Evidence for an ancient chromosomal duplication in Arabidopsis thalianaby sequencing and analyzing a 400-kb contig at the APETALA2 locus on

chromosome 4

Nancy Terryna, Leo Heijnenb, Annick De Keysera, Martien Van Asseldonck1;b,Rebecca De Clercqa, Henk Verbakelb, Jan Gielena, Marc Zabeaub, Raimundo Villarroela,

Taco Jesseb, Pia Neyta, Reneè Hogersb, Hilde Van Den Daelea, Wilson Ardilesa,Christine Schuellerc, Klaus Mayerc, Patrice Deèhaisa, Stephane Rombautsa,

Marc Van Montagua, Pierre Rouzeèa;d;*, Pieter Vosb

aLaboratorium voor Genetica, Departement Genetica, Vlaams Interuniversitair Instituut voor Biotechnologie (VIB), Universiteit Gent,K.L. Ledeganckstraat 35, B-9000 Ghent, Belgium

bKeygene N.V., Postbus 216, AgroBusiness Park 90, Keyenbergseweg 6, NL-6700 AE Wageningen, The NetherlandscMunich Information Center for Protein Sequence, Max-Planck-Institut fuër Biochemie, D-82152 Martinsried, Germany

dLaboratoire Associeè de l'Institut National de la Recherche Agronomique (France), Universiteit Gent, B-9000 Ghent, Belgium

Received 9 January 1999

Abstract As part of the European Scientists SequencingArabidopsis program, a contiguous region (396 607 bp) locatedon chromosome 4 around the APETALA2 gene was sequenced.Analysis of the sequence and comparison to public databasespredicts 103 genes in this area, which represents a gene density ofone gene per 3.85 kb. Almost half of the genes show nosignificant homology to known database entries. In addition, thefirst 45 kb of the contig, which covers 11 genes, is similar to aregion on chromosome 2, as far as coding sequences areconcerned. This observation indicates that ancient duplicationsof large pieces of DNA have occurred in Arabidopsis.z 1999 Federation of European Biochemical Societies.

Key words: Arabidopsis thaliana ; APETALA2 ; Genome;Sequencing

1. Introduction

In plant molecular biology and genetics, Arabidopsis thali-ana has long been recognized as a model organism [1]. By theend of 1993, a project designated European Scientists Se-quencing Arabidopsis (ESSA) was initiated with the aim ofsequencing large fragments of the A. thaliana genome. Cur-rently, a whole international team is working on the comple-tion of that genome [2].

Within the ESSA program, most e¡ort was concentratedinitially on chromosome 4 around the FCA locus [3] and theAPETALA2 (AP2) locus. The ap2 mutant had been described,mapped, and cloned before [4^6]. On the most recent map ofDean and Lister [7], AP2 was found between the markersm214 and g2486 at 93.2 cM (the whole chromosome 4 being116 cM).

Results of the sequencing and analysis of this latter region,as the result of cooperation between two laboratories, will bediscussed here. This region represents the second largest con-tig of Arabidopsis being analyzed in detail, preceded by the2 Mb FCA contig [3].

2. Materials and methods

2.1. Isolation and subcloning of an AP2-containing yeast arti¢cialchromosome (YAC)

The CIC YAC clone 7A10, size 420 kb (from the CIC A. thaliana(L.) Heynh. ecotype Columbia library) was isolated by using AFLP(¢led by Keygene N.V.) markers [8,9] derived from sequences of theAP2 gene and subcloned in the cosmid vector pCLD04541 [10], usinga partial Sau3AI digest. To isolate YAC-speci¢c cosmids, a totalnumber of 16 000 Escherichia coli clones were hybridized to gel-puri-¢ed YAC DNA. Approximately 450 YAC-speci¢c clones were iden-ti¢ed. AFLP ¢ngerprinting [9] was used to build a cosmid contig. Thisapproach resulted in a cosmid contig of approximately 260 kb.

2.2. Isolation of bacterial arti¢cal chromosome (BAC) clonesextending the cosmid contig

The BAC clones were isolated through hybridization of the BAClibrary with a probe containing the inserts of all cosmids of the 260-kbcontig, followed by contig building by AFLP. All AFLP markerspresent in the cosmid contig were also present in the BAC contig,indirectly proving colinearity of the cosmid contig with the genomicsequence. TAMU 8H13 and TAMU 10C14, which overlapped withthe cosmid contig, were selected for further sequencing (Fig. 1).

2.3. Construction of cosmid and BAC subclonesDNA from cosmids and BAC clones was sheared either by sonica-

tion (Misonix Inc., Farmingdale, NY; type XL2020) or by nebulizing(Lifecare Hospital Supplies, Harborough, UK). A 1.8^2.2-kb end-re-paired fraction was isolated and ligated in pUC18/SmaI/BAP (Phar-macia, Uppsala, Sweden) or a modi¢ed pUC19 vector (BamHI/SalIfragment replaced by a StuI-, SpeI-, and SalI-containing fragment).Individual colonies of the transformation were grown in 96- or 384-well microtiter plates.

FEBS 21569 2-3-99

0014-5793/99/$19.00 ß 1999 Federation of European Biochemical Societies. All rights reserved.PII: S 0 0 1 4 - 5 7 9 3 ( 9 9 ) 0 0 0 9 7 - 6

*Corresponding author. Fax: (32) (9) 264 5349.E-mail: pirou@gengenp.rug.ac.be

1Present address : University Hospital Nijmegen, Department ofHuman Genetics, Geert Grootteplein 24, NL-6500 HB Nijmegen,The Netherlands.

Abbreviations: AFLP, amplified fragment length polymorphism (is atrademark in the Benelux); BAC, bacterial artificial chromosome;EST, expressed sequence tag; YAC, yeast artificial chromosome

The genomic sequence of the contig has been submitted to theGenBank under accession numbers Z99707 and Z99708 (with 8361-bpoverlap). Cognate cDNAs have accession numbers AJ002596,AJ002597, and AJ002598.

FEBS 21569 FEBS Letters 445 (1999) 237^245

2.4. Sequencing strategyThe sequence of the partially overlapping cosmids 3A6, 4B6, 4E12,

5C9, 2H2, 5E2, 2F3, 2H7, and BAC TAMU 10C14 was determined ina non-random approach by sequencing the cosmid or BAC ends, aswell as a few random subclones. Primers were designed to isolateprimer-speci¢c clones from pools of subclones. A random sequencingapproach was taken for cosmids 2C7, 3A6, 4F4, 3F11, 6B5, 5E3,4G12, 4E3, and BAC TAMU 13H8. The ABI PRISM dye terminatorcycle sequencing ready reaction kit was used mostly (Perkin-Elmer,Foster City, CA). The reaction products were analyzed on an ABIPrism 377. The sequence with the corresponding electropherogramswere assembled into contigs using a home-made computer programcalled Sequence Assembly Facility Environment (SAFE; [11]) or the1994 version of the Staden sequence analysis program [12]. The meanredundancy of the assembled sequence is 5.

2.5. Sequence comparisonsSequence analysis was done initially by the Martinsried Institute for

Protein Sequences (MIPS) center (Martinsried, Germany) and re¢nedby the bioinformatics team of the Laboratory of Genetics (Ghent,Belgium). To this end, the sequence was cut into pieces of 7000 bpwith 1000-bp overlaps. Each piece was submitted to a BLASTX/non-redundant protein and BLASTN/non-redundant DNA search as wellas a BLASTN/EST search (mostly on the Beauty BCM server). Fromthe resulting ¢les, the homologues with the highest score and a reliableannotation were looked for.

When a homologue was found with a high score (P(N)6E3100)and with homology over its whole length, its protein sequence wasfollowed to check whether the transcript did not show any frameshiftsor stop codons and whether the intron borders were correct. Whenany doubt arose, NetGene2 predictions were used [13,14].

For genes with weak homologies, di¡erent prediction programswere used according to the speci¢c problems: NetStart when the startcodon was not obvious, GeneMark [15] together with GenScan [16]for exon predictions, GeneMark for exon frame predictions, GenScanto check that small exons were not missed, and NetGene2 for intronborder prediction. When an expressed sequence tag (EST) was foundthrough BLASTN/EST (most of the time for the 5P or 3P parts of thegenes), the EST was used to complement the information obtained byBLASTX.

When no homologues were found through BLASTX, and no ESTswere available, we relied upon prediction programs, namely GenScanand GeneMark together to locate potential exons, NetGene2 for the

intron borders, and NetStart [17] for the position of the start codon.In all cases, the coding sequence (CDS) was reconstructed, translated,and submitted to BLASTP, for a last homology search and check forgaps.

Updated FASTA searches done at MIPS on the genes in this contigcan be found at http://speedy.mips.biochem.mpg.de/arbi/data/ap2_contig.html. Full annotations as well as various genomic featurescan also be found at the Ghent site (http://spider.rug.ac.be/public/seq/ap-2.html).

3. Results and discussion

3.1. General overview of the contigFig. 1 gives an overview of the clones used for the sequenc-

ing program. Initially cosmid clones were used, but during theproject BAC clones became available and these were used tocontinue the contig sequencing (see Section 2).

The 103 predicted genes that are present in this region aresummarized in Tables 1 and 2 and in Fig. 2. Table 1 providesinformation on their putative function, highest related entry inthe public databases as well as EST sequences that correspondto the putative genes whereas in Table 2 the genes are classi-¢ed based on their homology.

In conclusion, the putative role of approximately half of thegenes could be established by sequence similarity to knowngenes. These genes have been classi¢ed into 15 classes accord-ing to their putative cellular role that will be used to describegenes identi¢ed in the genome program [3]. This list ispreliminary and new categories and subcategories will beadded as more of the genome is sequenced (updated versionavailable at http://muntjac.mips.biochem.de/arabi/fca/gene/funcat_table.html).

3.2. Annotation and re-annotation processAfter the ¢rst MIPS automatic annotation, manual re-an-

notation showed discrepancies on about four genes out of

FEBS 21569 2-3-99

Fig. 1. Chromosomal location of the AP2 gene. Overview of the cosmids and BACs used for sequencing. The upper line represents the chromo-somal region (approximately 10 cM) with some known markers in the region. The sequenced cosmids and BAC clone are indicated as well asthe region covered by the two database submissions covering this region (Z99707 and Z99708). The sequence Z99707 (206 440 bp) contains 50genes (1^50) with an 8361-bp overlap with Z99708 (198 555 bp) that contains 53 genes (51^103).

N. Terryn et al./FEBS Letters 445 (1999) 237^245238

FEBS 21569 2-3-99

Tab

le1

Ove

rvie

wof

the

103

pred

icte

dge

nes

inth

e39

7-kb

cont

ig

No.

SaF

unct

ion/

prod

uct

Pos

itio

nB

est

hom

olog

ueac

cess

ion

Nam

eSp

ecie

sB

LA

STP

scor

eE

STB

LA

STN

/EST

scor

eC

lass

bC

ateg

oryc

1W

Cyt

ochr

ome

P45

06

1^56

0P

2446

5cy

toch

rom

eP

450

avoc

ado

1.6E

-25

H76

015

2.1E

-76

312

2W

Cyt

ochr

ome

P45

097

8^27

31P

2446

5cy

toch

rom

eP

450

avoc

ado

7.6E

-60

H76

015

3.6E

-65

212

3W

Cyt

ochr

ome

P45

031

15^5

137

P24

465

cyto

chro

me

P45

0av

ocad

o1.

5E-6

7A

A39

5149

1.3E

-149

212

4W

Cyt

ochr

ome

P45

059

94^7

943

P24

465

cyto

chro

me

P45

0av

ocad

o1.

6E-2

1T

4364

05.

9E-4

73

125

WC

ytoc

hrom

eP

450

8851

^11

532

P24

465

cyto

chro

me

P45

0av

ocad

o1.

0E-6

8F

1357

36.

4E-1

042

126

CU

nkno

wn

1218

6^12

879

N38

505

9.9E

-131

513

7W

Unk

now

n17

504^

1775

5A

C00

2391

unkn

own

Ara

bido

psis

413

8W

Unk

now

n18

322^

2093

6D

1542

1R

ICE

5.7E

-22

613

9W

Cat

ion-

tran

spor

ting

AT

P-a

se21

332^

2569

7P

3033

6C

u-tr

ansp

orti

ngA

TP

ase

mou

se3.

3E-1

4Z

3373

12.

0E-1

043

7,22

10C

Myb

-rel

ated

2612

0^27

082

Z54

136

Myb

-rel

ated

Ara

bido

psis

4.2E

-69

T43

211

5.5E

-35

24.

19.0

411

WK

inas

e-lik

ere

cept

or37

499^

3989

5P

3354

3re

cept

orki

nase

TM

K1

Ara

bido

psis

4.5E

-50

H76

836

1.4E

-135

310

,01

12C

Unk

now

n42

632^

4306

6H

7666

34.

6E-3

45

1313

CP

seud

o-ge

ne50

097^

5130

0S0

4132

phot

osys

tem

IIox

ygen

-evo

lvin

gco

mpl

ex

pea

D49

160

RIC

E8.

2E-2

23

2,2

14C

Col

dac

clim

atio

npr

otei

n51

715^

5257

8U

7321

6co

ldac

clim

atio

nT

riti

cum

aest

ivum

4.4E

-56

AA

6506

471.

3E-3

52

11,0

5

15C

Put

ativ

ehi

ston

ebi

ndin

g53

662^

5530

8A

2568

0nu

clea

rhi

ston

ebi

ndin

gA

fric

ancl

awed

frog

3.3E

-07

313

16W

Thi

ored

oxin

5645

4^57

848

P73

920

thio

disu

l¢de

inte

r-ch

ange

prot

ein

Syn

echo

cyst

is4.

4E-2

3A

A39

4562

1.9E

-150

32,

2

17C

Put

ativ

ece

lldi

visi

onre

gula

tor

5817

2^60

529

U80

043

mis

ato

gene

Dro

soph

ilam

elan

o-ga

ster

3.5E

-15

R64

911

5.4E

-110

33,

22

18C

Unk

now

n61

577^

6233

5or

6300

96

13

19W

Glu

-t-R

NA

6682

1^66

868

K00

193

Glu

-tR

NA

Dro

soph

ilam

elan

o-ga

ster

34,

19

20W

Unk

now

n67

029^

6910

4F

1544

22.

1E-3

06

1321

WP

ecti

nest

eras

e70

560^

7286

4Z

9405

8pe

ctin

este

rase

tom

ato

1.6E

-123

H76

007

1.3E

-106

21,

059,

0122

WH

ydro

xyni

trile

lyas

e73

627^

7469

9Z

3427

1pi

r7a

rice

3.3E

-39

AA

6514

822.

5E-2

62

11,0

223

WP

seud

o-ge

nehy

drox

ynit

rile

lyas

e74

956^

7556

9Z

3427

1pi

r7b

rice

2.5E

-27

R30

024

9.5E

-20

311

,02

24C

Unk

now

n75

724^

7811

7U

6383

9nu

cleo

pori

nra

t2.

3E-0

2A

A23

1701

6.9E

-18

613

25C

Unk

now

n78

561^

8076

1A

C00

1229

unkn

own

Cae

norh

abdi

tis

eleg

ans

0.0E

+00

AA

3949

562.

9E-1

174

1326

WU

nkno

wn

8148

7^83

341

AA

6513

306.

3E-0

16

1327

WU

nkno

wn

8582

1^88

511

Z69

793

unkn

own

Cae

norh

abdi

tis

eleg

ans

1.2E

-08

T21

702

7.3E

-19

413

28C

Unk

now

n89

125^

9034

66

1329

CU

nkno

wn

9347

6^90

968

613

30*

WP

atat

inho

mol

ogue

100

626^

102

768

P15

478

pata

tin

toba

cco

9.2E

-57

H35

955

2.3E

-128

26,

231

WP

atat

inho

mol

ogue

103

950^

106

111

AJ0

0259

6pa

tati

nto

bacc

o4.

7E-2

28H

3595

62.

6E-1

063

6,2

32W

Pat

atin

hom

olog

ue10

804

2^11

042

3A

J002

596

pata

tin

toba

cco

1.1E

-153

H35

957

7.5E

-29

36,

233

CM

ethi

onyl

-am

ino-

pept

idas

e-lik

e11

059

9^11

250

9A

L00

8883

met

hion

ylam

ino-

pept

idas

eye

ast

8.8E

-70

AA

3946

716.

1E-2

32

6,07

34C

Unk

now

n11

333

7^11

553

46

1335

CU

nkno

wn

120

506^

121

725

AA

3947

792.

0E-1

545

1336

CC

altr

acti

n-lik

e12

230

4^12

326

3P

4121

0ca

ltra

ctin

Atr

iple

xnu

mm

ular

ia2.

0E-5

8A

A39

5741

6.5E

-136

29,

043,

25

N. Terryn et al./FEBS Letters 445 (1999) 237^245 239

FEBS 21569 2-3-99

Tab

le1

(con

tinu

ed)

Ove

rvie

wof

the

103

pred

icte

dge

nes

inth

e39

7-kb

cont

ig

No.

SaF

unct

ion/

prod

uct

Pos

itio

nB

est

hom

olog

ueac

cess

ion

Nam

eSp

ecie

sB

LA

STP

scor

eE

STB

LA

STN

/EST

scor

eC

lass

bC

ateg

oryc

37C

Unk

now

n12

392

2^12

505

76

1338

CH

eat

shoc

kfa

ctor

125

978^

127

024

U68

017

HS

fact

orH

SF4

Ara

bido

psis

2.2E

-188

R90

161

0.0E

+00

111

,05

4.19

.04

39W

Unk

now

n13

067

2^13

399

86

1340

WU

nkno

wn

136

224^

137

857

613

41C

Rib

onuc

leop

rote

in13

824

3^14

033

3S4

0774

ribo

nucl

eopr

otei

nX

enop

us7.

8E-2

2T

4334

31.

6E-0

53

4,22

42W

Kin

ase

143

838^

147

989

AB

0007

98pr

otei

nki

nase

sim

ilar

toN

PK

1A

rabi

dops

is1.

3E-2

3Z

2666

08.

9E-3

23

10.0

4.04

43C

Put

ativ

eni

coti

nate

phos

phor

ibos

yl-

tran

sfer

ase

148

171^

150

743

Z99

120

nico

tina

teph

osph

o-ri

bosy

ltra

nsfe

rase

Bac

illus

subt

ilis

5.3E

-47

N65

739

3.2E

-93

313

44C

Unk

now

n15

129

9^15

351

4A

C00

2330

2.9E

-07

C21

906

3.1E

-08

613

45C

AP

216

454

1^16

668

3U

1254

6ap

etal

a2A

rabi

dops

is1.

6E-2

831

4.19

.04

46W

Unk

now

n17

446

3^17

666

5D

6399

9de

hydr

ogen

ase

Syn

echo

cyst

is1.

8E-1

263

1347

*C

TIN

Y-l

ike

178

076^

178

665

AJ0

0259

8T

INY

Ara

bido

psis

2.6E

-182

AA

4048

104.

0E-1

152

4.19

.04

48W

Unk

now

n18

586

8^18

805

0Z

4642

94.

7E-7

65

1349

WC

yste

ine

prot

eina

se19

150

1^19

298

9P

2525

1cy

stei

nepr

otei

nase

rape

7.8E

-198

R84

153

4.3E

-81

26,

1350

CP

utat

ive

hom

eoti

cpr

otei

n19

395

8^19

825

6A

5763

2B

EL

1A

rabi

dops

is6.

3E-8

02

4.19

.04

51W

Unk

now

n83

63^1

104

9A

C00

3000

6.7E

-89

N65

197

2.4E

-124

513

52W

Unk

now

n13

796^

1620

2S5

7377

prob

able

mem

bran

epr

otei

nye

ast

4.4E

-13

H77

118

1.8E

-104

513

AA

6055

073.

9E-1

2053

CU

nkno

wn

1673

6^17

917

H77

079

7.2E

-11

613

54C

Unk

now

n19

236^

2010

5T

4588

19.

3E-1

475

1355

CU

nkno

wn

2162

2^22

727

U41

558

unkn

own

Cae

norh

abdi

tis

eleg

ans

7.5E

-11

L46

423

1.3E

-48

413

56C

Ger

anyl

gera

nyl-

pyro

-ph

osph

atas

esy

ntha

se24

988^

2610

3P

3480

2ge

rany

lger

anyl

-pyr

o-ph

osph

atas

esy

ntha

seA

rabi

dops

is1.

0E-1

79L

3747

71.

0E-0

81

20,2

57W

Ubi

quit

in-c

onju

gati

ngpr

otei

n27

468^

2837

8P

5249

1ub

iqui

tin-

conj

ugat

ing

prot

ein

yeas

t4.

0E-3

4T

4155

00.

0E+

003

6,07

58C

Unk

now

n31

311^

3325

5A

J001

694

mem

bran

epr

otei

nT

herm

otog

am

arit

ima

3.1E

-08

413

59W

Unk

now

n35

403^

3662

5R

6479

21.

0E-1

625

1360

WG

luco

sylt

rans

fera

se38

024^

3939

8Q

4028

7U

TP

-glu

cose

gluc

osyl

-tr

ansf

eras

eca

ssav

a9.

6E-1

062

1,05

61C

Am

inop

epid

ase

3963

6^42

894

AF

0385

91X

-pro

amin

opep

tida

sera

t5.

9E-1

35N

9600

83.

0E-6

82

6,13

62C

Phy

coer

ythr

in-l

ike

prot

ein

4340

1^44

974

D45

900

Led

i-3

Lit

hosp

ernu

mer

ythr

orhi

zon

1.50

E-4

9R

6494

91.

0E-1

582

12

63W

Put

ativ

eho

meo

tic

prot

ein

5330

3^54

968

X94

947

hom

eoti

cpr

otei

nto

mat

o2.

0E-2

33

4.19

.04

64W

G-b

oxbi

ndin

gG

BF

157

865^

5976

7X

6389

4G

-box

bind

ing

GB

F1

Ara

bido

psis

1.0E

-162

H37

673

3.0E

-43

14.

19.0

49,

165

CP

seud

o-ge

ne59

949^

6184

8Q

0076

5T

B2

hum

an2.

0E-2

03

1366

CSc

arec

row

hom

olog

ue62

097^

6355

7U

6279

8sc

arec

row

Ara

bido

psis

1.0E

-23

N65

163

1.0E

-175

34.

19.0

467

WP

utat

ive

stor

age

prot

ein

6928

0^71

174

Z54

364

vici

linM

atte

ucci

ast

ruth

iopt

eris

4.0E

-38

Z48

554

4.0E

-70

36,

268

WSp

licin

gfa

ctor

7200

8^75

478

S202

50sp

licin

gfa

ctor

hum

an8.

0E-7

0N

9670

41.

0E-1

123

4,22

69*

WP

utat

ive

salt

indu

cibl

e75

900^

7713

8A

J002

597

puta

tive

salt

indu

cibl

eA

rabi

dops

is0.

0E+

00R

6549

21.

0E-1

683

11,0

570

WT

rans

port

prot

ein

8013

4^81

937

Q10

286

suga

rtr

ansp

orte

rbe

etro

ot1.

6E-4

53

7,07

71W

Unk

now

n82

336^

8326

7H

3638

23.

0E-1

26

1372

WP

utat

ive

tran

scri

ptio

nin

itia

tion

fact

or84

089^

8633

3P

2905

3tr

ansc

ript

ion

init

iati

onfa

ctor

IIB

Afr

ican

claw

edfr

og5.

0E-2

63

4,19

N. Terryn et al./FEBS Letters 445 (1999) 237^245240

FEBS 21569 2-3-99

Tab

le1

(con

tinu

ed)

Ove

rvie

wof

the

103

pred

icte

dge

nes

inth

e39

7-kb

cont

ig

No.

SaF

unct

ion/

prod

uct

Pos

itio

nB

est

hom

olog

ueac

cess

ion

Nam

eSp

ecie

sB

LA

STP

scor

eE

STB

LA

STN

/EST

scor

eC

lass

bC

ateg

oryc

73W

Unk

now

n91

169^

9242

9U

8806

8se

c14

rice

5.0E

-21

R65

425

1.0E

-120

513

H76

207

3.0E

-74

74W

Unk

now

n93

093^

9752

76

1375

WU

nkno

wn

9995

3^10

070

9(z

inc

¢nge

rdo

mai

n)3

4.19

.01

76W

Unk

now

n10

234

1^10

407

0U

8388

1hy

drol

ase

Pse

udom

onas

313

77C

Put

ativ

eL

EA

prot

ein

104

647^

105

947

P20

075

LE

Aso

ybea

n2.

7E-1

7Z

2985

41.

0E+

131

312

78W

MA

DS

box

prot

ein

107

427^

108

470

Y12

776

Pic

eaab

ies

2.0E

-33

34.

19.0

479

CU

nkno

wn

108

955^

111

658

Z48

583

unkn

own

Cae

norh

abdi

tis

eleg

ans

1.0E

-83

413

80C

Unk

now

n11

456

8^11

495

66

1381

CU

nkno

wn

116

627^

116

959

613

82W

Unk

now

n12

190

0^12

421

86

1383

CU

nkno

wn

124

653^

126

204

HL

Hm

otif

AA

3950

501.

0E-1

385

1384

WU

nkno

wn

127

848^

129

498

D90

906

hypo

thet

ical

prot

ein

Syn

echo

cyst

is7.

0E-5

54

1385

WU

nkno

wn

134

183^

136

498

Q07

283

tric

hohy

alin

hum

an3.

0E-0

83

1386

CU

nkno

wn

140

167^

140

841

613

87W

Unk

now

n14

311

1^14

347

9N

9711

80.

0E+

005

1388

CU

nkno

wn

144

810^

147

488

P24

859

Sec1

4(P

I/P

C)

tran

spor

tpr

otei

nK

luyv

erom

yces

lact

is4.

0E-4

7A

A39

5164

1.0E

-08

313

89C

Put

ativ

eac

yltr

ansf

eras

e14

849

4^15

102

0X

9564

1se

rine

C-p

alm

itoy

ltr

ansf

eras

em

ouse

2.0E

-93

R90

586

8.0E

-81

31,

06

90W

Unk

now

n15

314

3^15

449

06

1391

WU

nkno

wn

156

519^

156

956

613

92W

MA

Pki

nase

158

205^

159

373

Q39

027

MA

Pki

nase

7A

rabi

dops

is0.

0E+

00T

4624

41.

0E-6

42

10.0

4.04

93W

Unk

now

n16

051

2or

161

813

T88

343

1.0E

-64

613

160

186

or16

261

894

WP

erox

idas

e16

370

1^16

497

0X

9880

4pe

roxi

dase

Ara

bido

psis

1.0E

-109

F13

569

1.0E

+11

32

1295

WP

utat

ive

ribo

som

alpr

otei

n16

536

1^16

590

0P

3621

2L

12ri

boso

mal

prot

ein

Ara

bido

psis

2.0E

-11

Z18

520

1.0E

-107

35,

01

AA

0421

500.

0E+

0096

CU

biqu

itin

-con

juga

ting

prot

ein

166

728^

167

649

P42

743

ubiq

uiti

n-co

njug

atin

gen

zym

eA

rabi

dops

is7.

0E-6

8A

A72

8692

6.0E

-85

32,

13

97C

Act

in-i

nter

acti

ngpr

otei

n16

914

6^17

203

7P

4668

1A

IP2

prot

ein

yeas

t1.

0E-1

44A

A04

2437

1.0E

-70

29,

04

98W

Unk

now

n17

284

1^17

487

2Q

0931

6un

know

nC

aeno

rhab

diti

sel

egan

s1.

0E-1

214

1399

WC

ytoc

hrom

eP

450

177

617^

181

645

S553

79cy

toch

rom

eP

450

Ara

bido

psis

2.0E

+91

N96

214

0.0E

+00

220

,110

0C

Unk

now

n18

200

8^18

236

86

1310

1W

L-G

alac

tosi

dase

prec

urso

r18

848

4^19

277

8P

4558

2L

-gal

acto

sida

sepr

ecur

sor

tom

ato

7.6E

-256

Z37

275

1.0E

-179

21,

05

102

WA

cid

phos

phat

ase

193

761^

195

881

AL

0221

41ac

idph

osph

atas

eA

rabi

dops

is7.

5E-1

202

10.0

4.07

103

WU

nkno

wn

196

431s

198

459

613

aS,

stra

nd.

bT

heho

mol

ogy

clas

si¢c

atio

nca

nbe

foun

din

Tab

le3.

cT

hefu

ncti

onal

cate

gory

tobe

foun

dat

the

MIP

Sw

ebsi

te(h

ttp

://m

untj

ac.m

ips.

bioc

hem

.de/

arab

i/fca

/gen

e/fu

ncat

_tab

le.h

tml)

.In

the

last

colu

mn

puta

tive

cogn

ate

EST

sar

ein

dica

ted

(s95

%id

enti

ty);

*w

ere

fully

sequ

ence

dan

dsu

bmit

ted

toth

eG

enB

ank

asco

gnat

ecD

NA

sw

ith

acce

ssio

nnu

mbe

rsA

J002

596^

AJ0

0259

8.T

henu

mbe

ring

isdo

neac

cord

ing

toth

eE

MB

L/G

enB

ank

entr

ies

Z99

707

(gen

es1^

50)

and

Z99

708

(gen

es51

^103

).

N. Terryn et al./FEBS Letters 445 (1999) 237^245 241

¢ve. In addition, the re-annotation pointed to a few frame-shifts caused by sequence errors that have been corrected thisway. Similar issues have been raised for annotation of bacte-rial genomes [18], but to a lesser extent. As the task of anno-tation of a higher eukaryote genome is much more di¤cultbecause of the larger size of the genome, the lesser gene con-tent, the split gene structure, and less homologies to knowndatabase entries, such a result is not surprising. It is clearly awarning for caution when using the present-day annotations,and it indicates that complete re-annotation of the genomewill be needed.

3.3. Statistical analysis of the contigAs can be deduced from Table 1, 59 genes can be found on

the Watson and 44 on the Crick strand. Gene density is quitehigh (3.85 kb/gene) compared to that of the FCA region [3]

(4.8 kb/gene) and the mean density of 4.1 kb/gene reported for6.7 Mb sequenced on chromosome 5 [19]. Nevertheless, re-gions of 1.2 Mb on chromosome 5 have been reported witha similarly high gene density of 3.84 kb/gene [20]. In total,42% of the genes are highly similar to Arabidopsis EST se-quences (s 95% similarity).

Table 3 represents a statistical analysis done on bothstrands of the 400-kb contig in terms of occurrence and sizeof genes, introns, and exons. Intergenic regions cover approx-imately 53% of the contig, whereas introns represent 16% andcoding sequences 31%. As a mean, the ¢rst intron is longerthan the others and the ¢rst and last exons longer than themiddle ones. It is noteworthy that local as well as strandheterogeneities were observed. For example, genes in the ¢rsthalf of the sequence have smaller introns than those in thesecond half of the contig, whereas exons have a similar aver-

FEBS 21569 2-3-99

Fig. 2. Schematic overview of the exons of the 103 genes present in the 379-kb contig. Genes are marked by numbers; each number representsa new gene. Gene 13 is not indicated.

Table 2Classes of similarities to genes

Class BLAST E value Type of matching protein Number Predicted function

1 identical same protein 4 42 6 10350 known protein 23 233 between 10310 and 10350 known protein 33 284 6 10310 hypothetical protein 8 ^5 s 10310 none, but has cognate EST match 11 ^6 6 150 none, no cognate EST match 24 ^

Total 103 55

N. Terryn et al./FEBS Letters 445 (1999) 237^245242

age size (data not shown). In addition, both exons and intronson the Watson strand are larger than those on the Crickstrand.

3.4. Finding of two U12-type intronsThe 102 protein-encoding genes were found or predicted to

contain 429 introns. All of these introns, except two, have theconsensus sequences of classical U2-type introns. The remain-ing two are the last intron (i4) from gene 15, which is distantlyrelated to the histone-binding protein from Xenopus, and thelast intron (i9) from gene 52, which is clearly similar to a smallyeast gene family encoding membrane proteins of unknownfunction. Interestingly, a paralogue of gene 52 was found inanother contig from chromosome 4 (al022224/al021637), butthis gene does not seem to have any intron. These introns,although having GT-AG borders, display the distinctive fea-tures of U12-type introns with their very conserved donor andbranch sites and the typical short distance between the poten-tial branch point and the acceptor site with no polypyrimidinetract, as shown in Fig. 3. Such introns were initially found inanimals where they have been shown to need speci¢c snRNAsfor their splicing. More recently, they have been found inplants too [21,22] and are considered to be of rare occurrence.The ¢nding of two members out of a total 429 may give a ¢rstestimation of their actual frequency in the Arabidopsis ge-nome. For the time being, these introns are not correctlypredicted by any gene prediction program.

3.5. Gene clusteringThere are three clusters of tandem gene repeats in the con-

tig. A cluster of ¢ve P450 genes is found at its 5P extremity.These genes have the same genomic structure (three exons)and their coding sequences are very similar to each other(74^83% similarity between copies 1^4, 60% between themand copy 5), their best homologue being an elicitor-inducedP450 from licorice. There is another P450 gene (gene 99) atthe other end of the contig that di¡ers from the genes in thecluster (low homology, 10 exons). The clustering of these ¢veP450 genes suggests that they might originate from a commonancestor late in evolution and could probably be involved indi¡erent steps of the same pathway (secondary metabolism,defence, etc.).

Three patatin genes (genes 30, 31, and 32) are also foundclustered. Patatin is the major storage protein of potato andhomologues have already been found in other species, but notyet in Arabidopsis. The patatins found in this cluster are sim-ilar to each other, the ¢rst and second copy being very close(90%). The third copy is more distant (67^69% similarity tothe others) and di¡ers speci¢cally at the N-terminus, suggest-ing a di¡erent subcellular localization of this member.

The third cluster is an imperfect tandem repeat of a geneencoding a protein with strong similarity to hydroxynitrilelyase, an enzyme that produces cyanide by hydrolyzing cya-nogen glucosides, which are secondary metabolites producedin a narrow range of plants. Only the ¢rst repeat contains acomplete and potentially functional gene with three exons.The gene in the second repeat seems truncated after the ¢rstexon, and would thus be a pseudo-gene. Paralogues of thisgene have been found in the FCA region [3] ; the ¢nding ofsuch a gene is unexpected, because Arabidopsis has not beenreported to produce cyanogen glucosides. It would be inter-esting to check whether these genes are functional and in-duced by pathogens and/or predator attack, and to examinewhich substrate their products would hydrolyze.

3.6. Genome duplicationDuring the process of gene search, several genes at the 5P

extremity of the contig were observed to have homologueslocated in the AC002391 contig from chromosome 2. To

FEBS 21569 2-3-99

Table 3Statistical analysis of the AP2 contig

Total Number Mean per gene Total size (bp) Mean size (bp) % W strandc (+) C strandc (3)

Genic regionsRNA CDS 1 71Protein CDSa 100 121 289 1214 31.2 56* 44*Without introns 15 13 707 914 7* 8*With introns 85 107 692 1270 49* 36*Intronsa;b

Total 429 5.05 63 530 148 16.2 160 133First 85 16 785 197 248 129Rest 344 46 745 136 137 134Exonsa;b

Total 526 6.19 107 692 205 230 180First 85 25 961 305 314 293Rest 346 55 625 161 176 141Last 85 26 106 307 364 229Intergenic regions

102 208 556 2045 52.6aStatistics on full-length genes.bMeans refer to intron-containing genes; for all genes, the mean numbers of introns and exons per gene are 4.29 and 5.26, respectively.cW and C strand columns refer to sizes, except for ¢gures followed by an asterisk, where they represent numbers. The region analyzed is the whole396-kb contig, except for gene 13, for which the prediction is uncertain (a pseudo-gene).

Fig. 3. U12 class introns. g15_i4, U12-type intron #4 of gene 15from this contig; g52_i9, U12-type intron #9 of gene 52 from thiscontig; p120_i6, U12-type intron #6 from the human p120 gene.The intron sequence is given in lowercase. L marks the branchpoint.

N. Terryn et al./FEBS Letters 445 (1999) 237^245 243

check the gene arrangement, a dot-plot comparison of bothcontigs was performed using the GCG compare/dot-plot andthe dotter [23] programs. As seen in Fig. 4, the ¢rst 45 kb(Watson strand) of the contig show similarities to a region ofcomparable size of the AC002391 contig (Crick strand,77 000^32 000). The extent of the duplication is perhaps larger,because no sequence is available yet on the 5P side of the AP2contig described here. The homology is patchy, being mostly

restricted to the coding sequence of the homologous genes.Among the 12 genes found in this region on the contig,nine have a counterpart in the chromosome 2 contig, theorder and the orientation of the genes being conserved. Thissituation strongly suggests an ancient duplication of this re-gion in the Arabidopsis genome. This duplication must indeedbe ancient, because most non-coding sequences (introns, inter-genic sequences) are not conserved and because the two copies

FEBS 21569 2-3-99

Fig. 4. Dot-plot comparison of a region on chromosome 2 (AC002391, bp 15 000^75 000) with the ¢rst 60 kb of the sequence described here(Z99707). Genes are numbered as in Table 1. Asterisks indicate non-homologous genes, the arrows inside the plot mark the regions of homolo-gous genes.

N. Terryn et al./FEBS Letters 445 (1999) 237^245244

di¡er by several insertions/deletions of various genes. Onesuch deletion is found in the chromosome 2 counterpart ofgene 9 for which homologies are found with several metal(Cu, Cd) transporters. On chromosome 4, gene 9 is predictedwith 13 exons, encoding a 819-amino acid protein. On chro-mosome 2, the corresponding gene is truncated on its 5P sideand is predicted to have three exons encoding a 221-aminoacid protein. The P450 cluster described above is the 5P-mostpart of the duplicated region. Whereas the chromosome 4copy contains ¢ve P450 members in tandem repeat, onlytwo P450 members are found in the chromosome 2 counter-part, separated by an insertion of two foreign genes and£anked by a P450 of di¡erent origin. The genetic distancesbetween the di¡erent P450 members suggests that the dupli-cation of the P450 genes on chromosome 2, and at least theduplications responsible for the four ¢rst copies on chromo-some 4, occurred after the duplication of the 45-kb regionitself. Although the Arabidopsis genome is of small size, du-plication of individual genes appears to be frequent. Duplica-tion on a larger scale has been suggested from mapping data[24], but, to our knowledge, this is the ¢rst demonstration of aduplication of a large region in Arabidopsis. The completionof the Arabidopsis genome sequence will tell us to which ex-tent these duplications occur. A comparative analysis of therepeats in various ecotypes and neighbor species will informus when this happened in the Arabidopsis evolutionary his-tory.

Acknowledgements: The authors wish to acknowledge the ABRCStock Center in Ohio for the distribution of ESTs, M. Bevan forcoordinating the ESSA program, H.W. Mewes, S. Klostermann,and N. Chalwatzis from MIPS for computer data analysis, PeterBreyne and Koen Goethals for critical reading of the manuscript,M. De Cock for help preparing it, and Rebecca Verbanck for the¢gures. This work was supported by the EU (Biotech ERBB102-CT93-0075 and ERBB104-CT96-0338) and by the Flemish govern-ment (VLAB-COT).

References

[1] Goodman, H.M., Ecker, J.R. and Dean, C. (1995) Proc. Natl.Acad. Sci. USA 92, 10831^10835.

[2] Arabidopsis Genome Initiative (1997) Plant Cell 9, 476^478.[3] Bevan, M., Bancroft, I., Bent, E., Love, K., Pi¡anelli, P., Good-

man, H., Dean, C., Bergkamp, R., Dirkse, W., Van Staveren, M.,Stiekema, W., Drost, L., Ridley, P., Hudson, S.-A., Patel, K.,Murphy, G., Wedler, H., Wedler, E., Wanbutt, R., Weitzenegger,T., Pohl, T., Terryn, N., Gielen, J., Villarroel, R., De Clercq, R.,Van Montagu, M., Lecharny, A., Kreis, M., Lao, N., Kavanagh,

T., Hempel, S., Kotter, P., Entian, K.-D., Rieger, M., Scholfer,M., Funk, N., Muller-Auer, S., Silvey, M., James, R., Montfort,A., Pons, A., Puigdomenech, P., Douka, A., Voukelatou, E.,Milioni, D., Hatzopoulos, P., Piravandi, E., Obermaier, B., Hil-bert, H., Duesterhoeft, A., Moores, T., Jones, J., Eneva, T.,Palme, K., Benes, V., Rechman, S., Ansorge, W., Cooke, R.,Berger, C., Delseny, M., Volckaert, G., Mewes, H.-W., Schueller,C. and Chalwatzis, N. (1998) Nature 391, 485^488.

[4] Koornneef, M., van Eden, J., Hanhart, C.J., Stam, P., Braaksma,F.J. and Feenstra, W.J. (1983) J. Hered. 74, 265^272.

[5] Bowman, J.L., Smyth, D.R. and Meyerowitz, E.M. (1989) PlantCell 1, 37^52.

[6] Jofuku, K.D., den Boer, B.G.W., Van Montagu, M. and Oka-muro, J.K. (1994) Plant Cell 6, 1211^1225.

[7] Dean, C. and Lister, C. (1996) in: Weeds World: The Interna-tional Electronic Arabidopsis Newsletter (Anderson, M., Ed.),Vol. 3 (iii), pp. 1^6, Nottingham Arabidopsis Stock Centre, Not-tingham.

[8] Zabeau, M. and Vos, P. (1993) European Patent Application EP534858A1.

[9] Vos, P., Hogers, R., Bleeker, M., Reijans, M., van de Lee, T.,Hornes, M., Frijters, A., Pot, J., Peleman, J., Kuiper, M. andZabeau, M. (1995) Nucleic Acids Res. 23, 4407^4414.

[10] Bancroft, I., Love, K., Bent, E., Sherson, S., Lister, C., Cobbett,C., Goodman, H.M. and Dean, C. (1997) in: Weeds World: TheInternational Electronic Arabidopsis Newsletter (Anderson, M.,ed.), Vol. 4 (ii), pp. 1^9, Nottingham Arabidopsis Stock Centre,Nottingham.

[11] Deèhais, P., Coppieters, J. and Van Montagu, M. (1996) Arch.Physiol. Biochem. 104, B38.

[12] Staden, R. (1996) Mol. Biotechnol. 5, 233^241.[13] Hebsgaard, S.M., Korning, P.G., Tolstrup, N., Engelbrecht, J.,

Rouzeè, P. and Brunak, S. (1996) Nucleic Acids Res. 24, 3439^3452.

[14] Tolstrup, N., Rouzeè, P. and Brunak, S. (1997) Nucleic Acids Res.25, 3159^3163.

[15] Borodovsky, M. and MacIninch, J.D. (1993) Comput. Chem. 17,123^133.

[16] Burge, C. and Karlin, S. (1997) J. Mol. Biol. 268, 78^94.[17] Pedersen, A.G. and Nielsen, H. (1997) in: Proceedings of the

Fifth International Conference on Intelligent Systems for Molec-ular Biology, (Gaasterland, T., Karp, P., Karplus, K., Ouzounis,C., Sander, C. and Valencia, A., Eds.), pp. 226^233, AmericanAssociation for Arti¢cial Intelligence Press, Menlo Park, CA.

[18] Galperin, M.Y. and Koonin, E.V. (1998) In Silico Biol. 1, 0007(http://www.bioinfo.de/isb/1998/01/0007/).

[19] Nakamura, Y., Sato, S., Kaneko, T., Kotani, H., Asamizu, E.,Miyajima, N. and Tabata, S. (1997) DNA Res. 4, 401^414.

[20] Kaneko, T., Kotani, H., Nakamura, Y., Sato, S., Asamizu, E.,Miyajima, N. and Tabata, S. (1998) DNA Res. 5, 131^145.

[21] Wu, H.-J., Gaubier-Comella, P., Delseny, M., Grellet, F., VanMontagu, M. and Rouzeè, P. (1996) Nature Genet. 14, 383^384.

[22] Sharp, P.A. and Burge, C.B. (1997) Cell 91, 875^879.[23] Sonnhammer, E.L.L. and Durbin, R. (1995) Gene 167, GC1^10.[24] Kowalski, S.P., Lan, T.-H., Feldmann, K.A. and Paterson, A.H.

(1994) Genetics 138, 499^510.

FEBS 21569 2-3-99

N. Terryn et al./FEBS Letters 445 (1999) 237^245 245