Amino Acid Sequence of the Penicillin-Binding Protein/DD-Peptidase of Streptomyces K15. Predicted Secondary Structures of the Low Mr Penicillin-Binding Proteins of Class A

Amino

acid

sequence

of

the penicillin-binding

protein/DD-peptidase

of Streptomyces

K15

Predicted

secondary

structures

of

the

low-M,

peniciliin-binding

proteins of class A Pilar PALOMEQUE-MESSIA,* Serge ENGLEBERT,* Melina LEYH-BOUILLE,*

MartineNGUYEN-DISTECHE,* Colette DUEZ,* Simone HOUBA,* Otto DIDEBERG,*

Jozef VAN BEEUMENt and Jean-Marie GHUYSEN*t

*Centred'Ingenierie desProteines, Universitede Liege, Institut de Chimie, B6, B-4000 SartTilman (Liege 1), Belgium,

andtLaboratorium voorMicrobiologie enMicrobiele Genetica, Rijksuniversiteit-Gent, K.L. Ledeganckstraat 35, B-9000 Ghent, Belgium

Thelow-Mrpenicillin-binding protein (PBP)/DD-transpeptidase ofStreptomycesK15is synthesized in the formofa

291-aminoacid-residue precursor possessing acleavable 29-amino acid-residue signalpeptide. Sequence-similarity searches

and hydrophobic-cluster analysis show that the Streptomyces K15 enzyme, the Escherichia coli

PBPs/DD-carboxy-peptidases 5 and 6, the Bacillus subtilis PBP/DD-carboxypeptidase 5and the spollA product(aputativePBPinvolvedin

the sporulation ofB. subtilis) are structurally relatedand form a distinct class A of

low-M,

PBPs/DD-peptidases. The distribution of thehydrophobic clusters along theaminoacidsequencesalso shows that theStreptomyces K15 PBP, and

by extension the other PBPs of class A, have similarityin the polypeptide folding, with the

,8-lactamases

of class A, with asreferencethe Streptomycesalbus G and Staphylococcusaureus,J-lactamasesof known three-dimensional structure.This

comparison allowsonetopredictmostof thesecondarystructures inthe PBPs and the amino acid motifs that definethe enzyme activesites.

INTRODUCTION

Thelow-Mrpenicillin-binding proteins (PBPs)/DD-peptidases catalyseruptureofthe C-terminal peptidebondofthebacterial wall peptidoglycan precursors and other R-D-alanyl-D-alanine-terminated analogues by sequential transfer ofthe R-D-alanyl moietytoanessentialserine residue(withreleaseofthe leaving

group D-alanine and formation of a serine ester-linked acyl-enzyme)and thentoanexogenousacceptor. Penicillin inactivates the DD-peptidase activity by immobilizing the essential serine residueinto a stable acyl-enzyme (Ghuysen, 1991).

Thelow-MrPBP/DD-peptidaseofStreptomycesK15 has been purifiedtoproteinhomogeneityinthepresenceof

cetyltrimethyl-ammonium bromide(Nguyen-Distecheetal.,1982).It ispeculiar

in at least two respects. (1) At variance with other low-Mr

PBPs/DD-peptidases that are anchored into the plasma

mem-brane bya non-cleaved C-terminal peptide segment (Spratt &

Cromie, 1988),theStreptomyces K15enzymecanbesolubilized

directly from the mycelium with the help of a high NaCl

concentration (M. Nguyen-Disteche & M. Leyh-Bouille,

un-published work). (2)Theefficacywithwhich thelow-Mr PBPs/DD-peptidases perform hydrolysis compared with transpeptidation

of thecarbonyldonordependsontheefficacywithwhichwater,

the leaving groupD-alanine andanexogenous peptide NH2-X

can attack the acyl (R-D-alanyl-)enzyme. Although all known

PBPs/DD-peptidases act mainly as DD-carboxypeptidases, the

relative acceptoractivityfor theacyl-(R-D-alanyl-)(Streptomyces

K15 enzyme)is water<< D-alanine <H2N-X. Hydrolysisofthe

carbonyl donor,inthe absenceofH2N-X, isnegligible because

attack of the acyl-enzyme by the leaving group D-alanine

regenerates theoriginal substrate. In thepresence ofa suitable

compound NH2-X, the carbonyl donor is quantitatively

con-verted intoatranspeptidated product, i.e. the Streptomyces K15

enzymefunctionsas astrictDD-transpeptidase(Nguyen-Disteche

etal., 1986).

Understanding thesepeculiar properties needsaprecise

knowl-edge of the structure of the Streptomyces K15

PBP/DD-pep-tidase.Genecloning andsequencing has yielded the amino acid

sequence of the protein. Pair-wise comparison with the other

low-M,

PBPs of known primary structure has led to the conclusion that(1) the Streptomyces K15 PBP, the Escherichia coli PBPs 5 and 6 (Broome-Smith etal., 1988), the Bacillus subtilisPBP5(Toddetal., 1986)and thespollA product,which isthought tobe aPBPspecificallyinvolved in sporulation (Wu

&Piggot, 1991), belongto a sameclass Aof

low-M,

PBPs/DD-peptidases, and(2) these PBPsadopt apolypeptide scaffolding

comparablewith that ofthe class A fl-lactamases of

Staphylo-coccus aureus (Herzberg & Moult, 1987) and Streptomyces

albusG(Didebergetal., 1987; Lamotte-Brasseur etal., 1991).

MATERIALS AND METHODS

Bacterialstrains, plasmidsand media

Streptomyces K15wasfrom thislaboratory.PlasmidpBR322

(Bolivar etal., 1977) was used for cloning experiments.

Escherichia coli HB1O1 (Boyer&Roulland-Dussoix, 1969)served

as host of the recombinant plasmids. Streptomyces K15 was

grown at 28°C in tryptone/soya broth (Oxoid) with vigorous

orbitalshaking.E.coliwasgrownat37°CinLuriaBertani(LB)

orM9CA medium(Maniatisetal.,1982)withvigorous shaking.

TransformantswereselectedonLB/agar plates containing25,ug

ofampicillin/ml or 12.5,goftetracycline/ml.

Abbreviations used: PBP,penicillin-binding

protein;

HCA,

hydrophobic-cluster analysis.

: Towhomcorrespondence should be addressed.

Thenucleotilesequence data

reported

willappearin theEMBL,GenBank and DDBK Nucleotide Sequence Databases under the accession number X59965.

N-Terminal VTKPTI AAVG

YAMNNG1T

GT TLYTKAADTRRSTGSTT|KI M(T) AKVXLAO^SNLN amino

acidsequence

A T

5'G-GGA-T_CC-TAC-GCC-ATG-AAC-AAC-GGC3' 3TTC-TAG-TAC-TGG-CGG-TTC-CAC-TTA-AGG5'

BamHl EcoRI

Senseprimer Antisenseprimer

5'ATG-AAC-AAC-GG

C-AC G-GG C-AC C-ACG-CT C-T AC-AC C-AAG-GCC-GCG-GAC-ACC-AGG-PCR

Sall fragment

-CGG-TCG-ACC-GGC-TCC-ACC-ACC-AAG-ATC-ATG-ACC3

3 |GC-TGG-CCG-AGG-T

GG-TGG-TTC-TAG-T ACJ5

Non-degenerated26-merprobe

Fig.

1. DNA

probe

construction

Onlypartof thesequenceof thePCRfragment is shown(seethetext).

Enzymes, antibiotics, oligonucleotides and recombinant DNA techniques

Restriction endonucleases and T4 DNA ligase were from Boehringer (Mannheim, Germany) and Bethesda Research Laboratories (Gaithesburg, MD, U.S.A.); lysozyme and tetra-cyclinewerefrom Sigma ChemicalCo.(St.Louis,MO, U.S.A.). Sequenasewasfrom UnitedBiochemical Corp.(Cleveland, OH, U.S.A.), bacterial alkaline phosphatase was from Amersham International (Amersham, Bucks., U.K.) and ampicillin was fromBeecham (Brussels, Belgium). Oligonucleotides (see below) were synthesized by Eurogentec (Liege, Belgium). The

Strepto-myces chromosomal DNA was preparedas described by

Hop-wood et al. (1985). Preparation ofplasmid DNA, transformation ofE. coli,digestion ofDNAwithrestrictionenzymes, treatment withbacterial alkaline phosphatase, ligation,agarose-gel electro-phoresis of plasmids and digestedDNAsandelution ofseparated DNA fragments were performed essentially as described in Maniatisetal. (1982).

Radioactiveoligonucleotide probe andscreening of gene libraries

On the basis of the known amino acid sequence ofthe

N-terminal region ofthe mature Streptomyces K15 PBP

(Leyh-Bouilleetal., 1989) and the known Streptomyces codon usage, two nucleotide primers were synthesized (Fig. 1). The sense

oligonucleotide IhadaBamHI siteatthe5'-endand encodedthe

sequence Y12AMNNG17 of the protein and the antisense

oligonucleotide 2 had an EcoRI site at the 5'-end and was

complementary ofthenucleotidesequenceencodingthe sequence

K38IMTAKV44. Polymerase chain reaction (PCR) was

per-formed on 1001ul samples containing the Streptomyces K15 chromosomal DNA(1.5

jug),

theprimers (1 ,Meach),the dNTPs

(200,UM

each),

the

Taq

DNA

polymerase (2.5 units;

Perkin-Elmer-Cetus, Norwalk, CT, U.S.A.) and

gelatin

(10,ug).

The buffer was 10

mM-Tris/HCl

buffer, pH 8.3,

containing

50 mm-KCIand 1.5mM-MgCI2. Sampleswerecovered with mineral oil andsubmittedto 30

amplification

cycles:

1mindenaturationat

94°C, 1.5 min annealing at 55

°C

and 2min

polymerization

at

72°C(Saikietal., 1988;Leeetal.,

1988).

The110

bp-DNA

PCR product was partially

sequenced

and the sequence translated exactly into the

expected

peptide

segment

(Fig. 1).

From this nucleotide sequence, a

non-degenerated

26-mer

probe (Fig. 1)

was

synthesized,

labelled with

[y-32P]ATP

(Maxam

&

Gilbert,

1980) and used to screen gene libraries by hybridization

[according to the procedure described in Focus (1984); BRL, Bethesda, MD, U.S.A.]. Southern blotting was performed as described in Maniatisetal. (1982).

Nucleotide sequence

DNA segments cloned into bacteriophages M13mpl8 and

M13mpl9 were sequenced by the dideoxynucleotide chain-termination method(Sanger et al., 1977, 1980). The sequencing reactions were conducted with the universal M13 sequencing primer or the two synthetic oligonucleotides shown in Fig. 2.

Zones of base compression due to high G+C content were

resolved by using dITP instead of dGTP (Sequenase kit). For eachpossible reading frame the codonusage wasanalysed with Staden's program (Staden & McLachlan, 1982; Staden, 1982), with the geneencoding the Streptomyces R61 PBP/DD-peptidase (Duezetal., 1987) asreference.

Amino

ac2'

analysis

Amino acid analysis was performed on a Multichrom B

analyser(Beckman, PaloAlto, CA, U.S.A.). Before hydrolysis thesampleswereoxidizedeither inthegasphaseorintheliquid phaseand thenhydrolysed in vacuowith 6M-HCIat 106°Cfor 24h.

Similaritysearches

TheGoad& Kanehisa (1982)procedure expressesthe extent

ofsimilarity betweenpairsof amino acid sequences bya score, the more negative the score, the better the similarity. The significance oftheextentofsimilarityfoundbetweentwoaligned amino acid sequences was estimated

by

the

SEQDP

program. Scores with atleast five deviations above that expected from a run of20randomized pairs ofproteins having thesame amino

acid

compositions

as the

proteins

under consideration indicate

significant similarity. Pair-wise aminoacidsequences

alignment

wasalsooptimized usingthe localhomology

algorithm

of Smith & Waterman (1981)

(Bestfit

program inthe GCG

package;

gap weight, 5.0; length

weight,

0.3).

Prediction of secondarystructuresbythehydrophobic-cluster analysis

Hydrophobic-cluster analysis (HCA)

was

performed

and the

HCA

sequence-similarity

scores wereestimated as described in

Gaboriaud et al.(1987). HCA is a powerful method for analysing proteins that areweakly related in the primary structure. It rests upon a representation of the protein sequences on an

a-helical

two-dimensional pattern (in which the hydrophobic residues tend to form clusters that usually correspond to the secondary structure elements) and compares thedistributionof the clusters alongthesequences. Clusters of similar shapes, sizes and relative positions express similarity in the polypeptide folding of the proteins. HCA offers the following advantages: (1) the two-dimensional plots allow distant information to become visible morereadily than with methods based only on single amino acid property/identity; (2) deletions or insertions are easily introduced between the secondary structures (Henrissat et al., 1990). Proteins with very close polypeptide folding have HCA score > 75 %. Scores of 55-65 % indicate similarity inthe secondary-structure topology (Gaboriaud et al., 1987; Henrissat etal., 1990). RESULTS

Genecloning and sequencing

The genomicDNAof Streptomyces K15 was partially cleaved withBamHI,withBglIIand with Bcll, and the DNA fragments

were cloned in pBR322 previously cut with BamHI and

de-(a)

km) 500 1000 13

,. 1 * 2

w 3 Il 4 *,

SphI N-Term. Sail

PCR A B C D 357bp SmaliC-Term Sphl E F G H

Fig.2. Restriction map of pDML223 (a) and strategy of nucleotide sequencing (b)

The bacteriophage vectors Ml3mpl8, M13mpl9, Ml3tgl3O and Ml3tgl31 were used to clone the various subfragments. The sub sequencesA, B, C, F,G andHwereobtainedby usingthe universal primer and the subsequences D and E by using the synthetic nucleotidescorrespondingtonucleotide residues 742 to723 of the sequence and nucleotide residues 710 to 729 (see Fig. 3). The positionscorrespondingtothe N-terminal and C-terminal residues of theDD-peptidaseprecursorareindicated.

phosphorylated. Of the 2990 ampicillin-resistant and tetracycline-sensitive E. coli transformants obtained, two clones from the BamHIlibrary and one clone from theBclI library gave a strong positive reaction with the radioactive probe after washing the filters at 75 °C

(Tm

-10°C). Two clones, one from each library, were analysed. PlasmidspDML220fromtheBamHIlibrary and

pDML221 from the Bcll library had acquired respectively a

10.7 kb and 9.8 kb insert. They were digested with restriction endonucleases, andthe DNA fragments(separated by gel electro-phoresis and transferred to nitrocellulose) were submitted to hybridization with the radioactive probe. Among the smallest hybridizing fragments, the 1.3kb SphI DNA fragments origin-ating from both pDML220 and pDML221 were subcloned into pBR322, yieldingpDML223and pDML224 respectively. As derivedfrom the restriction maps (Fig.2a), these two plasmids containedthe sameinsert, but in opposite direction.

Establishment of the nucleotide sequence of 90% of each

strand of the 1357bpSphI DNAfragment(by using the strategy shown in Fig. 2b) revealed an 876-nucleotide-residue open reading frame (Fig. 3). This open reading frame starts with a GTGcodon,presents thebiasedpattern of codon usage typical toStreptomycesgenes(thepercentageof

G.

C being 49, 45 and 98 at the first, second and third positions of the triplets respectively), terminates with an amber stop TAG codon and is followed by a putative terminator sequence. The 5'-ATGAAGGG-3' segment at position -9 to -2 upstream of the translation start codon (GTG) complements five bases out of eight of the 3'-end of the 16 S rRNA of Streptomyces lividans (Bibb & Cohen, 1982) and resembles a ribosome-binding site (Hopwoodetal., 1986).

Primary structure

The876-nucleotide-residueopenreading frametranslated into a 291-amino acid-residue protein precursor (Fig. 3). The 29-aminoacid-residueN-terminalregionhasthe features ofasignal peptide; it contains three arginine residues at positions -28, -26and -25, along hydrophobicstretch from Ala-25 to Pro-4 and the leader

peptidase-cleavage

site Ala--3-Thr-Ala--1. The aminoacidsequence of theproteinfrom

Val-1

toAsn-52 is exactly that ofthe N-terminal region ofthepurified

membrane-bound Streptomyces K15 enzyme (as determined by Edman

degradation; Leyh-Bouille etal., 1989) and the amino acid compositionofthepolypeptide chainfrom Val- 1toLeu-262 is in excellent agreement with thatofthemature protein (Table 1). Similaritysearches

Both the Goad & Kanehisa (1982)

procedure (Table

2) and Bestfit program (resultsnot shown) revealed

high similarity,

in the primary structure, between the Streptomyces K15 enzyme, the E. coli PBPs 5 and 6

(Broome-Smith

et

al., 1988)

and theB. subtilis PBP5

(Todd

etal., 1986) and

spollA product (Wu

& Piggot, 1991). No or

marginal

similarity

was observedwith the Streptomyces R61

PBP/DD-peptidase (Duez

et

al.,

1987),

the serine

fl-lactamases

ofclass

A,

Cand D

(Dehottay

et

al.,

1987;

Ambler, 1980; Jaurin &

Griindstr6m,

1981; Dale et

al., 1985)

and the

penicillin-binding

domains of the bifunctional

high-Mr

PBPs(Broome-Smith etal., 1985;Asohet

al., 1986;

Nakamura etal., 1983).

Hydrophobic clusteranalysisandamino acid

alignment

The upper part of Fig. 4 compares the distribution of the

hydrophobic

clusters

along

the amino acid sequences

(see

the

Materials

and methods

section)

of the

Streptomyces

K15

PBP,

theE.coliPBP5

(and by

extension

PBP6;

not

shown)

andthe B.

subtilisPBP5

(and by

extension,

the

spoIlA product;

not

shown).

Alignment of theconservedclusters

requires

veryfew deletions.

Ih) k

GCA

TGCGCACAGCATAGATACAACGCGTACGCATAGCAGTTCCCGTCCACCATCCGGACICCACGGAGGCCGCCGGACCCCA 82 CGCGCCCCTCCTCCCTCACCTTTCAAAGGGACCCCGCCCACGAACTGCACAACCATTCGTGCTACTTACGTGTCAGAGG

AGAAGCGGCATCCCCATGTGGCCGCACTCTTGGGGCATTTAACGTATTCGGGGTATTCGACTTGATAACTGGCATGAAG 240

Val Arq Leu Ar Arq Ala Ala Ala Thr Val Ile Thr Thr Gl Ala Leu Leu Ala Ala

GGT GTG CGC CTC CGT AGA GCT GCC GCC ACC GTC ATC ACC ACC GGC GCG CTG CTC GCG GCC 300

G1 Thr Leu G1 Ala Thr Pro Ala Thr Ala

lThr

Lys Pro Thr Ile Ala Ala Val G y

GGX

ACC CTC GGC GCG ACT CCC GCG ACC GCC ACC MG CCC ACC ATC GCC GCT GTG GGC

₃₆₀

20 30

Gliqr Ala Met Asn Asn Gl Thr Gl Thr ThrLeuTrThr Lus AloA aAsA hr Ar

GGCM

Tc WC TG AAC MC GGC ACO GGC ACC ACG CTC TIC ACC AAG GCC GCG

GAt

ACC AGG 420

*40 s50

Arq Ser Thr Gl Se Thr Thr L s lie Met Thr Ala

Lws

Vol Vol Leu Ala Gln Ser Asn

CG TCG ACC GO TCC ACC ACC AAG ATC ATG ACC GCC AAG GIG GTC CTG GCC CAG TCG MC

480

Leu Asn Leu Asp Ala Lqs Val Thr

CTG MC CTC

GAT

GCC MG GTC ACG

L4s Pro Ser Gln Ala H's Leu Ile

GG CCC TCG CAG GCG CAC CTG ATC

60

Ile Gln Lus Ala

Tr

Ser

ATC CAG AAG GCC TCAGC

80

Val Gi A Ls Val Thr

GTC GGC

GAE

AXG

GTC ACC

70

Asp Tjr Val Val Ala Asn

GAC TXC GTG GTC GCC AAC 540

90

Val Arq Gln Leu Leu T

GTC CGC CAG CTC CTG TXC 600

100 110

GlU Leu Met Leu Pro Ser GI Cs A¶ Ala Ala

Tr

Ala Leu Ala Am L s

Tr

Gl Ser

GGG CTG ATG CTG CCG TCC

GM

TI GAC GCC GCG T GCG CTG GCCGA GTCG ICG 660

120 130

Gi Ser Gln Ala Ala Ala Ar Val Ls Ser Phe Ile Gl Ls Met Asn Thr Ala Ala Thr

GGC AGC CAG GCT GCC GCG CG GTG AMG TCG TTC ATC GGC

AXG

AIG AAC ACC GCC GCG ACC 720

140

Asn Leu Gi1 Lou His Asn Thr Hts Phe Asp Ser Pho Asp

MC CTC GGT CTG CAC MC ACG CAC TTC GAt TCG TIC GAt

150

Glu Ile Glu A Gly Ala Asn

GG ATC GGC MC GGC GCC MC 780

160 170

Tr Ser Thr Pro Arg His Leu Thr Lgs Ile Ala Ser Ser Ala Met L AsneSr Thr Phe

TAC TCG ACG CCG CGG CAC CTG ACG AAG ATC GCC AGC AGC GCG ATG AXG MC TCG ACG TTC 840

Ar Thr Val Val

Lws

Thr Lys

CGC ACG GTC GTC AMGACC

AXG

180 190

Ala Tr Thr Ala Ls Thr Val Thr Ls Thr Gl Ser Ile

GCGTXC ACG GCG A G ACG GTC ACC AG ACC

GGM

AGC ATC 900

200 210

ArlThr Met As Thr Tr LsAsn Thr Asn Glu Leu Leu Ser Ser

Tr

Ser Gl Ala Ile

CGCACC AG GA ACGTIGG

AXG

AAC ACC AAC GGQ CTG CTC AGC AGC TIC AGC GGY GCG ATC 960

Gig Val Ls Thr Gl Ser Gl Pro

GGC GTG MG ACC GGC TCC GGC CCC

220

Glu Ala

Lws

r

GAG GCC AAG TAC

230

Cus Leu Val Phe Ala Ala Th Arq

TGC CTC GTC TTC GCC GCC ACC CGG1020

240 250

Gli GG Ls Thr Val Ile Glu Tr Val Leu Ala Scr Thr Scr Ile Pro Ala

Ar2

Glu Ser

GGC GGO

AXG

ACG GTC ATC GGC ACC GTC CTC GCC TCC ACG TCC ATC CCG GCC CGC GAG TCG1080

Asp Ala Thr Lus Ile Met

GA GOCC ACC

AXG

ATC ATG

260 A

Asn Tr Gi1 Phe AIa

AAC T C GGC TTC GCC | TAG CCCACGCACACGCAGAAGGGGCCGACC 1146

GCTTGTCGCGGTCG

CGCCCTTCTGCCGTTCGTGCGGCCCCTGGCGCCGTCTCAGGCCGGGTTGCCGGTCGAGACGAAGAG

CTGGATGTCCGGCATGTTCTTCGGS-TCCTCATCGGIITCCGCCGATTGGOCAGTTGGGCCATGACCGTGCCCTCGCCGTA1304

CGTGTICTCGTCCTGGTGCTTGATCGTCCACGACCAGCCCGCCGCCCGCATGC

Fig.3.Nucleotide sequence of the geneencodingtheStreptomycesK15PBP/DD-peptidaseprecursor and deducedamino acid sequence of theprotein

TheN-terminal valine residue and C-terminal leucine residue ofthematureenzymeareboxed.Theputative ribosome-bindingsite is underlined.

The invertedrepeatsof theputative transcriptionterminationsignalareshownbyhorizontalarrows.Ser* represents the active-site serine residue.

Theconserved clusterscoverapprox.30%of theproteins.Of the total 14 alignedclusters, ten for the

pair

Streptomyces Ki5-E. coli (pair I)andeightfor the

pair

Streptomyces K15-B. subtilis (pairII)have HCAscoreshigherthan 65%.Within theconserved area, the overall HCA scores are

800%

for

pair

I and71% for pair II. Many charged and additional hydrophobic residues are also conserved around the clusters showing considerable sequence similarity. The lower part ofFig.4 shows the amino

acid sequence alignment as derived from the

hydrophobic-cluster distribution. The cost of this comparison, expressed in percentage of residues of the

original

sequences that are not

effectively aligned,is 8%for bothpairs.Thealignmentgenerates

26% strict identities and 34% similarities for pair I, and 21 % strict identities and 29 % similarities for pair II. Note that, at variancewithStreptomyces K15 PBP, theE.coli andB. subtilis PBPs have a long C-terminal extension. These extensions are excluded from the present analysis.

As shown above, the Goad & Kanehisa (1982) and Bestfit procedures didnothighlight similarity,intheprimarystructure, between the three PBPs and the class A

,J-lactamases.

HCA,

however, revealedsimilarity, inthe

polypeptide folding,

between the PBPs and the ,-lactamases of

Staph.

aureus

(Herzberg

& Moult, 1987)andStrep.albus G

(Dideberg

et

al., 1987;

Lamotte-Brasseur etal., 1991).Thuspair-wise comparison

(upper

partof

Table 1. Amino acid composition oftheStreptomyces K15DD-peptidase

The amino acid composition is given as number of residues per

proteinmolecule asderived from chemical analysis (A) and from the

nucleotidesequence of the encoding gene (B).

Amino acid composition (residue/molecule) Amino acid A B Lys His Arg Cys Asp Asn f Met Thr Ser Glu Gln Pro Gly Ala Val Ile Leu Tyr Phe Trp Total 20 22 4 4 10 9 2 2 28

110

14 8 7 34 34 21 22 8 2 7 6 28 25 34 33 17 18 12 12 19 18 10 12 6 6 1 268 267

Fig. 5) of the Streptomyces K15 PBP with the Strep. albus G /-lactamase(pair III)and with theStaph.aureus/1-lactamase(pair IV)shows that the conserved areasrepresent 25 %(pairIII) and 30% (pair IV)of the proteins. Within the conserved areas, the overall HCA scores are 70% (pair III)and 73% (pair IV). The HCA scores for the individual clusters vary from 50% to89% (pair III) and from 22% to 100% (pair IV). The cost of the amino acid alignment (lower part of Fig. 5) does not exceed 12%. The sequences effectively aligned contain 15.6% strict identities and 23 % similarities for pair III, and 11.5% strict identities and 19% similarities forpair IV. The

significance

of

the similarity is expressed by a low but significant standard deviation of 9units (pair III) and 6 units (pair IV).

DISCUSSION

The Strep. albus G and Staph. aureus

,-lactamases

of class A areofknownthree-dimensionalstructure. These proteins consist ofan all-a domain (helices a2-a9) and an

a//,

domain (helices

al,alIO,al1andstrands

/11-/15).

Their active sites are defined by four conserved amino acid motifs. Using the ABL numbering

scheme, the S70VFK motif in the Streptomyces enzyme or

S70TSKinthe staphylococcal enzyme is central to the enzyme cavity with the essential S70 at theN-terminus of helixa2.The S13ODN motif (inboth enzymes), on a loop connecting helices

a4 and a5, occurs on one side of the cavity. The E166PELN

motif in the Streptomyces enzyme or E166IELN motif in the

staphylococcal enzyme is on a loop at the entrance of the cavity. Finally, the K234TG motif in the Streptomyces enzyme or

K234SG motif in the staphylococcal enzyme occurs on the

innermost

/13

strandofthe

fl-sheet

on the other side ofthe cavity. Onthe basisof the principlethat similarity in the hydrophobic-clusterdistribution expresses similarity in the polypeptide folding of the proteins, the present study leads to the following con-clusions.

(1)TheStreptomyces K15 PBP, the E. coli PBP5 (and 6) and the B. subtilis PBP5(andspoIIAproduct)arestructurally related (Fig. 4).Itisproposed that they belongto a same class Aof low-MrPBPs/DD-peptidases.

(2) The secondary structures in the Streptomyces K15 PBP (and the other PBPs of class A) have a spatial disposition comparable with that foundin the

,1-lactamases

of classA(Fig. 5).Identificationinthe PBPs of strands /

11,

/82,

/13, /14

and

/5

and of helices a2, a4,a5,a6,ac8 andacl 1 is not a matterofcontroversy. Identificationin the PBPsofclustersequivalent to helicesa3, 7, a9andalO in the

/1-lactamases

is hypothetical. Note, however, the presence of one glycine residue in helix a7 and of two asparagine residues on the N-terminal side of helix alO in all PBPs.

(3)Threeofthe four motifs that define the active site of the /-lactamases (see above) have homologues in the Streptomyces

K15, E. coli and B. subtilis PBPs. Thus the STTK, SLTK or

SMTK tetrad of the PBPs(astheyarelisted)arehomologues of

the S70XXK motifofthe

,-lactamases,

theSGC, SGN or SAN

Table2.Search forsequencesimilaritybetweenthe aminoacidsequencesof severallow-Mr PBPs andII-lactamasesbyusingtheGoad & Kanehisa

(1982)

method

Comparisonscores (belowthediagonal line)andsignificancein SDU(abovethediagonal line) areshown.

2. E.coli PBP5 3. E.coli PBP6 4. B.subtilis SpoIIA 5. B.subtilis PBP5 6. 7. 8.

Streptomyces Staph. aureusStrep. albus G

R61 PBP 8-lactamase /1-lactamase 1. StreptomycesK15PBP 2. E. coliPBP5* 3. E. coliPBP6* 4. B.subtilisSpoIIA* 5. B.subtilis PBP5* 6. StreptomycesR61 PBP 7. Staph.aureus,b-lactamase 8. Strep.albus G /-lactamase

~~~22

16 28 14 -0.5 6 3

-230 ~~~~106

29 12 0.9 0.4 -0.7 -225 -1270 24 21 2 0.6 0.8 -290 -460 -500 25 -0.3 0.3 -0.6 -180 -170 -300 -390 3 2 4 -40 -50 -70 -30 -80 3.6 0.01 -110 -50 -40 -70 -70 -60 61 -50 -25 -45 -20 -70 -35 -330

* TheC-terminal extensionsof the E. coliPBP5(fromP267 toG374),PBP6(fromT262toS373),the B. subtilisPBP5(fromE283 toW412)and

SpollA (fromP284 to K389)have been eliminated for thecomparison. 1.

Streptomyces

_ - s -rnL= , S: c ct * ~ ><QUIH>CHOEq X>24 X I0 * HH>4 ¢ k/) Hx'C > CCH 0iU3@1. EA < 4 1-4 1G>4 EnVD U)a CHN4 OP XH HI: .p E>Z* @ h: ~~~xmE. 0 W4 C S-4HH. H Hf_h74H *_@'C0 _CHd >o uH >0 *COX mHHC' CCH~~*>~~>4 0 'DcOH-4 0 Z 0 >H4In * Ce H Oa'Clp -,* 'oiH HHH HIi'000'*eo2 9>@6

-,i kor OHH*EqE fiM *

*F

oP

a;

W;z:;

:< >< *HHHVo . U

*H:

_LO-o

s _E* P;_U)0 >-A H,ix_K

_.Qz

_{7 f}

H:> Hf\vU:> uz C,)i~-l is Z0¢clfC .000.~~~~~~0 HC-)ES I SFju~_n V E 0a_ :H *>4CC. (D. :z' c)zcf H Z n xHt.- t-

x

ff <.

LD H OXn LC a rT ,x>4 040 0 cnE a4 L,i*D o *>412% 0 0 L 'C'CqXZ t E >cHO X 01- 00 0X : OH4H H L 2~~~*4 0'7 Co i 3 *a,>C>EPH4E,XX F, f X_ x* f uHZ 004~ HH > 4> Oe0Ct'CH 3g3 -EH4 Zr'' Z C 2 2_pHO_v)U 00XH >CX _0p[ZOH ¢ > <>HH L^iO H3H 00a , o CO> O H V H'CH >4C 0 U V HO Z V * <H >1* H .00 * E-;( ( * E S>q H x N(:) * 2* + -:HF1 > :> (D 1- E- <: E-4tn C PnI< lHCx g;Z,XU) Ui0

< ] X :> o~~~~~>~~~~~az :>>U)x4 >

t P4 H * - a~~~ 4 HX > apF I ~ ~~>>SZ~~~ a X [

=

H Icl

>

<~~0

01

E

1U

1- Enll x * @ NN4 1-o1 k S4 Eq > * o ; >4 Ef C: M b M @ X g D~~~~~C' cCflcCfl d-cl.-I 4z ~~~- <OL

an.

-R

bwd

O 'C ucfl CA 0) 0) c Q 0. Ct4 ;e 0 o c-i O -' *; g 00 C. o4 '4-04 I. O 04 _ Q C , 0O4 >' m o 0 0 .- 0 H ~ Ci ,-, 0. t 0. U '10 112, 1 -4 r-I c c K Cc 1991

m

u *@ 0c AA

o

u) -X4 >~~~~~~0 0~~~~ CY~~~~~ t) *.~~ a* o x00> * x 'U C H >U) *;|> *h-> X*x{ X S B B H<>404 t Kc r4;<z~ ~ z0 el,¢:<MM00s)K : .5 0>4 e. . 0>>H XL SOH 0V Z5 ZO t i: ) O 4< VH EA0 >40<oH (HO s0 E-4p U) *EH C H X @nH V)a 04

0~

01 LO E-4H <E z co >% z E4Er-1 OH<g Z S:H Xa > pn 4 a4z > z- X > .Z . 0 *-4 H

04H

n4 w a4a4ZkD 4 > >X4

*ZZZZ *>>XEqsazLI'I' I)(

a;>G b XL0 s GU% XL Z Wz1 a;>c>>4 0 < 14 :> a)M t V XH H U.X = X< Nzrv x>< wg u0>-el) i) 0 > E- p > XQ>4z u) < L1FQ > rz5 >4 H E-14 HIE-4; 4 O£<: im In X>, >a ° >H>- nV~~~L <>u < 4 ur_cn~~IV ' Z Z *o w| :gn *Esx> 13 u 0- Q *na ..~~ : C PHP >COZH~~~~0 C X;>>4 0E404o4 0 0>>4X H£Xw O P~~~~ ~<Eq~~;P Xxe X4> XLO LI) E- X :L4H-< O<y Z CY -.> soon. *HO4Z* E--I E-4 >4 I H HVU -CE CX 0 H0>:X el > m ~~~~~4

pe

04

S~~~~~~~

f

Ee

t>

Cox 7. Z UZ E- z_, SZ .v 6000

15N

xS

tv

L00H4O> * H uXg H> <>0 4 h* 000 OH>4 pH ~ c;z 5 H 0EX-*U' < - ° UF-~~x zco 00H ~~wa000 0ZH SH0x co U)~~ ~ ~ V (A 0.-I~~~~~~~~~ ._.

00

rJ~~~~~~~~~~i

triad arehomologuesof theS13ODNmotif,and the KTG motif is the homologue of the K234T(S)G motif. In addition, the

DSFD, DAD or ENKD groupings of the PBPs might be

homologues of the E166XELN motif of the

fl-lactamases.

The conserved motifsplay essential roles in thecatalytic mechanism of the

fl-lactamases.

Theirhomologues may fulfil similar func-tions inthe PBPs.

(4) At variance with the ,-lactamases and the Streptomyces K1 5 PBPwhosepolypeptidechainterminates about60residues downstream from theKT(S)G motif, the E. coli and B. subtilis PBPs have long additional C-terminal extensions themselves each terminated byanamphiphilic a-helical membrane anchor (Ferreira et al., 1988). With reference to the

fl-lactamases,

the Streptomyces K15 PBP appears to have much shorter al and

a 11 helices andtherebyitsf-sheetispartiallyuncovered on one face,thusexposinga zoneofhydrophobicity potential.This zone might be the site of interaction between the protein and the plasma membrane.

(5)TheStreptomyces K15 PBP(Leyh-Bouille et al., 1987) and E. coli PBP5 (Curtis &Strominger, 1978) are susceptible to thiol-specific reagents. In the Streptomyces K15 PBP, one cysteine residue(the SGCmotif)occurs within the active site on the loop connecting helices a4 and a5 and another occurs ten residues downstream of thelysine residue of the KTGmotif,i.e. onstrand

fl4.

Which ofthese two cysteine residues is the site ofreaction withp-chloromercuribenzoatecan bedefinitely established only by site-directed mutagenesis. The E. coli PBP5 possesses one singlecysteine residue at the third position immediately down-streamfrom the SGNmotif, i.e. in the immediate vicinity of the active site. Finally, one single cysteine residue, ten residues downstream from the KTG motif(i.e.onstrand

,/4)

isalsofound inthe B.subtilisPBP5 and in the SpoIIAprotein. Whether or not these two proteins are susceptible to thio-specific reagents is unknown.

The work inLiegewas supported in partby the Fonds de la Recherche

Scientifique Medicale (Contractn°3.4537.88),anAction Concertee with

the Belgian Government (Convention 86/91-90), the Fonds de Recherche de laFacult6deMedecineULg and a ConventionTripartite

between the Region Wallonne, SmithKline Beecham, U.K., and the

University ofLiege. Thistext presents researchresults of the Belgian

programme on Interuniversity Poles of Attraction initiated by the

Belgian State,Prime Minister'sOffice, SciencePolicy Programming.The

scientificresponsibility is assumed by its authors.P. P.-M. was afellow of

theEuropeanCommunities(TrainingContract n°010197).Someof the

work described in thispaperformspartofadissertation presentedby

P.P.-M. as partial fulfilment fora Ph.D. degree at the University of

Granada (Spain). The work in Ghentwas supported by the National

IncentiveProgramof Fundamental ResearchinLife Sciences initiatedby

theBelgian SciencePolicy Programming Department (Bio22).Wethank

D.Vaira (University of Liege) for assistance in PCR and D.Allaer

(Eurogentec, Liege) for helpful biochemical discussion.

REFERENCES

Ambler, R. P. (1980) Philos. Trans. R. Soc.LondonB 289, 321-331 Asoh, S., Matsuzawa, H., Ishino, F., Strominger, J. L., Matsuhashi, M.

& Ohta, T. (1986) Eur. J. Biochem. 160, 231-238

Bibb, M. J. & Cohen, S. N. (1982) Mol. Gen. Genet. 187, 265-277

Bolivar, F.,Rodriguez, R. L., Greene, P. J., Betlach, M. C., Heyneker,

H.L., Boyer, H.W., Crosa, H. J. & Falkow, S. (1977) Gene 2, 95-113

Boyer,H. W.&Roulland-Dussoix, D. (1969) J. Mol. Biol. 41, 459-472

Broome-Smith, J. K., Edelman,A., Youssif, S. &Spratt, B. G. (1985)

Eur.J.Biochem. 147, 437-446

Broome-Smith,J.K.,,Ioannidis, I.,Edelman,A.&Spratt,B.G.(1988)

Nucleic Acids Res. 16, 1617

Curtis, S. J. &Strominger,J. L.(1978)J.Biol.Chem. 253. 2584-2588

Dale, J. W., Godwin, D.,Mossakouska, D.,Stephenson, P. &Wall,S.

(1985) FEBS Lett. 191, 39-42

Dehottay, P., Dusart, J., De Meester, F.,Joris, B., Van Beeumen, J.,

Erpicum, T., Frere,J.-M.&Ghuysen,J.-M.(1987)Eur.J. Biochem.

166,345-350

Dideberg, O., Charlier, P., Wery, J. P.,Dehottay,P., Dusart,J.,Erpicum,

T., Frere,J.-M.&Ghuysen,J.-M. (1987) Biochem. J. 245, 911-913

Duez, C.,Piron-Fraipont, C., Joris, B., Dusart, J.,Urdea, M. S., Martial,

J. A., Frere, J.-M. & Ghuysen, J.-M. (1987) Eur. J. Biochem. 162,

509-518

Ferreira, L. C. S., Schwarz, U., Keck, W.,Charlier,P.,Dideberg,0. &

Ghuysen,J.-M.(1988)Eur.J.Biochem. 171, 11-16

Gaboriaud,C.,Bissery,V.,Bencheritt,T. &Mornon, J.P.(1987)FEBS

Lett. 224, 149-155

Ghuysen, J. M. (1991) Annu. Rev. Microbiol. 45, 37-67

Goad, W.B.&Kanehisa, M.T.(1982) Nucleic Acids Res. 10, 247-263

Henrissat, B.,Saloheimo, M., Lavaitte, S. & Knowles, J.K.C.(1990)

Proteins Struct.Funct. Genet.8, 251-257

Herzberg, 0. &Moult,J.(1987) Science236,694-701

Hopwood,D.A., Bibb, M.J.,Chater, K.F.,Kieser, T., Bruton, C. J.,

Kieser,H.M., Lydiate,D.J.,Smith, C. P.,Ward,J. M.&Schrempf,

H. (1985) Genetic Manipulation of Streptomyces: A Laboratory

Manual, The John InnesFoundation, Norwich

Hopwood,D.A.,Bibb, M. J., Chater, K. F., Janssen, G. R.,Malpartida,

F.&Smith, C.P.(1986) inRegulation of GeneExpression: 25 Years

On (Brooth, E. R. & Higgins,C.F., eds.), pp. 251-276, Cambridge

University Press, Cambridge

Jaurin, B. &Gruindstr6m, T. (1981) Proc. Natl. Acad. Sci. U.S.A. 78,

4897-4901

Lamotte-Brasseur, J., Dive, G., Dideberg,O., Charlier, P., Frere, J.-M &

Ghuysen,J.-M.(1991) Biochem. J. 279, 213-221

Lee, C. C., Wu, X., Gibbs, R. A., Cooks, R. G., Muzny, D. N. &

Caskey, C.T. (1988) Science 239, 1288-1291

Leyh-Bouille, M.,Nguyen-Dist6che, M., Bellefroid-Bourguignon, C. &

Ghuysen,J.-M.(1987) Biochem.J.241, 893-897

Leyh-Bouille, M., Van Beeumen, J., Renier-Pirlot, S.,Joris,B.,

Nguyen-Disteche, M.&Ghuysen,J.-M.(1989) Biochem. J. 260,601-604

Maniatis, T., Fritsch,E.F. & Sambrook, J. (1982)Molecular Cloning:

ALaboratory Manual, Cold Spring Harbor Laboratory, Cold Spring

Harbor

Maxam, A.M.&Gilbert,W. (1980)Methods Enzymol. 65, 499-560

Nakamura, M., Maruyama, I.N., Soma, M., Kato, J., Suzuki, H. &

Hirota, Y.(1983) Mol. Gen. Genet. 191, 1-9

Nguyen-Disteche, M., Leyh-Bouille, M. & Ghuysen, J.-M. (1982)

Biochem.J.207, 109-115

Nguyen-Disteche, M., Leyh-Bouille, M., Pirlot, S., Frere, J.-M. & Ghuysen,J.-M.(1986) Biochem.J.235, 167-176

Saiki,R.K.,Gelfand,D.H.,Stoffel, S., Scharf, S. J., Higuchi, R.,Horn, G.T.,Mullis,K. B. &Erlich,H.A.(1988) Science 239,487-491

Sanger, F.,Nicklen, S. & Coulson, A.R.(1977) Proc.Natl. Acad. Sci.

U.S.A.74, 5436-5467

Sanger, F., Coulson, A. R., Barrell, B. G., Smith, A. J. H. & Roc, B. A.

(1980) J. Mol. Biol. 143,161-178

Smith,J.F.&Waterman,M.S.(1981)Adv.Appl. Math 2, 482-489 Spratt,B.G.&Cromie, K.D.(1988) Rev.Infect. Dis. 10, 699-711 Staden, R.(1982) Nucleic Acids Res. 10, 2951-2961

Staden, R.&McLachlan,A. D.(1982)Nucleic Acids Res. 10, 141-156

Todd,J.A., Roberts,A.N., Johnstone, K., Piggot, P. J., Winter, G. &

Ellar,D.(1986)J. Bacteriol. 167,257-264

Wu,J. J.&Piggot,P.J.(1991)inGenetics andBiotechnologyofBacilli,

vol. 3(Zukowski, M.M., Ganesan, A.T. &Hoch, J.A., eds.), pp.

321-327,AcademicPress,NewYork

Received4March 1991/24 April 1991; accepted 8 May 1991