A family of homologous substrate-binding proteins with a broad range of substrate specificity and dissimilar biological functions

Biochimie (1995) 77, 744-750

© Soci6t6 fran~aise de biochimie et biologic mol6culaire / Elsevier, Paris

A family of homologous substrate-binding proteins with a broad range of substrate specificity and dissimilar biological functions LF Wu, MA Mandrand-Berthelot Laboratoire de GEnEtique MolEculaire des Microorganismes et des Interactions Cellulaires, CNRS-URAI486, Institut National des Sciences Appliqudes, 20, avenue A Einstein, 6962• Villeurbanne Cedex, France

(Received 7 February 1995; accepted 27 April 1995)

Summary - - The uptake of peptides is accomplished mainly by a family of homologous oligopeptide or dipeptide transporters in bac-

teria. Computer-aided sequence analyses expand members of the oligopeptide-binding protein family to nickel and heme permeases and other proteins, including an enzyme hyaluronate synthase. They are involved in human pathogenicity, bacterial virulence, substrate-sensing, bacterial conjugation and bacterial metabolic reactions distinct from nutrient uptake. These homologous proteins are found in both purple bacteria and Gram-positive bacteria, indicating the presence of a common ancestor before the appearance of the two eubacterial phyla. Nevertheless, the pheromone-binding proteins, involved in bacterial conjugation, and the hyaluronate synthase are present only in the low G-C Gram-positive eubacteria subdivision, which suggests that these proteins diverged from the common ancestor after the appearance of this subdivision. peptide / heme / nickel / binding-protein / evolution

Introduction

Bacterial cold osmotic shock-sensitive transport systems are complex permeases composed of a periplasmic substrate-binding receptor and a membranebound complex containing 2-4 proteins Ill. These penneases are energized by the hydrolysis of ATP, which is mediated by one of the proteins in the membrane complex. The ATP-hydrolysis constituent of the complex bears a high level of sequence homology to several eukaryotic membrane-bound proteins, such as the multidrug resistance protein and the cystic fibrosis protein. This universal class of permease is usually referred to as 'traffic ATPase' or 'ABC' (ATP binding cassette) transporters [2, 31. The periplasmic substrate-binding proteins of bacterial cold osmotic shock-sensitive transport systems are the primary specific receptors that initiate uptake of various small molecules such as sugars, amino acids, peptides and inorganic ions. In some cases, they also serve as the primary receptors for bacterial chemotaxis (for reviews see [4, 5]). In Gram-positive bacteria, the transport function of these extracellular proteins is fulfilled by homologous external membrane-bound lipoproteins.

To date, more than 50 sequences of the periplasmic substrate-binding proteins have been determined and analyzed by standard computer methods [6]. According to their sequence relatedness, the majority of these proteins are grouped into eight families. In addition, a class of bacterial DNA-binding proteins is also related in sequence, structure, and evolutionary origin to some of these substrate-binding proteins. The homology between all these proteins resides in that they bind similar substrates. The question raised in this review is whether other functionally unrelated proteins could be found to be homologous with the periplasmic substrate-binding proteins. The fifth family of the substrate-binding protein superfamily consists of, as originally reported, dipeptide-, oligopeptide-binding proteins and the nickelbinding protein required for nickel uptake in E coil [6, 7]. Recent additions to the databanks allowed us to expand this family to a variety of other proteins that are involved in human pathogenicity, bacterial virule~tce, substrate-sensing, and Gram-positive bacterial conjugation. The most intriguing result is that an enzyme, hyaluronate synthase of Streptococcus sp, is also related in sequence with this group of substratebinding proteins.

745

Bio|ogica| functions p e r f o r m e d by the m e m b e r s of the oHgopeptide-binding protein family By using the BLAST program, 17 protein sequences similar to oligopeptide-binding proteins were collected from GenPro databases (table I). Their relatedness was determined by statistical analysis using the RDF2 program [8]. The comparison scores, expressed as z value in standard deviation (SD) units, reveal that the 17 proteins are all homologous (table I). The comparison scores between the nickel-binding protein of E coli (NIKAEC) and the oligopeptide-binding proteins of Streptococcus pneumoniae (AMIASP) and Bacillus subtilis (OPPABS) and the has gene product of Streptococcus sp (HASNSS) are 3, 7 and 5 SD units, respectively, values which are below the level of reliability for establishing homology. However, the obvious relatedness of the nickel-binding protein (NIKAEC) with the oligopeptide-binding protein of E coli (OPPAEC; comparison score: 28 SD units) and of the latter with the above three proteins (OPPAEC/AMIASP: 25 SD units; OPPAEC/OPPABS: 93 SD units; OPPAEC/HASNSS: 45 SD units) suggests that all these proteins are related. The same argument is applicable to the comparison between OPPAEC with URE2EN and XP55SL.

Despite their sequence relatedness, the proteins belonging to the oligopeptide-binding protein family participate in quite different biological processes. Most members of this family were discovered as oligopeptide- and dipeptide-binding proteins of enteric bacteria (E coli and S typhimurium; OPPAEC, OPPAST and DPPAEC), and the Gram-positive species Lactococcus lactis (OPPALL), S pneumoniae (AMIASP) and B subtilis (OPPABS, APPABS and DCIABS). The principal role of the oligopeptide- and dipeptide-binding proteins is nutrient uptake [9]. These proteins display a rather broad substrate specificity and can transport any peptide of a given length, relatively independently of its amino acid composition. Some of these proteins exhibit other related roles such as recycling of cell wall peptides [10], peptidechemotaxis [11], adaptation to nutrient deficiency (DCIABS, [ 12]), and sporulation and competence presumably by recognizing signal molecules [ 13, 14]. Beside the above-cited functions, the peptide antibiotics-binding protein (SAPAST) encoded by the sapA gene of S typhimurium ensures an early step essential for the bacterial virulence [15]. Another distinct function is demonstrated by the pheromone-binding proteins (PRGZEF and TRACEF) encoded by prgZ and traC genes carried by plasmids pCF10 and pAD1 in the Gram-positive bacterium Enterococcus

Table I. Properties of members of the oligopeptide-binding protein superfamily. Abbreviation Gene

Organism

OPPAEC OPPAST DPPAEC OPPALL AMIASP OPPABS

E coli S typhimurium E coli L lactis S pneumoniae B subtilis

APPABS DCIABS SAPAST PRGZEF TRACEF HBPAHI ACCAAT NIKAEC XP55SL URE2EN HASNSS

Substrate specificity

Recycling of cell wall peptides Recycling of cell wall peptides Tap-dependent chemotaxis Recycling of cell wall pep|ides Recycling of cell wall peptides Sporulation, competence, recycling of cell wall appA B subtilis Oligopeptide Sporulation, competence, recycling of cell wall dciAE B subtilis Dipeptides Adap.tation to nutrient deficiency sapA S typhimurium Peptideantibiotics Bacterial virulence prgZ Efaecalis pCFI0 Pheromones Conjugation traC Efaecalis pAD1 Pheromones Conjugation hbpA H it~uenzae type b Haemin Pathogenicity to human accA A tumefaciens AgrocinopinesA, B Signaling, susceptibility to agrocln 84 nikA E coli Nickel Biosynthesis and activation of hydrogenase Not designed S lividans Unknown Unknown Not designed Enterobacteriaceae Unknown Potential nickel-binding protein has Streptococcus sp Biosynthesis of hyaluronate

oppA oppA dppA oppA amiA oppA/spoOKA

Oligopeptide Oligopeptide Dipeptides Oligopeptide Oligopeptide Oligopeptide

Function (other than transport)

Accession Comparison score number NikA OppA

J05433 X04194 M35045 L 18760 X 17337

33 28 33 17 3

284 162 49 13 25

X56347

7

93

-

31

25

X56678 X74212 L14285 L19532 M84028

12 35 24 18 37

95 26 58 46 37

L 14678

20

25

X73143 Y00142 L12007 Z 12624

280 56 16 5

28 5 8 45

Accession code is for GenBank. Comparison scores were calculated using the RDF2 program [8] with 50 shuffling for the second sequence against either NIKAEC (NikA) or OPPAEC (OppA) and expressed as z values in standard deviation (SD) units. As generally accepted, two sequences are homologous when the comparison score is higher than 10 SD. The probability to obtain this score by chance is < 0.7 x 10-23.

746

faecalis, respectively. They seem to be involved in pheromone sensing and play an important role in bacterial conjugation, which is a major mechanism of horizontal genetic transfer in Gram-positive cocci [16, 171. Other homologous proteins able to bind substrates distinct from oligopeptides or dipeptides are the haemin-binding protein A (HBPAHI) of Haemophilus influenzae, the opine-binding protein (ACCAAT) coded by the gene accA of the Ti plasmid harbored by Agrobacterium tumefaciens and the nickel-binding protein NIKAEC of E coli. H influenzae type b is recognized as the most important cause of meningitis in the United States [18]. This species is readily distinguished from nearly all ott'¢.r facultative Gramnegative bacteria by its absolute growth requirement for heme [19]. The haemin-binding lipoprotein HBPAHI is thus indispensable for its pathogenicity. The opine-binding protein ACCAAT is involved in the transport of opines, such as agrocinopines A and B, which can be utilized as growth substrates by pathogenic Agrobacterium strains [20, 21]. Intracellular opines derepress the repressor AccR-mediated opine catabolism and the tumor-inducing plasmid conjugal transfer, which are essential steps in the interaction between these bacteria and their plant host [22]. The periplasmic nickel-binding protein NIKAEC provides cells with nickel, which is a crucial component of the catalytic center of NiFe-hydrogenases essential for the anaerobic growth of E coli [71. Moreover, our recent genetic evidence suggests that NIKAEC also serves as the primary chemotactic receptor involved in the Tardependent repellent chemotaxis [231. The functions of two other homologous proteins, the major secreted protein of 55 kDa from Streptomyces iividans (XP55SL) 1241 and the polypeptide derived from the ORF2 of a plasmid-encoded urease gene cluster from several members of the family Enterobacteriaceae (URE2EN) [251, remain unknown. A search of the GenPro bank revealed that the URE2EN was similar in protein sequence with the nickel-binding protein (NIKAEC). Similarly, when URE2EN was used as a query, the two highest scored related proteins were XP55 of S lividans and NikA of E coli. The probabilities to obtain by chance such scores by the BLAST program are 8.2 x 10-13 and 8.2 x 10- 7, respectively. Statistical analysis using the RDF2 program confirmed that URE2EN and NIKAEC are homologous (table I). Since urease is also a nickel-containing metailoenzyme I261, and heteroexpression of Kiebsieila aerogenes urease in E coil depends on the NikA activity [601, URE2EN might be the nickel-binding protein that is involved in nickel transport and the activation of ureases. Urease acts as a virulence determinant for uropathogenic bacteria by catalyzing the hydrolysis of urea to ammonia and carbon dioxide, thus producing

alkaline conditions in the urinary tract. Increased urine pH can lead to formation of struvite stones that harbor the infecting organism, enhanced attachment of bacteria to the renal epithelium, direct renal tissue damage, and inactivation of certain antibiotics [27]. Most intriguingly, the last member of this family is an enzyme, the hyaluronate synthase encoded by has gene in Streptococcus sp (HASNSS). Group A Streptococcus (GAS) continues to be a major cause of morbidity and mortality throughout the world, both from infections and from the post-infectious sequelae of acute rheumatic fever and poststreptococcal glomerulonephritis. The characteristic mucoid colony morphology of these strains is due to abundant production of the streptococci hyaluronate capsule which protects the bacteria from immunological attack by the infected host and consequently amplifies infectious virulence [28 ]. Peptides can serve as important sources of nutrients for most species of bacteria. Small peptides also serve many specific biological functions as hormones, toxins or antibiotics. Transport of peptides is apparently the common characteristic of most members of the oligopeptide-binding protein superfamily. This capacity of peptide-uptake exhibits antagonistic effect on the bacteria. First, in contrast to the Opp and Dpp systems, the Sap system naturally transports peptide antibiotics, including protamine [15], in order to execute its virulence. Second, the natural substrates of the Acc system of A tumefaciens are agrocinopines A and B, but the antibiotic agrocin 84 produced by avirulent Agrobacterium radiobacter K84 can also enter the cells via the same system, resulting in susceptibility of nopaline- and agropine-type Agrobacterium sp to this antibiotic I201. Similarly, the proposed function of the ami system of S pneumoniae is recycling cell wall peptides which are utilized as nutrients. However, ami mutation can increase cell resistance to aminopterin, methotrexate and celiptium [29-31]. An intriguing question arising from these observations concerns the 'true' biological substrate(s) of each of these homologous binding-proteins and their pathway of evolution from a common ancestor, which led to their various roles and the adaptation to a variety of environments.

Evolutionary relationship of oligopeptide-binding protein family A phylogenetic tree of the 17 proteins listed in table I was constructed using their protein sequences and is presented in figure 1. Six subclusters are apparent. The oligopeptide-binding proteins of E coli (OPPAEC), S typhimurium (OPPAST), B subtilis (OPPABS) and the dipeptide-binding protein of B subtilis (DCIABS) make up the first subcluster. The

747 two plasmid-encoded pheromone-binding proteins of Efaecalis (TRACEF and PRGZEF) and the hyaluronate synthase of Streptococcus sp (HASNSS) form the second subcluster. The oligopeptide-binding proteins of S pneumoniae (AMIASP) and L lactis (OPPALL) represent the third subcluster. The heine-binding protein of H influenzae (HBPAHI), the dipeptide-binding protein of E coli (DPPAEC) and the peptide antibiotics-binding protein of S typhimuHum (SAPAST) make up the fourth subcluster. The opine-binding protein of A tumefaciens (ACCAAT) and the plasmidencoded potential nickel-binding protein found in the family of Enterobacteriaceae (URE2EN) form the fifth subcluster. The second oligopeptide-binding protein of B subtilis (APPABS), the major secreted protein of S lividans (XP55SL) and the nickel-binding protein of E coli (NIKAEC) represent the final subcluster. APPABS is located far away from the major oligopeptide-binding protein subcluster, as revealed by a branch length of 1921. This result is consistent with its relative low comparison score with the oligopeptide-binding protein of E coli (OPPAEC, table I,

25 SD), and with the observed physiological difference between these two binding-proteins. Unlike OppA, AppA is not able to transport tripeptides [14]. At present, the members of this family are representatives only of purple bacteria, and Gram-positive bacteria, mainly alpha and gamma purple eubacteria and low G-C Gram-positive bacteria subdivisions (fig 2). The protein XP55 of S iividans is the only member found in high G-C Gram-positive bacteria subdivision. Statistical analyses established that the 17 members are homologous (table I), indicating the presence of a common ancestor before the split between the two eubacterial phyla. Nevertheless, the pheromone-binding proteins involved in bacterial conjugation, and the hyaluronate synthase are present only in low G-C Gram-positive eubacterial subdivision (fig 2), which suggests that these proteins diverged from the common ancestor after the appearance of this subdivision. The appearence of 'new' catalytic functions of enzymes during the natural evolution can occur by one of two distinct mechanisms [37]. First, the forma-

OPPAEC (g) OPPAST (g)

TRACEF (i) " ; ~ HASNSS (I)

IABS (1)

11

PRGZEF (I)

442

734

334

470

OPPABS (I) Fig 1. Parsimony tree for

0)

315

559

997 362 ~ 4 8 0

645

OPPALL (1)

351

SAPAST (g)

395

442

513

280

,qga

HBPAHI (g)

311

477

831

DPPAEC (g) 563

NIKAEC (g) 552

ACCAAT (a)

637

XP55SL (h) APPABS (1)

~

URE2 EN (g)

the oligopeptide-binding protein family. The tree was constructed from the protein sequences using the Treealign program 1321. The standard parameters were used (mutation distance matrix derived from Doolittle's similarity measure; the insertion/deletion weight gk = I 1 + 3k). Relative distances are indicated adjacent to the branches. Heavy lines show the different subclusters. Protein names are the same as in table I. Their belongings to various eubacterial phyla are determined according to Woese 133] and Olsen et ai [34] and presented as (!) for low G-C Grampositive bacteria, (h) for high G-C Gram-positive bacteria, (a) for alpha purple bacteria and (g) for gamma purple bacteria.

748

NIKAEC MBPEC

li Dtn

264 342

t[Aivli a

s g

q t

yu k f a

e

365

Fig 2. Sequence alignment between nickel-binding protein (NIKAEC) and maltose-binding protein (MBPEC). Conserved amino acids are presented in capital letters while non-conserved residues are in lower case. Italic letters show residues specically involved in maltose chemotaxis as demonstrated by site-directed mutagenesis [35, 36]. Letters in shadowed boxes represent identical or similar amino acids. Numbers indicate the residue positions. tion of new active sites on the same protein framework may result in enzymes with similar protein fold but completely different functions. The second mechanism involves transformation of the old active site to a new function. Most of enzymes that have diverged by the second way are clearly homologous and assume closely related function. Evolution of the substrate-binding proteins might undergo both procedures. Thus modification of the sl~bstrate-bi~ding site would result in different substrate specificities, whereas acquisition of a chemoreceptor-interaction site would confer the chemotaxis function to some proteins. An intriguing observation is that various bacteria use different transport systems to ensure supplies of the same substrate. For example, high-specific nickel-uptake is achieved via the ABC transport system Nik in E coli [7], while the high-affinity nickel transporter of Alcaligenes eutrophus comprises only one extremely hydrophobic integral membrane protein encoded by the hoxN gene [38, 391. Both systems are required for the synthesis of nickel-containing hydrogenase and heterogenous expression of urease in the corresponding bacteria (see above and [38]). None of the bacteria has been reported to simultaneously contain both high-affinity nickel transport systems.

Similarly, plasmid-encoded urease gene cluster of the family Enterobacteriaceae comprises an O R F encoding a potential nickel-transporter homologous to Nik of E coli (see above), whereas the chromosomal urease gene complex of thermophilic Bacillus sp strain TB-90 contains a ureH gene whose product shows 23% amino acid identity to the HoxN protein, the high-affinity nickel transporter of A eutrophus [40]. Finally, although an ABC transport system was used for haemin uptake in both H influenzae [41] and Yersinia enterocolitica [42], the periplasmic haeminbinding protein of/-/influenzae (HBPAHI) is homologous to oligopeptide-binding proteins (s,~e above) and that of ¥ enterocolitica, ~ E M T , seems to be related to different periplasmic siderophore-binding proteins belonging to the eiighth family of the superfamily [6, 42]. As its homologous proteins, the size of H E M T is about half that of HBPAHI. The key question arising from this obsetwation is what made a given bacterium use one rather than another transport system. The question then becomes whether the 'choice' is strictly dependent on environment and host physiological difference, and finally, whether a bacterium is capable to 'choose" a system to adapt selective conditions during its evolution.

Table !!. Structural properties of periplasmic substrate specilic binding proteins Protein Name

From bacteria

Encoded Size,, by

MBPEC ABPEC GBPEC GBPST RBPEC LIVEC LBPEC SBPST PBPEC LAOST HISJST HISJEC OPPAST DPPAEC NIKAEC

E coil E coil E coil S typhimurium E coli E coli E coli S typhim,rium E coli S typhimurium S typhimurium E coli S typhimurium E coli E coli

malE araF mgIB mglB rbsB livJ livK c pstS argT hisJ hisJ oppA dppA nikA

370 306 309 305 271 344 346 310 319 238 238 238 519 507 502

Substrate Chemotactic PDB C o d e Structure Familyt' Reference Receptor Ligand Resolution (,~) Mal, Maltodextrin Ara Gal Glc Gai Glc Rib Leu, Iso, Val Leu, TFL Sulfate Phosphate Lys, Arg, Orn His His Oligopeptide Dipeptide Nickel

Tar Trg Trg Trg Tap Tar

IMBP I ABP 2GBP 3GB~,~" I DRI 2LIV 2LBP I SBP I ABH

+/+ + + + + +/+

2.3 1,7 !.9 2.4 1.7 2.4 2.4 2.0 1.7 1.8/! .9 C C 1.8 C C

I 2 2 2 2 4 4 6 6 7 7 7 5 5 5

[431 144] [45,46] [471 1481 [49] 150] [51 ] [52] [53] [54] I55] 1561 [571 [581

aSize of mature form in residues, baccording to 161. cGene designation of this binding protein is not reported. Abbreviations: Mal, maltose; Ara, arabinose; Gal, galactose; Glc, glucose; Rib, ribose; Leu, leucine; Iso, isoleucine; Val, valine; TFL, 5,5,5-trifluoroleucine; Lys, lysine: Arg, arginine; Orn, ornithine; His, histidine; C, the proteins have been crystallized.

749 Since 1984, the structures of I I periplasmic substrate-binding proteins belonging to six families have been solved with resolutions ranging from 1.7 to 2.4 ,~ (table II). They all share a strong similarity in the basic tertiary structural features. Unlike other periplasmic binding proteins, the crystal structure of the liganded form of OPPAST consists of three domains, each containing a I3-sheet. Two of the domains, linked by two segments, present surfaces that enclose the ligand and are structurally analogous to those present in other periplasmic substrate-binding proteins which contain only these two domains. The last domain (residues 45 to 168) has no counterpart among the periplasmic substrate-binding proteins from other families and its function is not knowo. Our recent study shows that NIKAEC mediates the Tar-dependent nickel repellent chemotaxis [23]. This firstly provides support to the notion that homologous periplasmic binding proteins can interact with different chernotactic receptors (NIKAEC with Tar and DPPAEC homologue from S typhimurium with Tap), and that the non-homologous nickel-binding protein NIKAEC and maltose-binding protein MBPEC can be recognized by the same chemotactic receptor like Tar. Furthermore, a search of the Brookhaven Protein Data Bank (April 1994 release) revealed that the protein sequence of MBPEC shows the highest comparison score with the sequence of nickel-binding protein NIKAEC. One of the three aligned segments comprising 24 residues (amino acid position 342 to 365 in MBPEC) exhibits 37% sequence identity with NIKAEC (fig 2). This region has been demonstrated, by both deletion mutation and site-specific mutagenesis, to be involved in maltose-chemotaxis [35, 36]. Furthermore, domains involved in specific interaction between the maltose-binding protein and the chemoreceptor Tar have been determined at the molecular level [59]. Recently, Charon et al have successfully crystallized the nickel-binding protein NIKAEC [58]. Determination of its structure and comparison with the structures of other substrate-binding proteins will provide very important information to explain the structure, function and substrate specificity of the periplasmic substrate-binding proteins. The development of molecular biology techniques has led to an accelerated discovery of new sequences. Sequence similarity has been widely used to interpret the function of newly discovered proteins. However, study presented in this mini-review indicates that homologous proteins can perform different functions. The most intriguing example is that various substrate-binding proteins showing different substrate specificities are related in sequence to an enzyme. Therefore, apparent homology between two sequences does not automatically imply that they are functionally related.

Acknowledgments We thank D~J Reizer for critical reading of this manuscript. We acknowledge the National Center for Biotechnology Information (NCBI) for BLAST network service.

References 1 Ames GF-L (1986) Bacterial periplasmic transport systems: structure, mechanism, and evolution. Annu Rev Biochem 55,397--426 2 Ames GF-L, Mimura CS, Shyamala V (1990) Bacterial periplasmic permeases belong to a family of transport proteins operating from Escheric'hia coli to human: traffic ATPases. FEMS Microbiol Rev 75,429-446 3 Higgins CF (1992) ABC transporters: from microorganisms to man. Annu Rev Cell Biol 8, 67-113 4 Furlong CE (1987) Osmotic-shock-sensitive transport systems. In Escherichia coli and Salmonella ~.phimurium: Cellular a~ld Molecular Biology (Neidhardt FC, JL lngraham, KB Low, B Magasanik, M Schaechter, ME Umbarger, ed) Am Soc" Microbiol, Washington, DC, 768-796 5 Bourret RB, Borkovich KA, Simon M1 (1991) Signal transduction pathways involving protein phosphorylation in prokaryotes. Annu Rev Biochem 60, 401--441 6 Tam R, Saier MHJ (1993) Structural, functional, and evolutionary relationships among extracellular solute-binding receptors of bacteria. Microbiol Rev 57, 320-346 7 Navarro C, Wu L-F, Mandrand-Berthelot MA (1993) The nik operon of Escherichia coli encodes a periplasmic binding protein-dependent transport system for nickel homologous to the peptide permease family. Mol Microbiol 9, ! 181-1191 8 Pearson WR, Lipman DJ (1988) Improved tools for biological sequence comparison. Proc Natl Acad Sci USA 85,244 A. 2448 9 Higgins CF, Gibson MM (1986) Peptide transport in bacteria. Methods Encymol 125,365-377 10 Goodell EW. Higgins CF (1987) Uptake of cell wall peptides by Sahnonella ~. phimurium and Escherichia coli. J Bacteriol 169, 3861-3865 I I Manson MD, Blank V, Brade G. Higgins CF (1986) Peptide chemotaxis in E c'oli involved the Tap signal transducer and the dipeptide permease. Nature 32 I, 253-256 12 Mathiopoulos C, Mueller JP, Slack FJ, Murphy CG, Patankar S. Bukusoglu G. Sonenshein AL (1991) A Bacillus subtilis dipeptide transport system expressed early during sporulation. Mol Microbiol 5, 1903-1913 13 Rudner DZ, LeDeaux JR, Ireton K, Grossman AD (1991) The spook krcus of Baciihts subtilis is homologous to the oligopeptide permease locus and is required tbr sporulation and competence. J Bacteriol 173, 1388-1398 14 Koide A, Hoch JA (1994) Identification of a second oligopeptide transport system in Bacilh¢s subtilis and determination of its role in sporulation. Mol Microhioi 13, 417-426 15 Parra-Lopez C. Baer MT, Groisman EA (1993) Molecular genetic analysis of a locus required for resistance to antimicrobial peptides in Sahnonella typhimurrain. EMBO J ! 2, 4053--4062 16 Wirth R (1994) The sex pheromone system of Enterococcus faecalis: more than just a plasmid-collection mechanism? Eur J Biochem 222, 235-246 17 Dunny GM, Leonard BAB, Hedberg PJ (1995) Pheromone-inducible conjugation in Enterococcus faecalis: interbacterial and host-parasite chemical communication. J Bacteriol ! 77, 871-876 i 8 Cochi SL, Broome CV (1986) Vaccine prevention of Haemophilus influen:ae type b disease: past, present, and future. Pediatr Infect Dis 5. i 2-19 19 Evans NM, Smith DD, Wicken AJ (1974) Haemin and nicotinamide adenine dinuccleotide requirements of Haemophilus influen:ae and Haemophilus parainfluenzae. J Med Microbioi 7,359-365 20 Ellis JG, Murphy PJ (1981) Four new opines from crown gall tumors-their detection and properties. Mol Gen Genet ! 81, 36-43 21 Hayman GT, von Bodman SB, Kim H, Jiang P, Farrand SK (1903) Genetic analysis of the agrocinopine catabolic region of Agrobacterium tumefac'iens Ti plasmid pTiC58, which encodes genes required for opine and agrocin 84 transport. J Bacteriol i 75, 5575-5584

750 22 Zambryski P, Temp6 i, Scheli J (1989) Transfer and function of T-DNA genes from Agrobacterium Ti and Ri plasmids in plants, Cell 6, 193-201 23 De Pina K, Navarro C, McWalter L, Boxer DH, Price NC, Kelley SM, Mandrand-Berthelot MA, Wu L-F (1995) Purification and characterization of the periplasmic nickel-binding protein NikA of Escherichia coli K12. Eur J Biothem 227, 857-865 24 Burner WV, Henner J, Eckhardt T t 1987,', The nucleotide sequence of the gene coding for XP55, a major secreted protein from Streptomyces lividans. Nucleic Acids Res 15, 3926 25 D'Orazio SEF, Collins MC (1993) Characterization of a plasmid-eneoded urease gene cluster Ibund in members of the family Enterohacteriaceae. J Bacteriol 175, 1860-1864 26 Hausinger RP (1987) Nickel utilization by microorganisms. Microbiol Rev 51, 22-42 27 Mobley HLT, Hausinger RP (1989) Microbial ureases: significance, regulation. and molecular characterization. Microbiol Rev 53, 85-108 28 Wessels MR, Moses AE, Goldberg JB, DiCesare TJ (1991) Hyaluronic acid capsule is a virulence factor for mucoid group A streptococci. Proc Nati Acad Sci USA 88. 8317-8321 29 Sicard AM, Ephrussi-Taylor H (19651 Genetic recombination in DNAinduced transformation of pneumococcus, il. Mapping the amiA region, Genetics 52, ! 207-1227 30 Trombe MC, l.aneelle G, Sicard AM 11984) Characterization of Strepto('o('cus pneunumiae mutant with altered electric transmembrane potential, JBacteriol 158, 1109-1114 31 Sautereau AM. Trombe MC (1986) Electric transmembrane potential mutation and resistance to the cationic and amphiphilic antitumoraldrugs derived from pyridocarbazole, 2-N-methylellipticinium and 2-N-methyl-9-hydro×yellipticinium, in Streptococcus pneumoniae. J Gen Microbiol ! 32, 2637-2641 32 Hein J (1990) Unified approach to alignment and phylogenies. Methods En:ymoi 183,626-644 33 Woese CR (1987) Bacterial evolution. Microbiol Rev 5 I, 221-271 34 Olsen GJ, Woese CR, Overbeek R (1994) The winds of (evolutionary~ change: breathing new life into microbiology. J Ba~'teriol 176. I-6 35 Martineau P, Saurin W, Hofnung M, Spurlino JC, Quiocho FA (19911) Progress in the identification of interaction sites on the periplasmic maltose binding protein from E ('oli. Biochimie 72, 397-4(12 36 Zhang Y, Conway C, Rosato M, Sub Y, Manson MD (1992) Maltose chemo. taxis involves residues in the N-tenninal and C-terminal domains on the same filce of maltose-binding protein. ,I Bmi Chem 267. 22813-22820 37 Mur?.in AG (1993) Can homolt)gous pa)teins evolve different enzymatic activities, TIBS 18, 4034(15 38 Eitinger T, Friedrich B (1091) Cloning, nucleotide sequence, and Iletcroh)gous expression o1' a Iligh-afliuity nickel transport gone from Ah'alig('n('s eutrophus, ,I Biol Chem 266, 3222-3227 39 Eitinger T, Friedrich B (1994) A topological model lot the high-aflinity nickel transporter ot' Ah',ligem.s eutrophus. Mol Mi('rohiol 12, 11125-1032 41) Macda M, Hidaka M. Nakamura A, Masaki H, Uozumi T (1994) Cloning, sequencing, and expression of Thermo/)hili(' Bacillus sp strain TB-90 urease gene complex in Es('herichia ~'oli, ,/lhwtcriol 176, 432-442 41 Hanson MS, Slaughter C, Hausen FJ (1992) The hbpA gene of Haemophilus inlluen:ae type b encodes a heme-binding lipoprotein conserved among heme- dependent tlaemophih,s species, h~,('t Immun, 60, 2257-2266

42 Stojiljkovic !, Hantke K (1994) Transport of haemin across the cytoplasmic membrane through a haemin-specific periplasmic binding-protein-dependent transport system in Yersinia enterocolitica. Mol Microbio/13, 719-732 43 Spurlino JC, Lu GY, Quiocho FA (1991) The 2.3-A resolution structure of the maltose- or maltodextrin-binding protein, a primary receptor of bacterial active transport and chemotaxis. J Biol Chem 266, 5202-5219 44 Quiocho FA, Vyas NK (1984) Novel stereospecilicity of the L-arabinose-binding protein. Nature 310, 381-386 45 Vyas NK, Vyas MN, Quiocho FA (1987) A novel calcium binding site in the galactose-binding protein of bacterial transport and chemotaxis. Nature 327, 635-638 46 Vyas NK, Vyas MN, Quiocho FA(1988) Sugar and signal-transducer binding sites of the Escheriehia coli galactose chemoreceptor protein. Science 242, 1290-1295 47 Zou JY, Flocco M, Mowbray SL (1993) The 1.7 ,~ refined X-ray structure of the periplasmic glucose/galactose receptor from Salmonella ~phimurium. J Mol Bioi 233,739-752 48 Mowbray SL, Cole LB (1992) The i,7 A X-ray structure of the periplasmic ribose receptor from Est'herichia t'oli. J Mol Bio1225, 155-175 49 Sack JS, Trakhanov SD, Tsigannik IH, Quiocho FA (1989) Structure of the L-leucine-binding protein refined at 2.4 A resolutiol~ and comparison with the Leu/lle/Val-binding p,otein structure. J Mo/Bio/206, 193-207 50 Sack JS, Saper MA, Quiocho FA (1989) Periplasmic binding protein structure and function; refined X-ray structure of the ieucine/isoleucine/valine-binding protein and its complex with leucine. J Mol Bio1206, 17 i- 19 I 51 Pflugrath JW. Quiocho FA(1988) The 2/~ resolution structure of the sulfatebinding protein involved in active transport in Salmonella typhimurium. J Mol Bio/200, 163-180 52 Luecke H, Quiocho FA (1990) High specificity of a phosphate transport protein determined by hydrogen bonds. Nature 347,402-406 53 Oh BH, Pandit J, Kang CH, Nikaido K, Gokcen S, Ames GFL, Kim SH (1993) Three-dimensional structures of the periplasmic lysine/arginine/ornithine-binding protein with and without a ligand. J Biol Chem 268, 11348-11355

54 Kang CH, Kim 3H, Nikaido K, Gokcen S, Ames GF (1989) Crystallization and preliminary X-ray studies of HisJ and LAO periplasmic proteins from Sahnonella typhimurium. ,I Mol Bio1207, 643-644 55 Trakhanov SD, Chirgadze N, Yusifov EF(IO89) Crystallization and preliminary X-ray crystaUographic data of a histidine-binding protein from Escherit'hia coli. J Mo/ Bio1207, 847-849 56 Tame JRH, Murshudov GN, D~)dson EJ, Neil TK. Dodson GG, Higgins CF, Wilkinson AJ (1994)The structural basis of sequence-independent peptide binding by OppA protein. St icme 2(54, 1578-1581 57 Dunten PW. Itarris Jil. Feiz V, Mowbray SL (1993) Crysttfilization and preliminary X-ray analysis of the periplasmic dipeptide binding protein li'om Es('ht.ri('hia ('oil. ,I Mol Bio123 I, !45-147 58 Charon M-H. Wu L-F, Piras C, de Pina K, Mandmnd-Berthelot MA, FontecillaCamps JC (1994) Crystallization and preliminary X-my diffraction study of the nickel-binding plx~tein NikA of Escheridtia co/i. ,I Mo/Bio1243, 353-355 59 Gardina P, Conway C, Kossman M, Manson M (1992) Aspartate and maltosebinding protein interact with adjacent sites in the Tar chemotactic signal transducer of Est'herichia (oli. J Bacteriol 174, 1528-1536 61) de Pina K (1992) Master Degree Thesis, University of Lyon !