Mol Gen Genet (1993) 239:188-195

© Springer-Verlag 1993

Cloning and sequencing of a gene involved in the synthesis of the capsular polysaccharide of Streptococcus pneumoniae type 3 Ernesto Garcia, Pedro Garcia, Rubens L6pez

Centro de Investigaciones Biol6gicas, CSIC, Vel/tzquez 144, 28006 Madrid, Spain

Received: 16 July 1992/Accepted: 15 December 1992

Abstract. A 4.5 kb ScaI chromosomal DNA fragment of a clinical isolate of Streptococcus pneumoniae serotype 3 was cloned in Eseherichia coli. Combined genetic and molecular analyses have allowed the localization, in a 781 bp EcoRV subfragment, of a gene (cap3-1) directly responsible for the transformation of an unencapsulated, serotype 3 mutant to the capsulated phenotype. Com- parison of the deduced amino acid sequence of CAP3-1 with the protein sequences compiled in the data banks revealed that the CAP3-1 polypeptide was highly similar to the amino-terminus of the GDP-mannose dehydroge- nase of Pseudomonas aeruginosa, an enzyme that partici- pates in the synthesis of the mucoid polysaccharide of this species. In addition, the 32 N-terminal amino acids of CAP3-1 perfectly matched structures common to NAD+-binding domains of many dehydrogenases. Our results indicate that the 4.5 kb ScaI fragment might also contain genes common to 13 different pneumococcal serogroups or serotypes tested. To the best of our knowl- edge, this is the first time that a gene of the capsular complex of S. pneumoniae has been cloned and se- quenced. The findings reported here provide new insights for the study of the molecular biology of the main viru- lence factor responsible for the pathogenesis of pneumo- coccal infections and might represent a basic step in the identification of cross-reactive antigens that should allow the preparation of new and improved vaccines.

Key words" Pneumoccocal capsules - Pneumococcus - Serotypes - Transformation

Introduction

In spite of the success of penicillin therapy, Streptococcus pneumoniae has remained, even in developed countries,

Communicated by W. Goebel

Correspondence to: R. Ldpez

a major etiologic agent of pneumonia and other severe diseases (Shan 1990). The work carried out by Austrian (1985) has underlined the importance of the development of an effective polyvalent vaccine for individuals at high risk of serious pneumococcal disease, and the use of a polyvalent vaccine, including 23 capsular polysaccharides, was approved in 1983 in the USA. Nevertheless, the first systematic trial of such vaccine has revealed, together with its partial effectiveness (Shapiro et al. 1991), the need for efforts to search for new and improved vaccines, with special emphasis on the identification of antigens common to a large number of pneumococcal serogroups or serotypes and that become protective in all the popula- tion at high risk (Broome and Breiman 1991). In this context, the reasons for the differences in virulence of the different capsular types are still unclear and the construc- tion of isogenic strains differing only in capsular types, together with the study of the genetics of capsule produc- tion, would be of great help.

Transformation of capsular types in S. pneumoniae was the first case of natural transformation to be de- scribed (Griffith 1928) and these studies led to the identi- fication of the transforming principle as DNA (Avery et al. 1944). Unfortunately, cloning of pneumococcal genes has been very difficult until recently (Dillard and Yother 1991) and no approaches for cloning the genes respon- sible for the synthesis of capsular polysaccharides of S. pneumoniae have been reported so far. No information is available on the molecular genetics underlying the synthesis of pneumococcal capsules of different antigenic types, but the cloning of the capsular genes of Escherichia coli (Roberts et al. 1988), Haemophilus influenzae (Kroll et al. 1989) and Neisseria meningitidis (Frosch et al. 1989), as well as genetic studies in pneumococcus (for a review, see M/ikel/i and Stocker 1969), suggested that these genes are genetically linked. To gain insight into the molecular mechanisms of capsular biosynthesis and transport in pneumococcus we took advantage of the technology developed in our laboratory to isolate, clone, sequence, and manipulate (Diaz et al. 1990) pneumococ- cal genes involved in the synthesis of the autolysins (Gar-

cia et al. 1985) and we have been able in E. coli to clone and identify a gene belonging to the capsular biosynthet- ic pathway of S. pneumoniae type 3.

Materials and methods

Bacterial strains and growth conditions. Relevant pneumococcal strains are described in Table 1. Encap- sulated, clinical pneumococcal isolates were kindly provided by J. Casal and A. Fenoll from the Laboratory of Pneumococci (Centro Nacional de Microbiologia, In- stituto de Salud Carlos III, Majadahonda, Madrid, Spain). Serogrouping or serotyping was carried out by the Quellung reaction as described by Lund and Henrich- sen (1978), with use of 46 antisera provided by the Staten Seruminstitut (Copenhagen). Also, we employed S. oralis NCTC 11427 (type strain) as a source of DNA. The E. coli strain TG1 (Sambrook et al. 1989) was the host for recombinant plasmids. S. pneumoniae and S. oralis were grown in liquid C medium (Tomasz 1970) contain- ing 0.08% yeast extract (C+Y) or 0.08% bovine serum albumin (C+Y+Alb ) without shaking or on recon- stituted tryptose blood agar base plates (Difco Laborato- ries) supplemented with 5% (v/v) defibrinated sheep blood. E. coli was grown in LB medium (Sambrook et al. 1989) with shaking.

DNA manipulation, plasmid construction, and genetic transformation. The DNA of the pneumococcal phage Cp-1 was prepared by proteinase K digestion of purified virions (Ronda et al. 1981). Sinai-digested and dephos- phorylated pUC18 was obtained from New England Biolabs. Plasmid pGL32 has been previously described (L6pez et al. 1986). The preparation of pneumococcal DNA and the transformation procedure for S. pneu- moniae have been described elsewhere (Tomasz 1970). Plasmid DNA was prepared by the alkaline method, including equilibrium centrifugation in CsCl-ethidium bromide gradients (Sambrook et al. 1989). Transforma- tion of E. coli was carried out using the RbC1 method (Sambrook et al. 1989). Restriction endonucleases and T4 DNA ligase were purchased from Amersham Searle, New England Biolabs or Boehringer Mannheim. All

Table 1. Relevant pneumococcal strains used in this study

Strains Genotype/phenotype Source/reference

M22 hex-4, end-l, exo-2, Lyt +, $2 - Ronda et al. (1988)

This work a This work b This work c J. Casal d

M22AlytA32 hex-4, end-l, exo-2, Lyt - , $2- M23 Lyt- , $3 + M24 Lyt- , $3- 406 Lyt +, $3 +

" Strain constructed by transformation of M22 with pGL32 b Strain constructed by transformation of M22AlytA32 with D N A from strain 406 c A spontaneous $3 mutant of strain M23 d A clinical type 3 isolate

189

these enzymes were used according to the recommenda- tions of the suppliers.

Restriction fragments and plasmids were analyzed by agarose gel electrophoresis. When required, restriction fragments were purified after electrophoresis using a Geneclean kit (Bio 101, La Jolla, Calif.) In some experi- ments, competent cells of S. pneumoniae were trans- formed with restriction fragments separated by low-melt- ing point (LMP) agarose gel electrophoresis as described elsewhere (Barany and Tomasz 1980).

Sanger dideoxy sequencing and sequence analysis. Both strands were sequenced by the dideoxy chain-termination method (Sanger et al. 1977) using ct-[35S]dATP (Amer- sham International, UK) and a T7 DNA polymerase sequencing kit from Pharmacia. Nucleotide and amino acid sequences were analyzed using DNASTAR software (Madison; Wis.) operated on an IBM PC system. Amino acid alignments were done with the AALIGN program and the PAM 250 matrix of Dayhoff (reviewed by Rob- son and Garnier 1988).

DNA hybridization. Dot blot hybridization was carried out as follows. DNA (about 1 gg) was denatured by heating to 37 ° C for 10 rain in 0.5 M NaOH, and filtered through a nylon membrane filter (Nytron-N, Renner GmbH, FRG) using a dot blotting apparatus (Bio-Rad). The membrane was wetted in denaturing solution (1.5 M NaC1, 0.5 M NaOH) for 1 rain, transferred to neutraliz- ing solution (1.5 M NaC1, 0.5 M TRIS-HC1, pH 7.2, 1 mM EDTA) for 1 min, and air dried. Afterwards, the DNA was covalently linked to the membrane by UV irradiation on a transilluminator. Hybridization was car- ried out at 65 ° C as previously described (Sambrook et al. 1989). Radioactive probes were synthesized by nick- translation using c~-[32p]dCTP and the random primer method. Radioactive spots were detected with Kodak X-Omat films and DuPont Cronex Lightning Plus inten- sifying screens at - 7 0 ° C.

Results

Construction of an $3- mutant as recipient for the localization of capsular genes

For cloning purposes, it is important that the phenotype of transformants is easily distinguishable from that of the mutant recipient. In the case of pneumococcal capsules, type 3 capsulated strains show a mucoid appearance that can be readily differentiated from that of unencapsulated mutants. In addition, since it has been shown that trans- formation of S. pneumoniae from one capsular type to another requires the transfer of a large segment of DNA (Bernheimer et al. 1967), it was important to use as recipient strain in transformation experiments, an unen- capsulated type 3 ($3 -) mutant differing by no more than a single mutation from the donor DNA. On the other hand, it has been shown that natural type 3 pneumococ- cal isolates are usually untransformable (Yother et al. 1986), even when the capsule is removed (Ottolenghi-

190

Fig. 1. Photograph of a transformed culture of strain M24. A culture of competent cells of the unencapsulated ($3-) pneumococcal strain M24 was transformed with pLGL1. The culture was plated on tryptose sheep blood agar and in- cubated at 37 ° C for 16 h. Highly mucoid ($3 ÷) transformed cells (large colonies) can be seen among those of the unencapsulated recipient. Magnification: x20

Nightingale 1972). Consequently, it was necessary to construct an isogenic, transformable, $3- strain, and the reason for following the strategy described below was threefold: (1) all the highly transformable pneumococcal strains are rough derivatives of type 2 strains (mostly derived from R36A, Avery's strain); (2) it has been estab- lished that unencapsulated derivatives of type 2 produce exclusively type 3 transformants when treated with donor DNA from type 3 cells; and (3) it is classically recognized that typical inter-type transformation in pneumococcus consists of the replacement of one cluster of type-specific genes by another cluster (Mfikel/i and Stocker 1969).

Consequently, we transformed the pneumococcal strain M22 (lytA +) with plasmid pGL32, which contains a frame-shift mutation in the lytA gene coding for the S. pneurnoniae lytic amidase, which makes this mutant una- ble to synthesize this enzyme (Ldpez et al. 1986). The availability of such a strain (designated M22AlytA32) (Table 1) should allow long periods of incubation at 37°C with no risk of suffering the autolytic process characteristic of S. pneumoniae, thus facilitating the screening of transformed clones for capsules following the technique described below. Subsequently, we trans- formed the M22AlytA32 strain (a $2- strain) with DNA obtained from a clinical type 3 isolate of S. pneumoniae (strain 406). The capsular phenotype of transformants was always confirmed by the Quellung test (Lund and Henrichsen 1978). A capsulated $3 + transformant was isolated (strain M23) and incubated in C ÷ Y medium for 24-36 h, and periodically tested on blood agar plates for appearance of spontaneous unencapsulated mutants. A rough ($3-) mutant (strain M24) was isolated after several cycles of culture. M24 did not show any evidence of capsule by the Quellung reaction. On the other hand, transformation of M24 with chromosomal DNA from strain 406 gave a high frequency of capsular transfor- mants (1-3%) (not shown). M24 is a stable strain in our experience since, despite frequent subculturing for more than 1 year, the strain has lost none of its distinctive features, and no $3 + revertant mutants have appeared in serial transfers.

Identification of restriction fragments containin9 capsular genes

To identify restriction fragments containing genes in- volved in capsular biosynthesis, DNA from the pneumo- coccal strain 406 was digested with several restriction enzymes, electrophoresed in 0.7 % agarose gels, and DNA fragments of various sizes were purified and used to transform competent cells of M24 to the capsulated phenotype. $3 ÷ transformants were observed with frag- ments of about 10 kb, 6.5 kb, and 4.5 kb when DNA from strain 406 was digested with EcoRI, PstI and ScaI, respectively.

Clonin 9 of pneumococcal capsular 9enes in E. coli

DNA fragments of about 4.5 kb, obtained by treatment of the DNA from strain 406 with ScaI, were ligated to Sinai-digested and dephosphorylated pUC18. The liga- tion mixture was used to transform E. coli TG 1 to am- picillin resistance, and those clones harboring recom- binant plasmids were identified by their white phenotype on plates containing 5-bromo-4-chloro-3-indolyl-[3-D- galactopyranoside (X-gal) and isopropyl-13-D-thiogalac- topyranoside (IPTG). It has been well documented that fragments of pneumococcal DNA cloned into heterolo- gous plasmids are able to transform competent cells of pneumococci (Garcia et al. 1985). Alkaline preparations of the recombinant E. coli plasmids obtained from the above transformants were tested for their ability to trans- form strain M24 to the $3 ÷ phenotype (Fig. 1). From about 300 recombinant clones, one transformant was isolated, which contained a plasmid (hereafter designated pLGL1) harboring a pneumococcal DNA insert of 4.5 kb (Fig. 2).

To localize within the 4.5 kb cloned fragment the region responsible for the transformation of the unen- capsulated M24 strain to the capsulated $3 +phenotype, pLGL 1 was digested with several restriction enzymes and the different DNA fragments were tested for their capac-

191

Ps N HSpHcXB EVPsHc P S HNCEVA EVPC NBgBNcSp EV KSE

if ( I I

1 2 kb I y / / / I / l l / / I / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / l PS~[

I I / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / ~ / - / I )7CD

I V / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / m ~l ind I1]

, / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / i i i BamHI

pLGL1

Ir/////////////////Ai i C/dI

F / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / J P

Fig. 2. Physical map of plasmid pLGL1 and localization of the region responsible for the transformation of M24 to the $3 + phenotype. Bars correspond to the 4.5 kb ScaI fragment of pneumococcal DNA, and the heavy line to the vector )lasmid pUC 18. Plasmid pLGL 1 was digested with the indicated restriction endonucleases, the DNA fragments were separated by electro-

I AccI

phoresis in low-melting point agarose and used to transform strain M24. Hatched bars correspond to the fragments which recombined with the $3- mutation of strain M24 to restore the $3 + phenotype. A, AccI; B, BamHI; Bg, BgIII; C, ClaI; E, EcoRI; EV, EcoRV; H, HindIII; Hc, HincII; K, KpnI; N, NsiI; Nc, NcoI; P, PvuII; Ps, PstI; S, SacI; Sp, SphI; X, XbaI

Ps H CEVAEVPCSp ( I h rUI

lkb

I pLGL1 Z2kb lacZ" H C EVAEVPC Sp

5.5kb EVA EV P t acz "

pLGL5 lac~,=V~_ P 4.0 k b

FL(.q pLGL4 lacZ" 3.1 kb

A

pLGL5 laeZ" 3.4 kb

~LSL15 lacZ---" 2.9 kb

Fig. 3. Construction of plasmids pLGL2, pLGL3, pLGL4, pLGL5 and pLGL 15. Plasmid pLGL2 is a subclone of pLGL 1, and pLGL3 is a derivate containing the 1.3 kb ClaI fragment of pLGL2. Plas- mids pLGL4 and pLGL5 are subclones of pLGL3, pLGL15 is a subclone of pLGL5. Open bars represent pneumococcal DNA, and closed bars, the vector plasmid pUC18. Abbreviations are as in Fig. 2. Arrows indicate the direction of transcription of the L3-ga- lactosidase (lacZ') gene of the vector plasmid

ity to transform M24 to the capsulated phenotype. The results obtained from these experiments showed that the transforming region was localized between ClaI and AccI sites (Fig. 2). The 1.4 kb ClaI fragment was subcloned from pLGL1 into AccI-digested pUC18 to construct pLGL3 (Fig. 3). In addition, two derivatives of pLGL3 (pLGL4 and pLGL5) were also constructed, and trans- formation experiments revealed that pLGL5 still trans- formed M24 to the encapsulated phenotype, whereas pLGL4 did not (not shown). Plasmid pLGL5 contains the 0.8 kb EcoRV fragment ofpLGL3 cloned into SmaI- digested pUC 18, and pLGL4 was constructed by remov- ing this fragment from pLGL3 (Fig. 3).

Nucleotide sequence of the 9ene responsible for transformation to the $3 + phenotype

The 781 bp EcoRV insert of pLGL5 was sequenced (Fig. 4). Two putative open reading frames (ORF), orien- ted oppositely, were identified, one truncated at the 5' end (ORF2). ORF1 contains a possible initiation codon (ATG) at nucleotide 145, and terminates at a TAA stop codon at position 559. If the initiation signal is used, ORE1 could code for a protein of 138 amino acids (predicted Mr = 15 347). Eleven nucleotides upstream of the initiation codon ATG, a putative Shine-Dalgarno sequence (GAGG) has been located. Although compila- tions of ribosome binding site sequences have shown that spacings of less than 5 and more than 9 nucleotides are rare (Stormo 1982), in vivo experiments suggest that within the range of 5 to 13 nucleotides there is only a small effect on translation (Shepard et al. 1982; Wood et al. 1984). Canonical - 1 0 (TATAAT) and - 3 5 (TTGACA) promoter regions were found 64 nucleotides upstream from the ATG start codon. These regions were spaced by 16 nucleotides. ORF2 seemed to lack the N-terminal amino acid sequence and contains > 365 bp with a TGA termination codon at position 416. Analysis of ORF2 using the CodonPreference program (The Wis- consin Package from the Genetics Computer Group) and the codon usage table for S. pneumoniae (Martin and Claverys 1991), did not conclusively answer the question of whether or not this ORF represents a potential protein coding region (not shown).

To determine which of the two ORFs contained a functional sequence able to complement the mutation in the unencapsulated phenotype, the 0.24 kb DraI-BstYI fragment of pLGL5 (Fig. 4) was cloned into pUC18 (previously digested with Sinai and BamHI) to create pLGL15 (Fig. 3). Transformation experiments demon- strated that the mutation present in the pneumococcal strain M24 was located in this fragment (data not shown). Therefore, it is evident that ORF1 represents the coding region of the gene, hereafter designated cap3-1, responsible for this transformation. This experiment

192

EcoRV -35 GATAT CTTTT CAAAG CTGATACTAAGG CA CAAA_AJiAA_AGT TTGATAT T C C C C TT GA CAAT 60 CTATAGAAAAGTTTCGACTATGATTC CGTGTTTTTTTTCAAACTATAAGGGGAACTGTTA

-10 Dra [

AGATAAAATTATTATATAATTAAACTATTGCTTTTTAAATAAAGTGAGAATATTAATAAT 120 TCTATTTTAATAATATATTAATTTGATAACGAAAAATTTATTTCACTCTTATAATTATTA

CAP3-1~ M K I A I A G S G Y V G

G C A G A G A A A ~ . ~ A C T G T A G T A A A A T G A A A A T T G C C A T T G C A G G A A G T G G T T A T G T A G G A 1 8 0 CGTCTCTTTCT CCTGACATCATTTTACTTTTAACGGTAACGTC CTTCAC CAATACATCCT

L S L A V L L A Q H H E V K V I D V I K CTGTCTTTAGCGGTGCTACTAGCTCAGCATCATGAAGTTAAGGTCATTGATGTTATAAAG 240 GACAGAAATCGC CACGATGATCGAGTCGTAGTACTTCAATTCCAGTAACTACAATATTTC

D K V E S I N N R K S P I K D E A I E K GATAAGGTAGAGTCGATAAACAATAGAAAATCTCCAATTAAGGATGAGG CGATTGAGA-A-A 300 CTATTCCATCTCAGCTATTTGTTATCTTTTAGAGGTTAATTC CTACTCCG CTAACTCTTT

BstYI Y L V E K E L N L E A S L D P A H V Y K

TACTTAGTTGAAAAAGAGTTGAATCTTGAAGCCTCCTTAGATCCTGCACACGTTTATAA-A 360 ATGAATCAACTTTTTCTCAACTTAGAACTTCGGAGGAATCTAGGACGTGTGCAAATATTT

Acc[ D V E Y A I I A T P T N Y D V D L N Q F

GACGTGGAGTATGCTATTATTGCTACTCCGACTAATTATGATGTAGACTTAAATCAGTTT 420 CTGCACCTCATACGATAATAACGATGAGG CTGATTAATACTACATCTGAATTTAGTCAAA

* N

D T S S V E A A I K T C M E Y N D T C T GATACATCTTCAGTTGAAGCTGCTATCAAGACTTGTATGGAATATAATGATACTTGTACA 480 CTATGTAGAAGTCAACTTCGACGATAGTTCTGAACATACCTTATATTACTATGAACATGT S V D E T S A A I L V Q I S Y L S V Q V

I V I K S L P P K F R N L S K T F S Q A ATCGTAATCAA/hAGTCTTCCTCCAAAATTTCGAAACCTCTCTAAAACCTTTTCCCAAGCT 540

TAGCATTAGTTTTCAGAAGGAGGTTTTAAAGCTTTGGAGAGATTTTGGAA/LAGGGTTCGA Fig . 4. Nucleotide sequence of the I T I L L R G G F N R F R E L V K E W A 781 b p E e o R V insert of pLGL5 and de-

L L L Y S S * rived amino acid sequence (in one-letter CTATTACTGTATTCCTCATAATTTCCATTATCATCAAATATTTTTACTTTGGTTGTAAAT 600 code). The bars delimit a possible pro- GATAATGACATAAGGAGTATTAAAGGTAATAGTAGTTTATAAAAATGAAACCAACATTTA moter for the cap3-1 gene; the putative R N S Y E E Y N G N D D F I K V K T T F ribosome binding site of the cap3-1 gene ATGTTAATAATTAGCCTATTAACGACGTCTCTATAATCATTTTCTAAGTCAAAGTTTTCA 660 is indicated by the broken overline. The TACAATTATTAATCGGATAATTGCTGCAGAGATATTAGTAAAAGATTCAGTTTCAAAAGT predicted open reading frames ( C A P 3 - I I N I I L R N V V D R Y D N E L D F N E and ORF2) and the direction of tran- GATTGATTACTGTAACTTTCAAGAGCAGACTCTATCTTACTTTCACTCAAAAATAATTTA 720 scription are also shown. The termina- CTAACTAATGACATTGAAAGTTCTCGTCTGAGATAGAATGAAAGTGAGTTTTTATTAAAT tion codons of both ORFs are indicated S Q N S Y S E L A S E I K S E S L F L K b y asterisks. Relevant restriction sites are

E c o R V underlined on one of the strands. The TCCTCGTATCTTATCTCCTCATCAATAACAAGTTTATTATCAATTAGAAATCCTCGATATC 781 sequence appears in the E M B L / AGGAGCATAGAATAGAGGAGTAGTTATTGTTCAAATAATAGTTAATCTTTAGGAGCTATAG G e n B a n k / D D B J nucleotide sequence D E Y R I E E D I V L K N D I L F G R Y data libraries under the accession num-

4ORF2 b e r Z 1 2 1 5 9

fully confirmed the previous observation (Fig. 2) that the left AccI fragment of pLGL1 transforms M24 to the capsulated phenotype.

Sequence similarity comparison

Similarity searches of the ORFs reported above with protein sequences compiled in the NBRF database and in the GENPRO (ORFs translated from sequences present in Genbank) showed no obvious similarity be- tween the incomplete ORF2 and any other previously published or unpublished sequences. However, the poly- peptide encoded by cap3-! (hereafter CAP3-1) was sim- ilar to the amino-terminus of the GDP-mannose dehy- drogenase (gene algD) of Pseudomonas aeruginosa (Fig. 5A). In fact, 42 out of the 138 residues (30.4%) are

identical and many other amino acids are conservative substitutions, which further increase the overall similar- ity. In order to analyze the significance of this, these proteins were subjected to RDF analysis (Lipman and Pearson 1985), the z value obtained, 8.1, after 40 random shuffles with a k-tuple value of 1, indicated that the similarity was significant. The gene algD participates in the production of the mucoid, alginate-containing cap- sule of P. aeruginosa (Deretic et al. 1987). It is interesting to note that the 33 N-terminal amino acid residues of CAP3-1 perfectly matched the j3a[3-fold common to the NAD+-binding domains of many dehydrogenases (Scrutton et al. 1990) centered on the highly conserved sequence Gly-X-Gly-X-X-Gly (where X is any amino acid). Furthermore, CAP3-1 retains several other charac- teristics of the NAD +-binding domain: (1) a hydrophilic residue (Lys) at the N-terminus; (2) a hydrophobic core

A CAP3-1MKIAIAGSGYVG-LSLAVLLAQHHEVKVIDVIKDKVESINNRKSPIKDEAIEKYL . . . . VEKELNLEASLDPAHV

GMDH MRISIFGLGYVGAVCAGCLSARGHEVIGVDVSSTKIDLINQGKSPIVEPGLEALLQQGRQTGRLSGTTDFKKAVL

CAP3-1 YKDVEYAIIATPT--NYDVDLNQFDTSSVEAAIKTCMEYNDTCTIVIKSL-PPKFRNLSKTFSQALLLYSS

GMDH DSDVSFICVGTPSKKNGDLDLGYIETVCREIGF-AIREKSERHTVVVRSTVLPGTVNNVVIPLIEDCSGKK

B An| ~ ~ ~ • [] • m o

2 K I A I A G S G Y V G L S L A V L L A Q H H E V K V I D V I K D 33

193

Fig. 5A, B. Comparison of the amino acid sequences of CAP3-1 and part 70 of the GDP-mannose dehydrogenase (GMDH) of Pseudornonas aeruginosa

75 and identification of a NAD +-bind- ing site. A Colons and periods in-

138 dicate identical matches and conser- vative substitutions, respectively (Lip- man and Pearson 1985). B The sym-

145 bols indicate the positions, in CAP3- 1, of amino acids characteristic of NAD+-binding sites: A, basic or hydrophilic; i , small and hydro- phobic; a, glycine; ©, acidic (Wierenga et al. 1986)

composed of six small residues (Leu-Ala-Val-Leu-Leu- Ala); and (3) a negatively charged residue (Asp) at the C-terminus of the motif (Fig. 5B) (Wierenga et al. 1986).

Plasmid pLGL1 as a probe to identify capsular genes common to the different pneumococcal serotypes

Preliminary indications for the involvement of the re- gions flanking the cap3-1 gene in capsular functions came from the observation that a 6.5 kb PstI fragment of strain 406 DNA, harboring the cap3-1 gene, was also capable of transforming rough spontaneous mutants (other than M24), that were not transformable by pLGL1, to the capsular serotype 3 phenotype (not shown). To evaluate whether the cloned 4.5 kb ScaI

A ,

12

10 t

5 8

Jl

16 17 g

B I 2 5 4 5 6

I0 f 7 8 9 o

0 12 13 t4 [5 16 i7

19 18 .m,L

Fig. 6A, B. Dot blot analyses of DNAs obtained from various strains of Streptococcus pneurnoniae expressing different capsular antigens. DNAs were blotted onto nylon membrane filters. The filters were incubated with 32p-labeled pLGL1 (A) or 3zP-labeled pLGL3 (B). DNA samples from: 1, S. oralis; 2, no DNA; 3, S. pneumoniae M31 (Sfinchez-Puelles et al. 1986); 4, phage Cp-1; 5, calf thymus; 6-18, clinical pneumococcal isolates belonging to serotypes or serogroups 1, 2, 3, 4, 5, 6, 7, 8, 9, 14, 19, 23, and 33, respectively; 19, plasmid pLGL1

fragment of pLGL1 includes capsular genes common to different pneumococcal serotypes, we tested for hy- bridization between pLGL1 or pLGL3 (Fig. 3), and DNAs prepared from several pneumococcal isolates be- longing to different serotypes or serogroups. Since 85 different serotypes have been described in S. pneumoniae (van Dam et al. 1990), only those reported to be most frequent (Fenoll et al. 1991; Shapiro et al. 1991) were used in this work.

The DNAs purified from several pneumococcal iso- lates belonging to 13 different serogroups or serotypes as well as samples prepared from species taxonomically related (i.e.S. oralis) (Bentley et al. 1991) or unrelated (i.e. calf thymus) to S. pneumonia< as well as from the pneumococcal phage Cp-1, were blotted onto a nylon membrane and probed with radioactive pLGL 1 (Fig. 6A) or pLGL3 (Fig. 6B). The dot blotting experiments in- dicated that the DNAs from all the pneumococcal se- rogroups assayed hybridized with pLGL1, whereas the heterologous DNAs did not. On the contrary, only DNAs purified from clinical isolates belonging to sero- types 3 and 5 gave hybridization signals with pLGL3.

Discussion

The possibility of deriving isogenic strains that differ by a single trait, as illustrated in this work by the prepara-

t ion of a pneumococcal strain that apparently contains a single mutation in a gene implicated in the synthesis of the capsular serotype 3, provides a powerful tool for investigating the genes responsible for the synthesis of this virulence factor. In this report we have described the molecular cloning of some determinants of capsular polysaccharide synthesis of S. pneumoniae serotype 3. By cloning in a multicopy plasmid of E. coli, it has been possible to identify a 781 bp EcoRV fragment of pneumococcal DNA capable of transforming an unen- capsulated mutant of serotype 3 to the capsulated phenotype (Fig. 3); a result that indicates that this region is essential, but, most probably, not sufficient, for production of the capsule of serotype 3.

The determination of the nucleotide sequence of this fragment together with subcloning and transformation experiments showed that the mutation responsible for the unencapsulated phenotype of strain M24 is located

194

in a gene that was designated cap3-1. Interestingly, the deduced amino acid sequence of cap3-1 was very similar to that of the amino-terminus of the GDP-mannose dehydrogenase of P. aeruginosa, which is implicated in the synthesis of the polysaccharide capsule of this species responsible for the majority of complications in patients suffering from cystic fibrosis (Deretic et al. 1987), and contains a consensus motif for NAD + binding, charac- teristic of many dehydrogenases (Scrutton et al. 1990). Type 3 capsular polysaccharide contains cellobiuronic acid in the repeating unit, a disaccharide consisting of ~-glucuronic acid ~(1 -~4) linked to ~-glucose. The cello- biuronic acid units are 13(1 --~3) linked to each other (van Dam et al. 1990). Since it has been demonstrated that at least one NAD-requiring dehydrogenase, namely UDP- glucose dehydrogenase, is directly involved in the me- tabolic pathways of type 3 capsule polysaccharide syn- thesis, being the key enzyme for converting UDP-glucose to UDP-glucuronic acid (Mills and Smith 1962), it is tempting to speculate that cap3-1 might be the gene coding for this enzyme. Nevertheless, further studies have to be carried out to identify this gene precisely.

In gram-negative bacteria (e.g.E. coli, H. influenzae, and N. rnenin9itidis) genes that are involved in sugar biosynthesis and polymerization are flanked by those for translocation of the capsule to the cell surface and the postulated postpolymerizational modification of the cap- sule (Frosch et al. 1989; Kroll et al. 1989; Roberts et al. 1988). Classical transformation experiments of the pneumococcal capsular type 2 into the capsular type 3 suggested that these capsular genes should be clustered, and that recombination between the conserved capsular biosynthetic genes would be expected to result in the conversion of one capsular type into another (Avery et al. 1944; Mfikel/i and Stocker 1969). The blotting experi- ments discussed here using recombinant plasmids con- taining fragments of pneumococcal DNA of different lengths, suggest that the 4.5kb ScaI fragment of pneumococcal DNA cloned into pLGL1 may include genes that are specific for the biosynthesis of the type 3 and 5 serotypes and genes that appear to be conserved in the different capsular serotypes (Fig. 6). On the other hand, pLGL3 also hybridized with DNA isolated from a serotype 5 strain (Fig. 6B). Provided that type 5 poly- saccharide contains glucuronic acid as does type 3 cap- sule (van Dam et al. 1990), it is possible that both sero- types share similar biosynthetic pathways. In spite of this apparent similarity, chromosomal DNA obtained from the type 5 pneumococcus was not capable of transform- ing strain M24 to the encapsulated phenotype (data not shown).

The increasing level of antibiotic resistance in S. pneu- moniae indicates the need for a return to the study of some of the obscure phenomena underlying the behavior of this common human disease (Klugman 1990). Since the pnemnococcal capsular polysaccharide is the main bacterial factor that has been proven to contribute to pathogenesis (Johnston 1991), understanding the molec- ular biology of this virulence factor will provide informa- tion about potential targets for future therapies. The data reported here represent the first step in the elucidation of

the mechanisms that make S. pneumoniae such a devas- tating microorganism.

Acknowledgements. We thank J.L. Garcia for his critical reading of the manuscript and valuable suggestions, and F. de Pablo for help with the preparation of the manuscript. We thank E. Cano and M. Carrasco for skillful technical assistance, M. Garcla for his help with the computer work, and A. Hurtado for the art work. Also we appreciate the help of D. Vicioso in performing the Quellung test. This work was supported by a grant from Comisi6n Interministerial de Ciencia y Tecnologia (SAL91-0898-C02~01).

References

Austrian R (1985) Life with the pneumococcus: notes from the bedside, laboratory and library. University of Pennsylvania Press, Philadelphia

Avery OT, MacLeod CM, McCarthy M (1944) Studies on the chemical nature of the substance inducing transformation of pneumococcal types. I. Induction of transformation by a deo- xyribonucleic acid fraction isolated from pneumococcus type III. J Exp Med 79:137-157

Barany F, Tomasz A (1980) Genetic transformation of Streptococ- cus pneumoniae by heterologous plasmid deoxyribonucleic acid. J Bacteriol 144:698-709

Bentley RW, Leigh JA, Collins MD (t991) Intrageneric structure of Streptococcus based on comparative analysis of small-subunit rRNA sequences. Int J Syst Bacteriol 41:487-494

Bernheimer HP, Wermundsen IE, Austrian R (1967) Qualitative differences in the behavior of pneumococcal deoxyribonucleic acids tranforming to the same capsular type. J Bacteriol 93: 320-333

Broome CV, Breiman RF (1991) Pneumococcal vaccine: Past, present, and future, N Engl J Med 325:1506-1508

van Dam JEG, Fleer A, Snippe H (1990) Immunogenicity and immunochemistry of Streptococcus pneumoniae capsular poly- saccharides. Antonie Van Leewenhoek 58:1-47

Deretic V, Gill JF, Chakrabarty AM (1987) Pseudomonas aerugino- sa infection in cystic fibrosis: nucleotide sequence and transcrip- tional regulation of the algD gene. Nucleic Acids Res 15: 4567-4581

Diaz E, Ldpez R, Garcla JL (1990) Chimeric phage-bacterial en- zymes: A clue to the molecular evolution of genes. Proc Natl Acad Sci USA 87:8125-8129

Dillard JP, Yother J (1991) Analysis of Streptococcus pneumoniae sequences cloned into Escherichia coti: effect of promoter strength and transcription terminators. J Bacteriol 173:5105-5109

Fenoll A, Martln Bourgon C, Mufioz R, Vicioso D, Casal J (1991) Serotype distribution and antimicrobial resistance of Strepto- coccus pneumoniae isolates causing systemic infections in Spain, 1979-1989. Rev Infect Dis 13:56-60

Frosch M, Weisgerber C, Meyer TF (1989) Molecular characteriza- tion and expression in Escheriehia coli of the gene complex encoding the polysaccharide capsule of Neisseria meningitidis group B. Proc Natl Acad Sci USA 86:1669-1673

Garcla E, Garcia JL, Ronda C, Garcia P, L6pez R (1985) Cloning and expression of the pneumococcat autolysin gene in Eseheri- chia coIi. Mol Gen Genet 201:225-230

Griffith F (1928) The significance of pneumococcal types. J Hyg (London) 27:113d 59

Johnston RB Jr (1991) Pathogenesis of pneumococcal pneumonia. Rev Infect Dis 13 (Suppl 6): $509-$517

Klugman KP (1990) Pneumococcal resistance to antibiotics. Clin Microbiol Rev 3:17t-196

Kroll JS, Zamze S, Loynds B, Mo×on ER (1989) Common or- ganization of chromosomal loci for the production of different capsular polysaccharides in Haemophilus influenzae. J Bacteriol 171 : 3343-3347

195

Lipman D J, Pearson WR (1985) Rapid and sensitive protein sim- ilarity searches. Science 227:1435-1441

Ldpez R, S/mchez-Puelles JM, Garcia E, Garcia JL, Ronda C, Garcia P (1986) Isolation, characterization and physiological properties of an autolytic-defective mutant of Streptococcus pneurnoniae. Mol Gen Genet 204:237-242

Lund E, Henrichsen J (1978) Laboratory diagnosis, serology and epidemiology of Streptococcus pneumoniae. Methods Microbiol 12 : 241-262

M/ikelfi PH, Stocker BAD (1969) Genetics of polysaccharide bio- synthesis. Annu Rev Genet 3:291-322

Martin B, Claverys JP (1991) Codon usage patterns for Streptococ- cuspneurnoniae and Escherichia coli. In: Dunny GM, Cleary PP, McKay LL (eds) Genetics and molecular biology of Streptococ- ci, Lactococci, and Enterocci. American Society for Microbiol- ogy, Washington, DC, pp 295-296

Mills GT, Smith EE (1962) Biosynthetic aspects of capsular forma- tion in the pneumococcus. Brit Med Bull 18:27-30

Ottolenghi-Nightingale E (1972) Competence of pneumococcal iso- lates and bacterial transformation in man. Infect Immun 6: 785-792

Roberts IS, Mountford R, Hodge R, Jann KB, Boulnois GJ (1988) Common organization of gene clusters for production of dif- ferent capsular polysaccharides (K antigens) in Escherichia coll. J Bacteriol 170:1305-1310

Robson B, Garnier J (1988) Introduction to proteins and protein engineering. Elsevier, Amsterdam

Ronda C, Ldpez R, Garcia E (1981) Isolation and characterization of a new bacteriophage, Cp-1, infecting Streptococcus pneu- rnoniae. J Virol 40:551-559

Ronda C, Garcia JL, Ldpez R (1988) Characterization of genetic transformation in Streptococcus oralis NCTC 11427: Ex- pression of the pneumococcal amidase in S. oralis using a new shuttle vector. Mol Gen Genet 215:53-57

Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: A laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York

S/mchez-Puelles JM, Ronda C, Garcia JL, Garcia P, Ldpez R, Garcla E (1986) Searching for autolysin functions. Charac- terization of a pneumococcal mutant deleted in the lytA gene. Eur J Biochem 158 : 289-293

Sanger F, Nicklen S, Coulson AR (1977) DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74: 5463-5467

Scrutton NS, Berry A, Perham R (1990) Redesign of the coenzyme specificity of a dehydrogenase by protein engineering. Nature 343 : 38-43

Shann F (1990) Pneumococcus and influenza. The Lancet 335:898-901

Shapiro ED, Berg AT, Austrian R, Schroeder D, Parcells V, Mar- golis A, Adair RK, Clemens JD (1991) The protective efficacy of polyvalent pneumococcal polysaccharide vaccine. N Engl J Med 325 : 1453-1460

Shepard HM, Yelverton E, Goeddel D (1982) Increased synthesis in E. coli of fibroblast and leukocyte interferons through altera- tions in ribosome binding sites. DNA 1:125-131

Stormo GD (1982) Translation initiation. In: Reznikoff W, Gold L (eds) Maximizing gene expression Butterworth, Boston, pp 195-224

Tomasz A (1970) Cellular metabolism in genetic transformation of pneumococci: Requirement for protein synthesis during induc- tion of competence. J Bacteriol 101 : 861-871

Wierenga RK, Terpstra P, Hol WGJ (1986) Prediction of the occur- rence of the ADP-binding [3eq3-fold in proteins, using an amino acid sequence fingerprint. J Mol Biol 187:101-107

Wood CR, Boss MA, Patel TP, Emtage JS (1984) The influence of messenger RNA secondary structure on expression of an immu- noglobulin heavy chain in Escherichia coli. Nucleic Acids Res 12:3937-3950

Yother J, McDaniel LS, Briles DE (1986) Transformation of encap- sulated Streptococcus pneumoniae. J Bacteriol 168 : 1463-1465