 94766, 1090 GT Amsterdam, The Netherlands

Received: 6 October 1995 / Accepted: 24 January 1996

Abstract.

The phylogenetic relationships among the

major groups of amniote vertebrates remain a matter of

controversy. Various alternatives for the position of the

turtles have been proposed, branching off either before or

after the mammals. To discover the phylogenetic posi-

tion of turtles in relation to mammals and birds, we have

determined cDNA sequences for the eye lens proteins

~A- and c~B-crystallin of the red-eared slider turtle

(Trachemys scripta elegans).

In addition, databases were

searched for turtle protein sequences, for which mam-

malian, avian, and outgroup orthologs were available.

All sequences were analyzed by three phylogenetic tree

reconstruction methods (neighbor-joining, maximum

parsimony, and maximum likelihood). Including the

c~-crystallins, 7 out of 12 proteins support a sister-group

relation of turtles and birds with all 3 methods. For each

of the other five proteins no topology was consistently

preferred by the three approaches. Analyses of the com-

bined amino acid data (1,695 aligned sites) also give

extremely strong evidence that turtles are nearer to birds,

indicating that mammals branched off before the diver-

gence between turtles and birds occurred.

Key words:

Testudines - -

Trachemys scripta elegans

- - Tetrapod phylogeny - - Molecular evolution - -

c~-crystallin

Correspondence to: W . W . de Jong at his N i j m e g e n address

Introduction

The major groups of amniote vertebrates diverged from

a common ancestor during a relatively short period at the

end of the Paleozoic era, about 300-250 million years

ago (Carroll 1987; Laurin and Reisz 1995). Until re-

cently there was a broadly accepted view of the relation-

ships between these amniote groups, based on paleon-

tological and morphological evidence. This held, as

already implied by Haeckel (1866), that mammals rep-

resent the sister group of all other extant amniotes. The

turtles (Testudines) were generally considered to be the

next group to have branched off, followed by Lepido-

sauria (tuatara, lizards, and snakes), with the final diver-

gence occurring between crocodiles and birds (together

the Archosauria) (Carroll 1987).

In the past decade this view has seriously been chal-

lenged with the advent of more sophisticated cladistic

analyses. The hypothesis of a sister-group relation be-

tween m a m m a l s and birds (Gardiner 1982, 1993;

Lcvtrup 1985), reviving the clade Haematothermia

(Owen 1866), has stirred much debate. Also, the position

of the turtles is controversial. Some authors unite Testu-

dines and Archosauria, to the exclusion of Lepidosauria

(LCvtrup 1977; Hennig 1983; Ax 1984). Alternatively,

the turtles have been proposed to be the first to branch off

from the amniotes, followed by the mammals (Gaffney

1980). This latter view has indeed been adopted in some

recent text books (e.g., Chaline 1990, p 78; Ridley 1993,

p 465). Various authors have, however, provided re-

newed support for the classical branching pattern, based

on morphological characters (Carroll 1987; Gauthier et

(3)

al. 1988; B e n t o n 1990; L a n r i n a n d Reisz 1995). A total e v i d e n c e approach, c o m b i n i n g m o r p h o l o g i c a l a n d m o - lecular data, also resulted in the traditional p h y l o g e n y (Eeruisse a n d K l u g e 1993). T h e b i r d - c r o c o d i l i a n sister- group relationship has further b e e n c o n f i r m e d b y a r e c e n t exhaustive study of m o l e c u l a r s e q u e n c e data (Hedges 1994). H o w e v e r , f r o m a m o l e c u l a r p o i n t of v i e w the position o f the turtles relative to m a m m a l s a n d other a m n i o t e s has n o t b e e n well studied. O n l y few turtle se- q u e n c e s have b e e n u s e d in p h y l o g e n e t i c analysis, a n d results f r o m different m o l e c u l e s have b e e n c o n t r a d i c t o r y (Hedges et al. 1990; Marshall 1992; Eeruisse a n d K l u g e 1993; V a n de Peer et al. 1993).

As a c o n t r i b u t i o n to s o l v i n g the p h y l o g e n e t i c p o s i t i o n of the turtles, we d e t e r m i n e d c D N A s e q u e n c e s for the eye lens proteins e~A- a n d e~B-crystallin f r o m a turtle a n d a n a l y z e d these s e q u e n c e s together with all i n f o r m a t i v e turtle protein s e q u e n c e s retrieved f r o m the databases. Be- cause birds are the best-represented n o n m a m m a l i a n amniotes in the databases, we l i m i t e d ourselves to r e s o l v i n g the f o u r - t a x o n case ((turtles, m a m m a l s , birds) outgroup). This w o u l d allow us to decide w h e t h e r turtles b r a n c h e d off f r o m the a m n i o t e tree before or after the m a m m a l s . It w o u l d s i m u l t a n e o u s l y c o n t r i b u t e to further settling the H a e m a t o t h e r m i a c o n t r o v e r s y . W h i l e f o u r - t a x o n approaches have b e e n w i d e l y used in p h y l o g e n e t i c studies (e.g., G r a u r et al. 1991; Steel et al. 1993), o n e should realize that these c a n b e m i s l e a d i n g in certain cases, as p o i n t e d o u t b y P h i l i p p e a n d D o u z e r y (1994). Therefore, if available, s e q u e n c e s of m o r e than o n e species were used for each of the four m a j o r clades. I n addition to the c~A- a n d ~ B - c r y s t a l l i n sequences, exhaustive database searches y i e l d e d s e q u e n c e s o f ten proteins for w h i c h orthologs are available f r o m turtles, m a m m a l s , birds, a n d outgroups. W e applied the three m a j o r tree c o n s t r u c t i o n m e t h o d s - - n e i g h b o r - j o i n i n g , m a x i m u m p a r s i m o n y , a n d m a x i m u m l i k e l i h o o d - - o n all sequences, apart a n d c o m - bined.

Materials and Methods

Sequence Analysis of eL-Crystallins. Total RNA was isolated from eyes of three juvenile red-eared slider turtles, Trachemys scripta elegans, by the lithium chloride/urea method (Auffray and Rougeon 1980). RNA was reverse transcribed using SuperScript reverse transcriptase (Gibco BRL/Life Technologies) and oligo(dT) primer, e~-Crystallin sequences were amplified from the resulting single-stranded cDNA by the polymerase chain reaction (PCR) method, using Taq polymerase (Gibco BRL/Life Technologies). Degenerated oligonucleotide primers were designed to amplify cDNA sequences coding for amino acid positions 12-160 of ~A-crystallin and positions 9-61 of c~B-crystallin. For e~A- crystallin, the primers were as described earlier (Caspers et al. 1994); for e~B-crystallin we used 5'-ATACTGCAGGATATCACCATTCA- CACCC-3' and 5'-ATAAAGCTTACCTC[C,T]GAGAGTCCCG- T[A,C,G,T]TC-3'. Hybridization temperatures were 45°C for the aA- crystallin amplification and 55°C for the c~B-crystalIin amplification. Reamplification with the same primers was necessary to obtain enough amplification product. Amplification products were made blunt-ended and 5'-phosphorylated with Klenow DNA polymerase and T4 poly-

581

nucleotide kinase (Gibco BRL/Life Technologies), respectively, and ligated into pGEM-3Zf(+) phagemid vector (Promega), cut with Sinai.

Constructs from several separate amplifications were sequenced in both directions using the Sequenase 2.0 DNA sequencing kit (United States Biochemical). The sequences have been deposited in the GenBank data base (accession nos. U31938 and U31939).

Data Base Searches. The SwissProt (version 31.0), PIR (version 44.0), EMBL (version 42.0), and GenBank (version 83.0) databases were searched for amino acid sequences and protein-coding nucleic acid sequences from turtle species. Only sequences for which at least a mammalian, avian, and outgroup homolog were available were used. Sequences suspected not to obey the orthohigy criterion (i.e., resulting from a gene duplication rather than a speciation event) were excluded (e.g., gastrirdcholecystokinin, parvalbumin, and ribonuclease). How- ever, as long as we found no evidence against the hypothesis that homologous sequences diverged from one another close in time to the ancestral divergence of their species lineages, they were treated as orthologs (Goodman et al. 1987); in cases where two different protein sequences of one species were available ([3-hemoglobin, insulin), both were included in the analyses. Where several turtle sequences were known, species from all available families were included. If available, the mammaiian sequences used were human, bovine, mouse, and a marsupial. For birds, chicken, and, if available, another species, pref- erably a paleognath, were used. When possible, multiple outgroups were used. Species names and database accession numbers are given in the legends of Fig. 2 and Table 1.

Phylogenetic Analyses. Because of the divergence times dealt with in this study, we used amino acid rather than nucleotide sequences for phylogenetic analyses. Sequences were aligned with PILEUP from the GCG package (Devereux et al. 1984). Phylogenetic analyses were performed with maximum parsimony (PROTPARS), neighbor-joining (Saitou and Nei 1987) (PROTDIST and NEIGHBOR, or TREECON) using Kimura distances (Kimura 1983), and maximum-likelihood methods (Felsenstein 1981) (PROTML) based on the JTT model (Jones et al. 1992). The programs PROTPARS, PROTDIST, and NEIGHBOR are from the PHYLIP package (Felsenstein 1993), TREECON has been written by Van de Peer and De Wachter (1994) and PROTML by Adachi and Hasegawa (1992). Confidence in the maximum-parsimony and neighbor-joining methods was assessed by bootstrapping (Felsen- stein 1985) (SEQBOOT and CONSENSE (Felsenstein 1993) for the programs from the PHYLIP package). Confidence in the maximum- Iikelihood method was assessed by the differences in log-likelihood from the highest-likelihood tree (Bishop and Friday 1985).

Results

o~-Crystallin Sequences

~ - C r y s t a l l i n s b e l o n g to the small heat-shock protein family (Caspers et al. 1995). T h e y occur in the vertebrate eye lens as m u l t i m e r i c c o m p l e x e s c o m p o s e d of two types o f h o m o l o g o u s subunits, c~A- a n d oLB-crystallin ( G r o e n e n et al. 1994). Both s u b u n i t s are e n c o d e d b y s i n g l e - c o p y genes ( K i n g a n d Piatigorsky 1983; Q u a x - J e u k e n et al. 1985), w h i c h avoids the p r o b l e m o f paralogy in c o m - parative studies. ~ A - C r y s t a l l i n protein s e q u e n c e s have already c o n t r i b u t e d to r e s o l v i n g the r e l a t i o n s h i p s be- t w e e n m a m m a l s , lizards, crocodiles, a n d birds (Stapel et

(4)

582 A

T u r t l e T e g u A l l i g a t o r C h i c k e n T i n a m o u H u m a n E l e p h a n t M o u s e B o v i n e S l o t h O p o s s u m K a n g a r o o F r o g B u l l f r o g l l l l l l l l l l l l l l l l l l l l l l l l l l i i i 1 2 2 2 3 3 3 3 3 4 4 5 5 5 5 5 5 6 7 7 7 7 8 8 8 9 9 9 9 0 2 2 2 2 2 2 3 3 3 3 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 3 6 7 8 0 6 7 1 2 3 7 9 0 4 0 2 5 6 7 8 9 0 2 4 6 3 6 9 0 1 3 6 1 2 3 4 7 8 9 0 2 5 8 2 3 4 6 7 8 9 0 1 2 3 4 5 6 8 A P L F S Y L L F D L F F I H L T V L E D K T L D E S I M D F I N S N V S A I T S A M S G P V Q S N M D T S Y S E P . . , I . F F . , E . L . , Q . . . F . . . V I E . . . A . . . A A . . . T . P . H N . . • . . I . F F . . E . L L . . . S . . . M . . . . I . , V . . S . . . V . V . . . . S . . . . V . . I H . D . • . . I . F F . L E . L . , Q . S . . . M . . . . I . . . S A . . . S . . . P . . . . P . H . . . • . . I . F F . L E . L . . Q . S . . . E . . M . . . . I . . . S . . . S . . . A . . . P . H . . , T . F Y . F F . . E . . L . Q . . . . D . . V F . , T V Q . . . L S . . . C . , I . T G L . A T H A , A , , F Y . F F . . E . , L . Q . . . . D . Q V . , , T V Q . . . L S , . . C , , I . , G , , A . H . , A . . F Y . F F . . E . . L . Q . . . . D . . V F . , T V L E . . . L S . . . G L . A G H . , A T , F Y . F F . . E . . L . Q . . . . D . . V F . , T V Q E . . . L S . . . I P . G V , A G H . . A . . F Y , F F . , E . . L . Q . . . . D . . V F , , T V L . . . . T A , . . L S . . . I - - - V , P . H . . T . S . Y . F F , . E . . L . Q . . . R V Y . , T V L . Y . S . . . S . . . I H . , . E S . H . D S . S . Y . F F . . E . . L . Q . . . V F . , T V L . , . S . . . S . . . I H . D . , A . H . D S • . F Y N V F M . . F . L , Q F G F . D . R . N , D T . L . , . S . . L . S , S , , I . . , M M . . L V S . H . . . • . F Y N V F M , , F , L V . . G F M D . R . N . D T . L . , , S . . L . S . S . . I . . . M M . G L . S . H . , .

B

T u r t l e C h i c k e n D u c k H u m a n M o u s e B o v i n e B u l l f r o g l l l l l l 1 1 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 4 5 5 5 5 5 5 6 9 0 3 4 5 6 7 8 9 1 2 3 5 7 2 3 5 6 8 9 0 1 2 3 8 9 0 1 3 4 5 7 1 L I P L F S F L T T R I D S L S S E F P T S G A L L R S - F L T L . V . . . . W . . S , . . I . Q . . L . . . P S . M . . - . F M . . . . W . A S . . . I . Q . . L . A . P S . M , . - I F M . W , . F . P . H S S . L . F . L . D . . . . T S Y . . P S . . A F W , , F . P . H S S . L . F . L . D , S . A T S Y . . P S , . A I W , . F , P . H S S . L . F . L . D , . A , T S Y . . P S . , A I W F Q F Y . . F G N K M E C I Q A D . . S . V - Y F K Y - . . L I

Fig. 1. Comparison of variable sites in turtle and other vertebrate

c~A-crystallins (A) and otB-crystallins (B). Deduced amino acid se-

quences of turtle

Trachemys scripta elegans etA- and c~B-crystallin

cDNAs (positions 12-160 and 9-61, respectively) were aligned with

corresponding sequences of e~-crystallins from other species, selected

to cover the variety within extant vertebrates. Only those positions are

shown at which different residues occur in these data sets. Amino acid

position numbers are those of the complete a-crystallin chains.

Dots

indicate where residues are identical to the turtle sequences;

dashes

denote deletions. Species names and database accession numbers are

given in the legends of Fig. 2 and Table 1.

al. 1984; de Jong et al. 1985; Eernisse and Kluge 1993;

Hedges 1994). oLB-Crystallin has only recently been used

in comparative studies (Lee et al. 1993; Lu et al. 1995).

The nucleotide sequences corresponding to amino

acid residues 12-160 of ~A-crystallin and to residues

9-61 of o~B-crystallin of the red-eared slider turtle

(Trachemys scripta eIegans) were determined. The de-

duced amino acids sequence of aA-crystallin accounts

for almost the entire length of this 173-residue polypep-

tide. Residues 9-61 of aB-crystallin represent the largest

part of the protein-coding information in the first exon of

the aB-crystallin gene. These turtle ~A- and otB-

crystallin sequences are aligned in Fig. 1 with corre-

sponding sequences of representative other vertebrate

ot-crystallins.

The neighbor-joining tree based on this oLA-crystallin

data set c l e a r l y includes the turtle in a t u r t l e -

lepidosaurian-archosaurian clade (bootstrap value of

93%) (Fig. 2A). Also the maximum likelihood tree and

the most parsimonious tree for these c~A-crystallin se-

quences support such a clade--in the case of the parsi-

mony tree, with a bootstrap value of 92% (not shown).

With all three tree construction methods the turtle is the

first to branch off within this clade. However, the statis-

tical support for the resulting lepidosaurian-archosaurian

clade is very weak (bootstrap value of 45% in the neigh-

bor-joining and 42% in the parsimony tree). This data set

also strongly groups the placental mammals together, as

well as the two marsupials, but gives only poor support

for mammalian monophyly. Comparison of the a B -

crystallin sequences in Fig. 1B reveals that position 53

probably represents a synapomorphous insertion of one

amino acid in the mammalian lineage. However, the

character state at this position is not known in lepido-

saurs and crocodiles. Analysis of the ~B-crystallin data

set supports a turtle-bird relationship with all three tree

construction methods (Table 1, see below).

Sequences from Databases

To combine the phylogenetic information from these

newly determined ot-crystallin sequences with that from

other proteins available in the databases, exhaustive

searches were performed for turtle sequences that might

enable the resolution of the turtle-mammal-bird rela-

tionship. Amino acid sequences from ten additional sets

of orthologous genes were found to be suitable for this

purpose (Table 1). Selection criteria are described in Ma-

terials and Methods.

Each data set was subjected to maximum parsimony,

neighbor-joining, and maximum-likelihood analyses.

The support for the three alternative branching orders of

mammals (M), turtles (T), and birds (B) with respect to

outgroup sequences is summarized in Table 1. This table

o v e r w h e l m i n g l y demonstrates the evidence for a

(M(T,B)) topology. Seven out of 12 protein sequences

support a sister-group relation of turtle and birds in all

analyses. Prolactin and nicotinamide adenine dinucleo-

tide dehydrogenase subunit 2 (ND2) support a (M(T,B))

topology with two of the methods, while cytochrome b,

myoglobin, and somatotropin support a (T(B,M)) topol-

ogy with maximum parsimony and maximum likelihood.

In the case of prolactin the likelihood support is equal for

all three topologies.

Combining all proteins in a single analysis provides

the strongest measure for the resolution of the four-taxon

case under investigation. To that end a composite out-

group sequence was constructed from the phylogeneti-

cally nearest outgroups (see footnote f of Table 1). For

the ingroups, a turtle, chicken, human, and mouse were

the only taxa available for all proteins. The combined

analyses of all amino acid sequences (1,695 aligned

sites) supports the sister-group relationship of turtle and

birds almost at the highest possible levels (Table 1, Fig.

2B).

(5)

583

A

Distance 0.1 ₄ 99

50 [

93 Tegu Turtle 43 Human

7~

Elephant Mouse 92 Bovine Sloth 90 Opossum

L Kangaroo

1001

Frog

I

Bullfrog Distance 0.1 I I

loo[

Chicken Tinamou Alligator Chicken Turtle Outgroup Human Mouse

Fig. 2. Amniote relationships inferred from (A) c~A-crystallin amino acid sequences (see Fig. 1A), and (B) the I2 combined protein sequences from Table 1. Neighbor-joining trees are shown, constructed with TREECON (Van de Peer and De Wachter 1994) using Kimura distances, with bootstrap values from 1,000 replications. Distance is proportional in the minimum number of mutations per residue. With the combined protein sequences, the fact that branch lengths in the reptile- bird lineage are considerably shorter than the branches leading to the mammalian species, also observed in the maximum-likelihood tree (not shown), is not due to a general acceleration in the evolutionary rate in the mammalian lineage but rather to particular more divergent sequences (mouse insulin and human cytochrome c and somatotropin).

Species names and database accession numbers for c~A-crystallins are: chicken Gallus gallus (P02504), tinamou Eudromia elegans (L25850), alligator Alligator mississippiensis (P06904), tegu lizard Tupinambis

teguixin (P02506), human Homo sapiens (P02489 with minor correc-

tion according to L25781), elephant Loxodonta africana (P02498), mouse Mus musculus (P02490), bovine Bos taurus (P02470), sloth

Choloepus hoffmanni (P02486), opossum Didelphis virginiana

(P02503), kangaroo Macropus rufus (P02502), frog Rana esculenta (up to amino acid position 70)/R. temporaria (from position 71 onward) (P02507 and P02508), and bullfrog R. catesbeiana (X85205). For species names and database accession numbers used in the combined protein analysis, see legend of Table 1.

Discussion

Previous studies o f protein and nucleic acid sequences failed to give conclusive evidence about the position o f turtles among the amniotes. 18S r R N A placed turtles outside a b i r d - m a m m a l clade (Hedges et al. 1990; Eer- nisse and Kluge 1993), although weighting the nucleotide positions generated the classical amniote tree (Van de Peer et al. 1993), or nearly so (Marshall 1992). 28S r R N A sequences, which had earlier been inconclusive (Hedges et al. 1990), did group turtles within a b i r d - reptile clade, but as sister group to crocodiles, while salamanders were also included in the b i r d - r e p t i l e clade in this analysis (Eernisse and Kluge 1993). Studies o f turtle insulin (Cascone et al. 1991), prolactin (Yasuda et al. 1990), somatotropin (Yasuda et al. 1989), and tyrosinase ( Y a m a m o t o et al. 1992; Morrison et al. 1994) noted that these proteins were closer to avian than to m a m m a - lian orthologs, a - and [3-hemoglobin sequences, and a c o m b i n e d analysis o f a - and [3-hemoglobin, myoglobin, and cytochrome c p l a c e d turtles in a clade with

croco-

diles and birds, excluding a m a m m a l - l e p i d o s a u r clade (Eernisse and Kluge 1993). In earlier studies, 13-hemoglobin and m y o g l o b i n grouped birds with mammals, both in m a x i m u m - p a r s i m o n y ( G o o d m a n et al. 1987; Hedges et al. 1990) and in m a x i m u m - l i k e l i h o o d analyses (Bishop and F r i d a y 1988). However, different taxonomic sampling m a y have played a role in these deviating results. Finally, mitochondrial t R N A sequences contained little phylogenetic information for inferring the position of turtles among the amniotes, while flanking protein- coding sequences (ND2) supported a placement of turtles as a sister group to a b i r d - c r o c o d i l e clade (Seutin et al. 1994).

It might be expected that, like in the case of the avian sister-group controversy (Hedges 1994), an extended set of various proteins will enable a more convincing resolution of the branching order o f turtles, mammals, and birds within the amniotes. W e therefore d e t e r m i n e d sequences of two additional turtle proteins, c~A- and c~B- crystallin. Database searches yielded a further ten proteins suitable for our purpose. W e analyzed these se-

(6)

inferred by maximum-parsimony, neighbor-joining, and maximum-likelihood analyses ~

Number of sites

Bootstrap support for alternative sister-group relationships in maximum parsimony analyses b Protein e Alignable Variable Informative (M(T,B)) (T(B,M)) (B (T,M))

e~A-Crystallin* 149 63 7 92 (89) 3 (+3) 5 (+3) ~B-Crystallin* 53 33 4 94 (61) <1 (+7) 6 (+5) Cytochrome b 167 64 6 29 (+0) 41 (155) 16 (+1) Cytochrome c* 104 22 2 69 (41) 26 (+1) 3 (+2) c~-Hemoglobin* 142 99 10 72 (321) 16 (+0) 2 (+2) 13-Hemoglobin* 146 117 19 67 (468) 15 (+1) 12 (+4) Insulin* 51 14 3 91 (25) 4 (+3) 3 (+3) Myoglobin 153 123 17 14 (+4) 85 (403) 1 (+6) ND2 60 53 2 45 (255) 25 (+0) 7 (+2) Prolactin 203 170 3 41 (525) 7 (+0) 37 (-2) Somatotropin 193 119 9 30 (+0) 42 (226) 28 (+ 1) Tyrosinase* 277 125 9 58 (249) 16 (+3) 25 (+2) Combined proteins f 1695 850 64 98 (1,728) 1 (+25) <1 (+29)

Bootstrap support for alternative sister-group relationships in neighbor-joining analyses c

Differences in log-likelihood from best trees for alternative sister-group relatioz!ships in maximum likelihood analyses d

Protein e (M(T,B)) (T(B,M)) (B(T,M)) (M(T,B)) (T(B,M)) (B(T,M)) aA-Crystallin* 97 3 <1 ML -1.9 _+ 2.2 -1.9 -+ 2.2 c~B-Crystallin* 99 <1 <1 ML -11.6 -+ 7.0 -11.6 _+ 7.0 Cytochrome b 24 29 36 -1.8 + 3.8 ML -2.7 + 3.1 Cytochrome c* 44 42 7 ML -1.3 + 3.6 -2.4 _+ 2.7 e~-Hemoglobin* 94 4 <1 ML -5.4 + 4.0 -5.4 -+ 4.0 [3-Hemoglobin* 87 9 <1 ML -0.3 -+ 8.6 -7.8 _+ 5.2 Insulin* 100 <1 <1 M L -32.5 _+ 1.6 -32.5 _+ 1.6 Myoglobin 50 50 <1 -2.1 ___ 2.6 ML -1.6 _+ 3.0 ND2 20 42 23 ML -2.0 _+ 2.2 -0.8 _+ 3.4 Prolactin 85 3 12 ML ML ML Somatrotropin 71 19 10 -0.1 + 4.2 ML -2.0 _+ 2.4 Tyrosinase* 50 22 29 M L -1.4 _+ 2.2 -1.5 + 2.1 Combined proteins f 99 1 <1 ML -38.7 _+ 16.5 -51.4 _+ 14.6

a Values in boldface type emphasize the typology supported by the three different methods of tree construction. Asterisks (*) indicate the proteins that support a sister-group relationship of turtle and bird in all analyses b Bootstrap values in % (1,000 replications) that support the respective branching orders are given, followed (in parentheses) by the number of steps in the majority-rule consensus parsimony tree or the number of extra steps required for the alternative sister-group relationships. These are not necessarily the numbers of steps in the actual most parsimonious trees. The numbers of informative sites refer to the branching orders of manunals, turtles, and birds, not to branching orders within any of these clades c Bootstrap values in % (1,000 replications) that support the respective branching orders are given

d The highest likelihood trees are indicated by ML. The differences of log-likelihoods, witli their SEs, are given for the alternative topologies

Species names and database accession numbers for the proteins used in the analyses are: for aA-crystallin as given in the legend of Fig. 2, excluding alligator, tegu, elephant, and sloth, in order to obtain a similar data set as for the other proteins; for c~B-crystallin, human (P02511), bovine (P02510), mouse (P23927), chicken (Q05713), duck Anas platyrhynchos

(Q05557), and bullfrog (X87114); cytochrome b, green turtle Chelonia

mydas (L12716), leatherback turtle Dermochelys coriacea (L12712),

snapping turtle Chelydra serpentina (L12713), human (P00156), bovine (P00157), mouse (P00158), opossum Monodelphis domestica (Q04911), chicken (P18946), titmouse Pants inornatus (P29638), and clawed toad

Xenopus laevis (P00160); cytochrome c, snapping turtle C. serpemina

(P00022), human (P00001), bovine (P00006), mouse (P00009), kangaroo

M. giganteus (P00014), chicken (P00016), ostrich Struthio camelus

(P0001g), and bullfrog (P00024); e~-hemoglobin, Western painted turtle

Chrysemys picta bettii (P13273), human P01922), bovine (P01966),

mouse (P01942), opossum D. virginiana (P01976), chicken (P01944), ostrich (P01981), clawed toad X. tropicalis (P07428), and axolotl Ambys-

toma mexicanum (P02015); [3-hemoglobin, Western painted turtle

(P13274), human (P02023), bovine (P02070), mouse (P02088 and P02089), oppossum D. virginiana (1702109), chicken (1702112), ostrich (P02123), clawed toad X. laevis (P02132 and P02133), and newt Triturus

cristatus (P10785 and P10786); insulin, slider turtles Trachemys scripta

and Chrysemys dorbigni mouse (P01325 and P01326), opossum D. vir-

giniana (P18109), chicken and ostrich (P01332), and clawed toad X. laevis

(P12706 and P12707); myoglobin, green turtle (P02202), map turtle

Graptemys geographica (P02201), human (P02144), bovine (P02192),

mouse (P04247), opossum D. virginiana (P02193), chicken (P02197), penguin Aptenodytes forsteri (P02199), carp Cyprinus carpio (P02204), tuna Thunnus albacares (P02205), and shark Galeorhinus australis

(P14397); nicotinamide adenine dinucleotide dehydrogenase subunit 2 (ND2; including 5~5 residues of cytochrome c oxydase subunit I), loggerhead turtle Caretta caretta (Sentin et al. 1994), terrapin turtle Mala-

clemys terrapin (Seutin et al. 1994), human (P03891,'P00395), bovine

(P03892/P00396), mouse (P03893/P00397), opossum D. virginiana

(P41305/P41310), chicken (P18937/PI8943), quail Coturnix japonica

(P24971/P24984), clawed toad X. laevis (P03894/P00398), and carp (P24972/P24985); prolactin, green tttrtle (P33090), human (P01236), bovine (P01239), mouse (P06879), lungfish Protopterus aethiopicus

(P33091), carp (P09585), and eel Anguilla anguiUa (P33096); somatotropin, green turtle (P34005), human (P01241), bovine (P01246), mouse (P06880), chicken (P08998), duck (P11228), and bullfrog (P10813); tyrosinase, snapping turtle Trionyx sinensis ($56789), human (P14679), mouse (Pl1344), chicken ($54189), quail ($56788), and frog R. ni-

gromaculata (Q04604). Due to computational limitations, in the maxi-

mum-likelihood analyses one of the mouse (P02089), toad (P02133), and newt (P10786) sequences from the {3-hemoglobin data set and tuna from the myoglobin data set had to be removed

f For the combined protein analysis, the turtle (green turtle for cytochi'ome b and myoglobin, loggerhead for ND2), chicken, human, and mouse (P02088 and P01325, respectively, in the cases of the recent duplications of the hemoglobin [3 and insulin genes) sequences were used. A composite outgroup sequence was constructed from the nearest available outgroups: frog e~A-crystaUin, bullfrog c~B-crystaUin, clawed toad cytochrome b, frog cytochi'ome c, toad c~-hemoglobin, toad [3-hemoglobin (P02132), toad insulin (P12706), carp myoglobin, toad ND2, lungfish prolactin, frog somatotropin, and frog tyrosinase

(7)

585

q u e n c e s w i t h three c o n c e p t u a l l y d i f f e r e n t m e t h o d s o f p h y l o g e n y reconstruction. A p p l y i n g all three m e t h o d s - - n e i g h b o r - j o i n i n g , m a x i m u m p a r s i m o n y , and m a x i m u m l i k e l i h o o d - - a v o i d s d i s c u s s i o n s a b o u t t h e i r r e s p e c t i v e m e r i t s ( S w o f f o r d and O l s e n 1990; H a s e g a w a and Fuji- w a r a 1993; H i l l i s et al. 1993, 1994; T a t e n o et al. 1994; D e B r y and A b e l e 1995; H u e l s e n b e c k 1995). O b s e r v e d c o n g r u e n c e a m o n g b r a n c h i n g patterns d e r i v e d f r o m in- d e p e n d e n t data sets and by d i f f e r e n t m e t h o d s p r o v i d e s the strongest o b t a i n a b l e e v i d e n c e for p h y l o g e n e t i c robustness.

It is clear f r o m T a b l e 1 that a sister-group relation o f turtles and birds, to the e x c l u s i o n o f m a m m a l s , is ex- t r e m e l y w e l l supported, e s p e c i a l l y w h e n the s e q u e n c e s are c o m b i n e d . U n f o r t u n a t e l y , the p r e s e n t l y a v a i l a b l e protein data sets are n o t y e t s u f f i c i e n t to c o m p l e t e l y r e s o l v e the b r a n c h i n g o r d e r o f the a m n i o t e s . T h e p o s i t i o n o f the l e p i d o s a u r s e q u e n c e in the oLA-crystallin trees is v e r y w e a k l y supported. O t h e r proteins g r o u p e d l e p i d o s a u r s w i t h m a m m a l s , w h i c h does n o t c o r r e s p o n d to any m o r - p h o l o g y - b a s e d p h y l o g e n y ( E e r n i s s e and K l u g e 1993). H o w e v e r , m o l e c u l a r analyses, c o m b i n i n g a m i n o acid and n u c l e i c acid data, h a v e a l r e a d y f i r m l y e s t a b l i s h e d the sister-group r e l a t i o n s h i p o f c r o c o d i l e s and birds ( H e d g e s 1994). T h e p r e s e n t analysis further refutes the H a e m a - t o t h e r m i a h y p o t h e s i s and restores the p o s i t i o n o f the turtles as h a v i n g b r a n c h e d o f f after m a m m a l s . It is quite l i k e l y that additional m o l e c u l a r data w i l l further c o n f i r m the c l a s s i c a l v i e w o f a m n i o t e p h y l o g e n y as b a s e d on m o r p h o l o g i c a l and p a l e o n t o l o g i c a l e v i d e n c e .

Acknowledgments.

This work was carried out under the auspices of the Netherlands Foundation for Life Sciences (SLW) with financial aid from the Netherlands Organization for Scientific Research (NWO).

References

Adachi J, Hasegawa M (1992) Computer science monographs, no. 27: MOLPHY: programs for molecular phylogenetics. I. - - PROTML: maximum likelihood inference of protein phylogeny. Institute of Statistical Mathematics, Tokyo

Auffray C, Rougeon F (1980) Purification of mouse immunoglobulin heavy chain mRNAs from total myeloma tumor RNA. Eur J Bio- chem 107:303-314

Ax P (1984) Das phylogenetische System. G Fischer Verlag, Stuttgart, pp 101-114

Benton MJ (1990) Phylogeny of the major tetrapod groups: morphological data and divergence dates. J Mol Evol 30:409-424 Bishop MJ, Friday AE (1985) Evolutionary trees from nucleic acid and

protein sequences. Proc R Soc B 226:271-302

Bishop MJ, Friday AlE (1988) Estimating the interrelationships of tetrapod groups on the basis of molecular sequence data. In: Benton MJ (ed) The phylogeny and classification of the tetrapods. Vol 1: amphibians, reptiles, birds. Clarendon Press, Oxford, pp 33-58 Carroi1 RL (1987) Vertebrate paleontology and evolution. WH Free-

man, New York, pp 1-15, 192-216

Cascone O, Turyn D, Dellacha JM, Machado VLA, Marques M, Vita N, Cassan C, Ferrara P, Guillemot JC (1991) Isolation, purification, and primary structure of insulin from the turtle

Chrysemys dorbigni.

Gen Comp Endocrinol 84:355-359

Caspers G-J, Leunissen JAM, de Jong WW (1995) The expanding small heat-shock protein family, and structure predictions of the conserved "e~-crystallin domain." J Mol Evol 40:238-248 Caspers G-J, Wattel J, de Jong WW (1994) aA-Crystallin sequences

group tinamou with ratites. Mol Biol Evol 11:711-713

Chaline J (1990) Paleontology of vertebrates. Springer Verlag, Berlin DeBry RW, Abele LG (1995) The relationship between parsimony and maximum-likelihood analyses: tree scores and confidence estimates for three real data sets. Mol Biol Evol 12:291-297

de Jong WW, Zweers A, Versteeg M, Dessaner HC, Goodman M (1995) et-Crystallin A sequences of

Alligator mississippiensis

and the lizard

Tupinambis teguixin:

molecular evolution and reptilian phylogeny. Mol Biol Evol 2:484-493

Devereux J, Haeberli P, Smithies O (1984). A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res 12: 387-395

Eernisse DJ, Kluge AG (1993) Taxonomic congruence versus total evidence, and amniote phylogeny inferred from fossils, molecules, and morphology. Mol Biol Evol 10:1170-1195

Felsensteih J (1981) Evolutionary trees from DNA sequences: a maximum likelihood approach. J Mol Evol 17:368-376

Felsenstein J (1985) Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783-791

Felsenstein J (1993) PHYLIP: phylogeny inference package, version 3.5c. Distributed by the author, Department of Genetics, University of Washington, Seattle, WA

Gaffney ES (1980) Phylogenetic relationship of the major groups of amniotes. In: Panchen AL (ed) The terrestrial environment and the origin of the land vertebrates. Academic Press, London, pp 593- 610

Gardiner BG (1982) Tetrapod classification. Zool J Linn Soc 74:207- 232

Gardiner BG (1993) Haematothermia: warm-blooded amniotes. Clad- istics 9:369-395

Gauthier J, Kluge AG, Rowe T (1988) Amniote phylogeny and the importance of fossils. Cladistics 4:105-209

Goodman M, Miyamoto MM, Czelusniak J (1987) Pattern and process in vertebrate phylogeny revealed by coevolution of molecules and morphologies. In: Patterson C (ed) Molecules and morphology in evolution: conflict or compromise? Cambridge University Press, Cambridge, pp 141-176

Graur D, Hide WA, Li W-H (1991) Is the guinea-pig a rodent? Nature 351:649-652

Groenen PJTA, Merck KB, de Jong WW, Bloemendal H (1994) Struc- ture and modifications of the junior chaperone c~-crystallin: from lens transparency to molecular pathology. Eur J Biochem 225:1-19 Haeckel E (1866) Generelle Morphologie der Organismen. G Reimer,

Berlin, Tafel XII

Hasegawa M, Fujiwara M (1993) Relative efficiencies of the maximum likelihood, maximum parsimony, and neighbor-joining methods for estimating protein phylogeny. Mol Phylogen Evol 2:1-5 Hedges S B (1994) Molecular evidence for the origin of birds. Proc Natl

Acad Sci USA 91:2621-2624

Hedges SB, Moberg KD, Maxson LR (1990) Tetrapod phylogeny inferred from 18S and 28S ribosomal RNA sequences and a review of the evidence for amniote relationships. Mol Biol Evol 7:607~33 Hennig W (1983) Testudines. In: Hennig W (ed) Stammesgeschichte

der Chordaten. P Parey, Hamburg, pp 132-139

Hillis DM, Allard MW, Miyamoto MM (1993) Analysis of DNA sequence data: phylogenetic inference. Methods Enzymol 224:456- 487

Hillis DM, Huelsenbeck JP, Cunningham CW (1994) Application and accuracy of molecular phylogenies. Science 264:671-677 Huelsenbeck JP (1995) The robustness of two phylogenetic methods:

four-taxon simulations reveal a slight superiority of maximum likelihood over neighbor joining. Mol Biol Evol 12:843-849 Jones DT, Taylor WR, Thornton JM (1992) The rapid generation of

(8)

mutation data matrices from protein sequences. Comput Appl Biosci 8:275-282

Kimura M (1983) The neutral theory of molecular evolution. Cam- bridge University Press, Cambridge, p 75

King CR, Piatigorsky J (1983) Alternative RNA splicing of the murine aA crystallin gene: protein-coding information within an intron. Cell 32:707-712

Lanrin M, Reisz RR (1995) A reevaluation of early amniote phylogeny. Zool J Linn Soc 113:165-223

Lee DC, Kim RY, Wistow GJ (1993) An avian aB-crystallin: non-lens expression and sequence similarities with both small (HSP27) and large (HSP70) heat shock proteins. J Mol Biol 232:1221-1226 LCvtrup S (1977) The phylogeny of the Vertebrata. J Wiley & Sons,

London, pp 180-184

Lcvtrup S (1985) On the classification of the taxon Tetrapoda. Syst Zool 34:463-470

Lu S-F, Pan F-M, Chiou S-H (1995) Sequence analysis of frog c~B- crystallin cDNA: sequence homology and evolutionary comparison of aA, aB and heat shock proteins. Biochem Biophys Res Commun 216:881-891

Marshall CR (1992) Substitution bias, weighted parsimony, and amniote phylogeny as inferred from 18S rRNA sequences. Mol Biol Evol 9:370-373

Morrison R, Mason K, Frost-Mason S (1994) A cladistic analysis of the evolutionary relationships of the members of the tyrosinase gene family using sequence data. Pigment Cell Res 7:388-393 Owen R (1866) On the anatomy of vertebrates, vol I. Longmans,

Green, London, p 7

Philippe H, Douzery E (1994) The pitfalls of molecular phylogeny based on four species, as illustrated by the Cetacea/Artiodactyla relationships. J Mamm Evoi 2:133-152

Quax-Jeuken Y, Quax W, van Rens G, Meera Khan P, Bloemendal H (1985) Complete structure of the aB-crystallin gene: conservation of the exon-intron distribution in the two nonlinked a-crystallin genes. Proc Natl Acad Sci USA 82:5819-5823

Ridley M (1993) Evolution. Blackwell Scientific Publications, Boston

Saitou N, Nei M (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4:406-425 Seutin G, Lang BF, Mindell DP, Morais R (1994) Evolution of the

WANCY region in amniote mitochondrial DNA. Mol Biol Evol 11:329-340

Stapel SO, Leunissen JAM, Versteeg M, Wattel J, de Jong WW (1984) Ratites as oldest offshoot of avian stem--evidence from a-crystallin A sequences. Nature 311:257-259

Steel MA, Lockhart PJ, Penny D (1993) Confidence in evolutionary trees from biological sequence data. Nature 364:440-442 Swofford DL, Olsen GJ (1990) Phylogeny reconstruction. In: Hillis

DM, Moritz C (eds) Molecular systematics. Sinauer, Sunderland, MA, pp 411-501

Tateno Y, Takezaki N, Nei M (1994) Relative efficiencies of the maximum-likelihood, neighbor-joining, and maximum-parsimony methods when substitution rate varies with site. Mol Biol Evol 11:261- 277

Van de Peer Y, Neefs J-M, De Rijk P, De Wachter R (1993) Recon- structing evolution from eukaryotic small-ribosomal-subunit RNA sequences: calibration of the molecular clock. J Mol Evol 37:221- 232

Van de Peer Y, De Wachter R (1994). TREECON for Windows: a software package for the construction and drawing of evolutionary trees for the Microsoft Windows environment. Comput Appl Biosci 10:569-570

Yamamoto H, Kudo T, Masuko N, Miura H, Sato S, Tanaka M, Tanaka S, Takeuchi S, Shibahara S, Takeuchi T (1992) Phylogeny of regu- latory regions of vertebrate tyrosinase genes. Pigment Cell Res 5:284-294

Yasuda A, Kawauchi H, Papkoff H (1990) The complete amino acid sequence of prolactin from the sea turtle

(Chelonia mydas).

Yasuda A, Yamaguchi K, Papkoff H, Yokoo Y, Kawauchi H (1989) The complete amino acid sequence of growth hormone from the sea turtle