Gene, 102 (1991) 255-259 0

1991 Elsevier

GENE

Science

Publishers

255

B.V. 0378-I 119/91~$03.50

04046

Primary structure of human, chicken, and Xenopus laevis pll, substrate, annexin II (Recombinant

DNA;

cDNA

cloning;

nucleotide

sequence;

Ca2 + binding;

a cellular ligand of the Src-kinase

EF-hand;

S-100 proteins)

Eckhard Kube, Klaus Weber and Volker Gerke Department of Biochemistry, Max-Planck-Institute for Biophysicar Chemistry. D-3400, GBttingen [F.R. G.) Tel. (49)551201486 Received by H.G. Zachau: 7 November Accepted: 19 December 1990

1990

SUMMARY

The pl 1 protein is a member of the S-100 family of Ca2’ -binding proteins and serves within the cell as a ligand of the tyrosine kinase substrate, annexin II. To obtain more structural information on this molecule, we have isolated and characterized pl f cDNA clones from several different species. A comparison of the deduced amino acid (aa) sequences reveals that mammalian and avian pl 1 are highly similar (at least 90% identical at the aa level), whereas p 11 from Xenopus la&s shows a considerable degree of sequence variation (the aa sequence identity drops to approx. 60% when compared to mammalian or chicken ~11). Interestingly, the C-terminal 18 aa, which are unique to pl 1 within the S-100 family, show a relatively high conservation among species. This high evolutionary conservation is in line with a structurally and/or functionally important role of this C terminus, e.g., in annexin II binding.

INTRODUCTION

The p 11 protein belongs to a recently identified family of small dimeric proteins, which share sequence homologies with S-100& and S-lOOp(for review see Kligman and Hilt: 1988). characteristically, the members of the S-100 protein family show two helix-loop-helix structures (EF-hands) thought to be involved in Ca2 +-binding (for review see Persechini et al., 1989). The pl 1 protein is unique within the S-100 family for several reasons. First, pl 1 has suffered

Correspondenceto: Dr. V. Gerke, Planck-Institute

for Biophysical

Tel. (49)551201477: Abbreviations:

of Biochemistry,

D-3400 Gottingen

Max(F.R.G.)

Fax (49)551201578.

aa, amino acid(s); bp, base pair(s); cDNA,

mentary to RNA; kb, kilobase( nucleotide; ORF, open reading II-p1

Department

Chemistry,

DNA comple-

nt, nucleotide(s); oiigo, oligodeoxyriboframe; ~11, small subunit of the annexin

1 complex; pll, gene (DNA) encoding pII; SDS, sodium dodecyl

sulfate; src, protein encoded by the src gene of Rous sarcoma 0.15 M NaCl10.015 M Na, I citrate pH 7.6.

virus;

SSC,

crucial aa deletions and substitutions in both EF-hand loops which render the two Ca2+-binding sites inactive (Gerke and Weber, 1985; Glenney, 1986). Second, the primary structure of pll which was determined for the porcine, bovine and murine molecule so far shows a unique C-terminal extension following the last helix of the second EF-hand (Gerke and Weber, 1985; Saris et al., 1987). Third, pl 1 has been shown to interact in vitro and in vivo with another cellular protein, the tyrosine kinase substrate annexin II (also known as ~36, lipocortin II, and calpactin I heavy chain; for review see Gerke et al., 1990). Annexin II is a member of the annexin family of Ca2’- and phospholipid-binding proteins (for review see Klee, 1988; Crompton et al., 1988). At least some of its biochemical properties, e.g., the Ca2 + -dependent binding to and aggregation of lipid vesicles, are modulated by pl 1 binding (Powell and Glenney, 1987; Drust and Creutz, 1988). The complex formation between p 11 and annexin II is mediated through an amphiphatic a-helix formed by the N-terminal 12 aa of the annexin II molecule (Johnsson

256 et al., 1988; Becker et al., 1990). The binding site for this annexin II helix on the p 11 molecule, however, has not been mapped so far but chemical modification experiments suggest that CYSTS(located in the unique C-terminal extension) or its immediate environment is involved in this interaction (Johnsson and Weber, 1990). To obtain more information

about evolutionarily conserved residues be involved in annexin II binding we characterized pll cDNA clones from comparison of the different sequences malian and avian pl 1 are remarkably Xenopus

luevis

pll

shows

in p 11 which might have isolated and several species. A reveals that mamsimilar whereas

a high

degree

M

s

of sequence

A AGAATACACTCACAAGCCACTCCGCTGCTCGCCTCTCCGCCCCGCGTCCAGCTCGCCCAG CTCGCCCAGCGTCCGCCGCGCCTCGCCAAGGCTTCAACGGACCACACCAAA~CCATCT P

CAAATGGAACACGCCATGGAAACCATGATGTTTACATTTCACAAATTCGCTGGGGATAAA QMEHAMETMMFTFHKFAGDK GGCTACTTAACAAAGGAGGACCTGAGAGTACTCATGGAAAAGGAGTTCCCTGGATTTTTG GYLTKEDLRVLMEKEFPGFL GAAAATCAAAAAGACCCTCTGGCTGTGGACAAAATAATGAAGGACCTGGACCAGTGTAGA ENQKDPLAVDKIMKDLDOCR GATGGCAAAGTGGGCTTCCAGAGCTTCTTTTCCCTAATTGCGGGCCTCACCATTGCATGC DGKVGFQSFFSLIAGLTIAC AATGACTATTTTGTAGTACACATGAAGGGAAAGAAGGGAAAGAAGTAGGCAGAAATGAGCAGT NDYFVVHMKQKGKKk TCGCTCCTCCCTGATAAGAGTTGTCCAAAGGGTCGCTTAAGGAATCTGCCCCACAGCTTC CCCCATAGAAGGATTTCATGAGCAGATCAGGACACTTAGCAAATGTAAAAATAAAATCTA ACTCTCATTTGACAAGCAGAGAAAGAAAAGTTAAATACCAGATAAGCTTTTGATTTTTGT ATTGTTTGCATCCCCTTGCCCTCAATAAATAAAGTTCTTTTTTAGTTCCA~~

B 1 CCCCTGCGCCGCGCGACCCCACAGCCGCCACC~C~GT~CC~GATGGAGCACGCCATGG 1 M

E

H

A

M

Y

F

V

V

61 AGA~GC:GA~GT~CA~CT:CC~CA~AT~CG~GG~TG~CA~GAACTACCTGAGCAAGGAGG NYLSKE 10 E 121 ACCTGCGTGCGCTGATGGAGAAGGAGTTCCCCGGATTCCTGGAGAACCAGCGCGACCCTA 30DLRALMEKEFPGF LENQRDP 181 TGGCGCTGGATAAGATCATGAAGGACCTGGACCAGTGCCGGGATGGCAAAGTGGGCTTCC 50MALDKIMKDLDQCRDGKVGF z7” ~GA~CTTCTTCT~ACTGGTGG~TG~AC~GA~CA~CG~CT~CA~TG~CTACTTCGTGGTGC L v F F 301 ACATGAAGCAGAAGG~GCGGAAGTGAGAGAAGGCCCCGCAGCCCCAATAAAGTGTTTTAT R K * 90H M K Q K 361 ATGA,,

Fig. 1. The nt sequences by asterisks.

of human (A) and chicken (B) ~12 and the deduced

The canonical

polyadenylation

signals in the 3’-nontranslated

from HT29 cell mRNA and screened with two synthetic p11.3: 5’-AAICCAGGGAACTCCTTITCCAT;

~11.4: 5’-CTTCT”T(G,C)CCCTTCTGCCTTCATGT. Hybridization was carried out at 48°C in 6 x SSCjl Tris

x

HCI pH 7.5/l x lo5 cpm per ml of 5’ end-labeled

aa sequences. portions

The start codons

are underlined.

are boxed and the stop codons

(A) A human

cDNA library in lgtl0

oligos:

Denhardt’s

solution/l00

ng per ml of yeast tRNA (Boehringer)/O.OS

oligo. After hybridization,

filters were washed

% Na

pyrophosphate/20

in 6 x SSC/O.OS% Na. pyrophosphate

and then exposed to Kodak X-OMAT film with an intensifying screen at -80°C. Of 5 x lo4 plaques screened one clone hybridized The cDNA from this clone was purified, subcloned into Ml3mpl8 (New England Biolabs), and sequenced by the chain-termination et al., 1977) using a T7 sequencing

kit (Pharmacia).

of the human pll cDNA (covering

nt 134-589).

are indicated was prepared

(B) A chicken Hybridization

embryo libroblast

was carried

cDNA library in lgtll

out in 48% formamide/4.8

was screened

x SSCjl

with both probes. method (Sanger

with an NcoI-Hind111

x Denhardt’s/lO%

mM at 60°C

dextran

fragment

sulfate/O.1 y0

SDS/20 mM Tris HCl pH 7.5/l x lo5 cpm per ml of “P-1abelled DNA fragment. Final washing of the nitrocellulose filters was at 40°C in 2 x SSC/O.l % SDS. Of 4 x lo4 plaques screened, two clones reacted with the probe. cDNA from both clones was subcloned and sequenced as described in part A of this legend.

257 library in i,gtll (kindly provided by Dr. R. Hynes, MIT, Cambridge, MA) at low stringency flinal wash in 2 x SSCjO.l% SDS at 40°C). Sequence analysis of two purified cDNA clones revealed an identical ORF (nt 33-326) encoding 97 aa (Fig. 1B). The deduced aa sequence of this ORF is highly similar to the mammalian pl 1 sequence. A comparison of the sequence shown in Fig. 1B with the N-terminal 49 aa of chicken pl 1 which were determined by direct gas-phase sequencing of the purified protein (Johnsson, 1989) unambiguously identities the cloned cDNA as chicken p 11. In both species, human and chicken, the mature protein starts with an unprocessed Pro residue (Johnsson, 1989) indicating that the initiator Met is re-

variation. Only three regions of the molecule, including the C-terminal extension, are highly conserved among all species.

EXPERIMENTAL

AND DISCUSSION

(a) Isolation and characterization of cDNA clones for human and chicken pll A lgtl0 cDNA library prepared from HT29 (a human adenocarcinoma cell line) mRNA (kindly provided by Dr. D. Louvard, Pasteur Institute, Paris, France) was screened with two oligos whose sequences were derived from bovine and murine pll cDNA sequences (Saris et al., 1987). The cDNA of a phage clone hybridizing with both probes was

moved after translation. (b) cDNA cloning and expression of Xenopus pll The human cDNA was also employed to screen a Igt 10 cDNA library made from Xenopus kidney cell RNA (kindly provided by Dr. I. Mattaj, EMBL, Heidelberg, F.R.G.). Of 5 x IO4 plaques screened, three positive clones were identified under low-stringency conditions (final wash in 2 x SSCjO. 17; SDS at 3.5“C). Sequence analysis revealed that the three cDNAs were identicai with respect to their nt sequence but differed slightly in their 5’ ends. The clone depicted in Fig. 2A (591 nt) most likely represents the fulllength cDNA since its length matches with the size of the Xenopus pll mRNA (0.6 kb as determined by Northern-

characterized by sequence analysis. Fig. 1A shows the nt sequence of this human pll cDNA and the deduced aa sequence. The ORF (nt 112-405) encodes a polypeptide of 97 aa, which is highly similar to porcine, bovine, and murine pl 1 (Gerke and Weber, 1985; Glenney and Tack, 1985; Hexham et al., 1986; Saris et al., 1987; Masiakowski and Shooter, 1988). In addition, the deduced aa sequence is identical in the respective parts to partial human pll sequences obtained by direct protein sequencing (Johnsson, 1989). The human pll cDNA was in turn used as a hybridization probe to screen a chicken-embryo tibroblast cDNA

I3

A

*

CCGCAGGACACAAGAACCAGAACCAGTCAAG~G~GG~CC~CT~TG~GC~AG~AC~TT~ S CATGGAGA~AA~GC~GT~GA~TT~CC~~A~GT~TG~AG~AG~GA~GA~CT~CATGA~CGG M E

M

GGACGACCTGCAGAAGCTTCTGGACAGCGAGT~CT~CGAGTTCCTGAAGAACC~AA~CGA DDLQKLLDSE E F L

K

N

R

D

CCCCATGACTGTGGACAAAATCATGAAGGACCTGGATGACTGCCGCAAAGGGCAAGTGAA PMTVDKIMKDLDDCRKGOVN CTTCCGCA~CTACT~CTCTCTCATAGCCGGCCTCCTCATCGCCTGCAACGAGTATTACGT Y F R SLIAGLLIACNEYYV

kb

CAAACACATGAAAAAGA~GTGAAAGAGAGGAGAGCGCCTGAGGCCTCGATGC~G~~AAAC * K H M K K

-0.6

TCGCCAGAGCACCGCCCTCACCGCAGCCAATAiTCAATATTGAAATTCG~TCAGCAG~~ACTCCCAC TCTCCCCTGAACAAATGCACAATAAACTCTAAA~TTGTGGTTTTTTTG~TATTTGCCTTT CAGTTTCCTCCTTATTTCCTCTATTCCAGATTTACCCTCCATAGGGAATAAATGGACCAG

Fig. 2. Sequence

of cDNA

and expression

tb

o

TGTCATCATCATTAAAGGGAATATAACCACAAACA,,

ofXenopus

luevis pll. (A) A &IO

cDNA library made from Xenopus kidney-cell

RNA was screened

with the

NcoI-Hind111 fragment of human pll cDNA. Hybridization was carried out as described in Fig. IB legend with the exception that 30% formamide was used instead of 482,. The final washing was at 35°C in 2 x SSCjO.1 Y0 SDS. The cDNA from positive clones was subcloned and sequenced as described before (Fig. 1). The start and stop codons is underlined. electrophoresed

(Panel through

B) Northern-blot a

which determine analysis

of Xenopus

1y. agarose gel and transferred

out at 42°C in the presence

of 50% formamide

the ORF are boxed and marked RNA

isolated

to GeneScreenPlus

using the Xerro~~s pll

from oocytes (DuPont)

cDNA

with an asterisk,

respectively.

(0) or tailbud-stage

embryos

following the manufacturer’s

(A) as a probe.

The polyadenylation (tb). RNA

protocol.

signal

was glyoxylated,

Hybridization

was carried

258 Human. Bovine & Pomne

Rat

bSOMEHAMETMMFTFHKFAGDKGYLTKEDLRVLMEKEFPG ~~~~~L___~___~_~_~~~ ~~~ __~~~L~~~R~~~~~Dt.

MOUSe

~~~~~~~______y____N

Chlcken

US

~~~A

ARK -R ~~~

xenopusVA-MEL--S--K-LL~~~~~~~E~N~MNRD~-OK-LDS-~SE Fig. 3. Sequence

comparison

of mature

~~~

~~

R AKRON

M

~~

~~~~

L

~MT

given at the top are indicated

and Shooter by dashes.

~K~O

blot analysis, Fig. 2B). All cDNA clones contain the identical ORF of 291 nt encoding a polypeptide of 96 aa (Fig. 2A). Both cDNA and deduced aa sequence show a rather low degree of identity (approx. 60%) when compared to the mammalian and avian pll sequences. This finding explains the weak hybridization signal of the Xenopus clones probed with human pll cDNA which is lost upon highstringency washing. The expression of Xenopus pll in two different stages of development was studied by Northern-blot analysis. Fig. 2B shows that a 0.6-kb transcript is recognized by the Xenopus pl I cDNA. Thus, the Xenopus pll mRNA corresponds in its size to pll transcripts from other species (Saris et al., 1987; Masiakowski and Shooter, 1988; E.K., unpublished observations). Interestingly, the Northern blot reveals very little expression of the pl I mRNA in Xenopus oocytes (only a faint band is visible on overexposed films, not shown) whereas the pf 1 transcript is readily detected in total RNA from the Xenopus tailbud stage. (c) Comparison of the pll sequences from different species Fig. 3 shows a comparison of the different pl 1 sequences which were either determined by direct protein sequencing or deduced from the corresponding cDNA sequences. While mammalian and avian pl 1 are highly homologous (the different sequences are at least 90n/rl identical at the aa level) the Xenopus sequence is remarkably divergent. Although the majority of the differences can be explained by conservative aa replacements it seems that species variations cluster in the regions of the two EF-hand loops and the last helix of the first EF-hand. A close inspection of the different primary structures reveals that all p 11 molecules sequenced so far have suffered crucial deletions or substitutions in the EF-hand loops. In particular, some aa, whose side-chain oxygens are believed to coordinate the Ca’+ ions in classical EF-hand proteins like S-100 or calmodulin, are absent in the different pl 1 molecules (for review on the EF-hand see Kretsingcr, 1987). Thus, it can be expected that neither pl I species is able to bind Ca2 + , a prediction already proven experimentally for the porcine and bovine protein (Gerke and Webcr, 1985; Glenney, 1986). Interestingly, the C-terminal extension (aa 78-96) which is unique to pll within the S-100 protein family, shows a relatively high degree of scqucncc conservation.

N

are from Gerke

(1988; rat pl I), and this work (human,

Lines above the sequences

L

V

~~/

L

V

~~

~~ IN

v D

pl 1 molecules from different species. The aa sequences

et al. (1987; bovine and mouse pl I), Masiakowski with the sequence

-0

R

L

EYK

KR

and Weber (1985; porcine

chicken

mark the positions

R

YC

pl I), Saris

and Xenopus ~11). Residues

of the mutated

EF-hand

identical

loops,

Such a conservation can be expected for a structurally and/or functionally important domain, e.g., the annexin II binding site. A direct involvement of the unique C terminus of pl 1 in annexin II binding is in line with the finding that alkylation of Cys x2 in pll abolishes pll-annexin II complex formation (Johnsson and Weber, 1990).

ACKNOWLEDGEMENTS

We thank Drs. D. Louvard, R. Hynes and I. Mattaj for cDNA libraries and Drs. B. Fouquet and W. Franke (DKFZ, Heidelberg, F.R.G.) for Xenopus RNA samples. Thanks also to Dr. N. Johnsson and members of the annexin lab for helpful discussions, and Dr. P. Newman for critical reading of the manuscript. This work was supported by the Deutsche Forschungsgemeinschaft (OS 70/l-3) and by the Bundesministerium fur Forschung und Technologie. The cDNA sequences described in this report are listed in the GenBank database under the accession Nos. M38591 (human pll), M38592 (chicken pll), and M38593 (Xenopus luevis ~11).

REFERENCES Becker, T., Weber, K. and Johnsson, short amphiphatic of annexin

N.: Protein-protein

helices; a mutational

II for pll.

EMBO

analysis

recognition

via

of the binding

site

J. 9 (1990) 4207-4213.

Crompton, M.R., Moss, S.E. and Crumpton, M.J.: Diversity cortinicalpactin family. Cell 55 (1988) l-3. Drust,

D.S. and Creutz,

calpactin

C.E.: Aggregation

at micromolar

Gerke, V. and Weber, K.: The regulatory complex

of viral tyrosine-specific

quence to the 29 17-2920.

S-100

of chromaflin

levels of calcium.

protein

Nature

in the lipogranules

chain in the p36-kDa protein

of glial

kinases cells.

by

33 1 (1988) 88-9 1. substrate

is related

EMBO

in se-

J. 4 (1985)

Gerke, V., Johnsson, N. and Weber, K.: Intestinal Ca’ ’ /lipid-binding proteins. In: Smith, V.L., and Dedman, J.R. (Eds.), StimulusResponse Proteins. Glenney

Coupling:

The Role of Intracellular

CRC Press, Boca Raton,

FL, 1990,

Jr., J.R.: Phospholipid-dependent

tyrosine

kinase

substrate

(calpactin)

Ca”

Calcium

Mediator

pp. 31 l-338.

-binding

and its 33-kDa

by the 36-kDa core. J. Biol.

Chem. 261 (1986) 7247-7252. Glenney Jr., J.R. and Tack, B.F.: Aminoterminal sequences of p36 and associated ~10: identification of the site of tyrosine phosphorylation and homology 7884-7888.

with S-100. Proc. Natl.

Acad.

Sci. USA 82 (1985)

259 Hexham,

J.M., Totty, N.F., Waterfield,

mology

between

associated Commun. Johnsson,

the subunits

with p36 of pig lymphocytes. 134

N.: Biochemische

Untersuchungen

strat zelltransformierender Ttibingen,

Biochem.

M.J.: Hopolypeptide

Biophys.

F.R.G.

an Protein

Proteinkinasen Proteine.

Res.

I, einem Sub-

und Mitglied einer neuen Ph.D. Thesis, University

of

1989.

N. and Weber,

K.: Alkylation

of cysteine

the complex formation

with the tyrosine-protein

(annexin

1, lipocortin

2, calpactin

82 of pl 1 abolishes kinase substrate

2). J. Biol. Chem.

p36

265 (1990)

14464-14468. Johnsson,

N., Marriott,

substrate a short

of src tyrosine amino-terminal

Biochemistry

D. and Hilt, D.C.: The S-100 protein

Kretsinger,

K.: ~36, the major

cytoplasmic

protein kinase, binds to its pll subunit via amphiphatic helix. EMBO J. 7 (1988) phospholipid-

(and

27 (1988) 6645-6653.

membrane-)

family. Trends

R.H.: Calcium

binding

coordination

and calmodulin

versus convergent evolution. Cold Spring Harbor 52 (1987) 499-510. Masiakowski, P. and Shooter, E.M.: Nerve growth genes for two proteins

Biochem.

fold: divergent

Symp. Quant. factor

related to a family of calcium-binding

induces

Biol. the

proteins

in PC12 cells. Proc. Nat]. Acad. Sci. USA 85 (1988) 1277-1281. Persechini, A., Moncrief,N.D. and Kretsinger, R.H.: The EF-hand family of calcium-modulated proteins. Trends Neurosci. 12 (1989) 462-467. Sanger, F., Nicklen, S. and Coulson, A.R.: DNA sequencing with chainterminating

G. and Weber,

2435-2442. Klee, C.B.: Ca’+ -dependent proteins.

Kligman,

Sci. 13 (1988) 437-443.

( 1986) 248-254.

Familie Ca’+ /Lipid-bindender Johnsson,

M.B. and Crumpton,

of S-100 and a 10 kDa

inhibitors.

5463-5467. Saris, C.J.M., Kristensen, Hunter,

T. and Tack,

Proc.

Natl.

Acad.

T., D’Eustachio, B.F.: cDNA

Sci.

USA

74 (1977)

P., Hicks, L.J., Noonan,

sequence

D.J.,

and tissue distribution

ofthe mRNA for bovine and murine pll, the S-loo-related

light chain

of the protein-tyrosine

I). J. Biol.

kinase

Chem. 262 (1987) 10663-10671.

substrate

p36 (calpactin