Cloning and Characterization of Two Serratia marcescens Genes ...

5 downloads 997 Views 419KB Size Report
Mailing address: Department of Microbiol- ... Electronic mail address: regue .... EMBL version 34 databases by using the BLAST network service at the National.
JOURNAL OF BACTERIOLOGY, Oct. 1996, p. 5741–5747 0021-9193/96/$04.0010 Copyright q 1996, American Society for Microbiology

Vol. 178, No. 19

Cloning and Characterization of Two Serratia marcescens Genes Involved in Core Lipopolysaccharide Biosynthesis† ´ ,1 NURIA CLIMENT,1 SANTIAGO FERRER,1 JOAN FRANCESC GUASCH,1 NURIA PIQUE ´ 1* SUSANA MERINO,2 XAVIER RUBIRES,2 JUAN M. TOMAS,2 AND MIGUEL REGUE Department of Microbiology and Parasitology, Health Sciences Division, Faculty of Pharmacy, University of Barcelona, 08028 Barcelona,1 and Departamento de Microbiologı´a, Facultad de Biologı´a, Universidad de Barcelona, 08071 Barcelona,2 Spain Received 3 April 1996/Accepted 16 July 1996

Bacteriocin 28b from Serratia marcescens binds to Escherichia coli outer membrane proteins OmpA and OmpF and to lipopolysaccharide (LPS) core (J. Enfedaque, S. Ferrer, J. F. Guasch, J. Toma´s, and M. Reque´, Can. J. Microbiol. 42:19–26, 1996). A cosmid-based genomic library of S. marcescens was introduced into E. coli NM554, and clones were screened for bacteriocin 28b resistance phenotype. One clone conferring resistance to bacteriocin 28b and showing an altered LPS core mobility in polyacrylamide gel electrophoresis was found. Southern blot experiments using DNA fragments containing E. coli rfa genes as probes suggested that the recombinant cosmid contained S. marcescens genes involved in LPS core biosynthesis. Subcloning, isolation of subclones and Tn5tac1 insertion mutants, and sequencing allowed identification of two apparently cotranscribed genes. The deduced amino acid sequence from the upstream gene showed 80% amino acid identity to the KdtA protein from E. coli, suggesting that this gene codes for the 3-deoxy-manno-octulosonic acid transferase of S. marcescens. The downstream gene (kdtX) codes for a protein showing 20% amino acid identity to the Haemophilus influenzae kdtB gene product. The S. marcescens KdtX protein is unrelated to the KdtB protein of E. coli K-12. Expression of the kdtX gene from S. marcescens in E. coli confers resistance to bacteriocin 28b. chromosome (the rfa gene cluster). This cluster contains 17 genes in E. coli K-12 organized into three transcriptional units; a similar organization is found in S. typhimurium, and most of these genes are strongly conserved between the two species

Studies on characterization of genes involved in lipopolysaccharide (LPS) core biosynthesis have been limited to Escherichia coli and Salmonella typhimurium. In both cases, it has been found that those genes are clustered in a region of the

TABLE 1. Bacterial strains, cosmids, and plasmids used Strain, cosmid, or plasmid

Strains S. marcescens N28b E. coli NM554 DH5a XL1blue Cosmids Supercos 1 CosFGR2 Plasmids pBluescript SK pSKF41 pBA1767 pJK2252 pCP2089 a

Relevant characteristicsa

Reference or source

Wild type

10

recA13 araD139 D(ara-leu)7696 D(lac)174 galU galK hsdR2 rpsL mcrA mcrB recA1 D(lacZYA-argF)U169 F80dlacZDM15 endA1 hsdR17 supE44 thi-1 gyrA96 relA1 recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F9 proAB lacIqZDM15 Tn10]

Stratagene 13 Stratagene

Apr Kmr cloning vector Apr Kmr Supercos 1 containing 35-kb chromosomal S. marcescens N28b Sau3A insert

Stratagene This work

Apr ori of ColE1 Apr pBluescript SK containing 5.5-kb BamHI fragment from S. marcescens N28b including the kdtA gene pGEM4 carrying 5.5-kb rfaFCLKZ1 BglII fragment from E. coli K-12 pGEM3 carrying 3.4-kb rfaQG1 DkdtA BglII-SalI fragment from E. coli K-12 pGEM5Z carrying 5.0-kb rfaPSB1 NcoI fragment from E. coli K-12

Stratagene This work 14 21 21

Apr, ampicillin resistant; Kmr, kanamycin resistant.

(for a review, see reference 28). In contrast to other enteric bacteria, Serratia marcescens and Proteus spp. are resistant to polymyxins B and E2 and to the polycationic antimicrobial proteins found in granulocyte granules (31). This characteristic has been related to differences in the LPS core composition between these two bacteria and E. coli K-12 and S. typhimurium (24), suggesting that there could be important differences among the genes involved in LPS core biosynthesis among different enterobacteria. To detect E. coli clones con-

* Corresponding author. Mailing address: Department of Microbiology and Parasitology, Health Sciences Division, Faculty of Pharmacy, University of Barcelona, 08028 Barcelona, Spain. Phone: 34-34024496. Fax: 34-3-4021886. Electronic mail address: regue @farmacia.far.ub.es. † This work is dedicated to the memory of Henry C. Wu, who died on 12 February 1996, with deep sorrow, respect, and gratitude. 5741

5742

GUASCH ET AL.

J. BACTERIOL. TABLE 2. Oligonucleotide primers used for DNA sequencing Primer

FIG. 1. Gel analysis of LPS on a Tricine buffer system from E. coli NM554 (lanes 1 to 4) and E. coli DH5a (lanes 5 to 8) without cosmid (lanes 1 and 5), harboring Supercos 1 (lanes 2 and 6) or CosFGR2 (lanes 3 and 7), and cured of CosFGR2 (lanes 4 and 8).

taining such S. marcescens genes, we decided to use S. marcescens bacteriocin 28b as a screening agent. We have cloned and sequenced the structural gene (bss) of bacteriocin 28b, which kills E. coli K-12. The predicted amino acid sequence of the C-terminal part of this bacteriocin has been shown to have a high degree of similarity to the Cterminal domains of pore-forming colicins (30). Similar bacteriocins are produced by most S. marcescens biotypes (11). The bacteriocin 28b system has several unique features: the structural gene appears to be located on the chromosome (11), the SOS induction of bss transcription requires a transcription activator (8), and no colicin immunity-like gene was found downstream of the structural gene (29). The apparent absence of an immunity gene suggested that bacteriocin-producing S. marcescens could avoid killing by lacking appropriate bacteriocin receptors (29). We have recently reported that E. coli LPS core and outer membrane proteins OmpA and OmpF are involved in bacteriocin 28b binding (6). Since E. coli overproducing the S. marcescens Omp4 protein showed a decrease in outer membrane proteins OmpA, OmpF, and/or OmpC, leading to partial resistance to bacteriocin 28b (12), it was expected that screening for bacteriocin 28b resistance could be used to isolate E. coli clones expressing S. marcescens genes involved in LPS core biosynthesis. In this work, we report the cloning of S. marcescens rfa-like genes and the characterization of the kdtAX region.

FIG. 2. Bacteriocin 28b sensitivity assay. Cells grown on LB were adjusted to an A600 of 0.8. Cells (10 ml) were mixed with different bacteriocin dilutions (10 ml) in a microtiter plate for 15 min at 378C. LB (200 ml) was added, and the cultures were incubated for 2 h at 378C. The percentage of survival was estimated from the A600 ratios of bacteriocin-treated and bacteriocin-untreated cultures for each strain. h, NM554 harboring vector plasmid pSK; É, NM554 harboring pSKF41. AU, arbitrary units.

Oligonucleotide

I.....................................................59-CTCCCGAGATCTGATCAA-39 O...................................................59-CCCCTACTTGTGTATAAG-39 F412..............................................59-TTTGCACCACGCCTCTGA-39 F413..............................................59-GACGGCGATCGTTTTATC-39 S41 ................................................59-ACCCAGGTGGTGATCGGC-39 S131 ..............................................59-AGCGCAGGCAGTCCGGCA-39 F411..............................................59-CGATATGCTGCACCTGAC-39 S4M1 ............................................59-GCGGCGGCTTCGTCGCCC-39 S4M2 ............................................59-ACCTTCAACTTTAAAGAC-39

MATERIALS AND METHODS Bacterial strains, plasmids, and growth conditions. Bacterial strains and plasmids used in this study are listed in Table 1. Bacteria were grown in LB Miller broth and LB Miller agar (18) supplemented with ampicillin (50 mg/ml), chloramphenicol (50 mg/ml), or kanamycin (30 mg/ml) when needed. General DNA methods. DNA manipulations were carried out essentially as previously described (26). DNA restriction endonucleases, T4 DNA ligase, E. coli DNA polymerase (Klenow fragment), and alkaline phosphatase were used as recommended by the suppliers. Recombinant clones were selected on LB Miller agar plates containing the appropriate antibiotics. Construction of an S. marcescens N28b genomic library. S. marcescens N28b genomic DNA was isolated and partially digested with Sau3A as described by Sambrook et al. (26). Supercos 1 was first digested with XbaI, dephosphorylated, and digested with BamHI and then ligated to Sau3A genomic DNA fragments. DNA packaging by using Gigapack Gold III (Stratagene) and infection of E. coli NM554 were carried out as instructed by the manufacturer. Bacteriocin 28b production and sensitivity assay. Bacteriocin 28b was prepared as previously described (30). The overlay test was used for qualitative bacteriocin sensitivity assays (23). Quantitative bacteriocin sensitivity assays were performed as described by Enfedaque et al. (6). Gel analysis of LPS. Cultures for analysis of LPS were grown in LB broth at 378C. LPS was isolated by sodium dodecyl sulfate (SDS)-proteinase K digestion from an outer membrane fraction obtained by differential centrifugation (2) and then was separated by polyacrylamide gel electrophoresis using an SDS-tricine buffer system and visualized by silver staining as previously described (22). Southern blot hybridization. DNA fragments containing E. coli rfa genes were labelled with digoxigenin as described by the manufacturer (Boehringer Mannheim). BamHI-digested plasmid pSKF41 was electrophoresed, denatured, and transferred to a Hybond B membrane. After baking, the membrane was prehybridized and hybridized in 53 SSC (13 SSC is 0.15 M NaCl plus 0.015 M sodium citrate)–0.5% blocking reagent (Boehringer Mannheim)–0.1% Sarkosyl–0.02% SDS. Washing, antibody incubation, and signal detection with p-nitroblue tetrazolium chloride and 5-bromo-4-chloro-3-indolylphosphate (Boehringer Mannheim) were done as described by the manufacturer. Isolation of plasmids carrying Tn5tac1 insertions. E. coli XL1blue carrying recombinant plasmid pSKF41 was mutagenized by using l::Tn5tac1 (4). After 1 h of phenotypic expression, kanamycin (30 mg/ml) was added to the cultures. After overnight growth, plasmid DNA was isolated and retransformed into E. coli XL1blue. Kanamycin-resistant transformants were assayed for bacteriocin-insensitive phenotype, and the points of Tn5tac1 insertions were determined by restriction mapping. DNA sequencing. Primers used for DNA sequencing were purchased from Pharmacia. Double-stranded DNA sequencing was performed with 59-[a35 S]deoxyadenosine thiotriphosphate (NEN-Dupont), by using the Sanger dideoxy-chain termination method (27), according to instructions included in the T7 DNA sequencing kit (Pharmacia LKB Biotechnology). Compressions were resolved by using deaza T7 sequencing mixes (Pharmacia LKB Biotechnology). DNA and protein sequence analysis. The DNA sequence was translated in all six frames, and all open reading frames (ORFs) greater than 100 bp were inspected. Deduced amino acid sequences were compared with those of DNA translated in all six frames from the nonredundant GenBank version 76 and EMBL version 34 databases by using the BLAST network service at the National Center of Biotechnology Information (1). Multiple sequence alignments were done by using the PileUp program from the Genetics Computer Group (Madison, Wis.) package on a VAX 4300. Hydropathy profiles were calculated as described by Kyte and Doolittle (15). Phylogenetic studies were performed by using the parsimony method from the PHYLIP package (7). Nucleotide sequence accession number. The nucleotide sequence data reported in this paper have been submitted to GenBank and assigned accession number U52844.

RESULTS AND DISCUSSION Cloning of an S. marcescens genomic region determining bacteriocin 28b resistance in E. coli. A cosmid-based genomic

VOL. 178, 1996

S. MARCESCENS kdtA AND kdtB GENES

5743

FIG. 3. Restriction map of pSKF41, including adjacent regions in pSK (striped box). The positions of Tn5tac1 insertions used for DNA sequencing, and the corresponding bacteriocin 28b sensitivity phenotype (R, resistant; S, sensitive), are also shown.

library of S. marcescens N28b chromosomal DNA was constructed and introduced into E. coli NM554. Kanamycin (30 mg/ml)-resistant clones were screened for bacteriocin 28b resistance by the agar overlay technique. Gel analysis of LPS from bacteriocin-resistant clones was performed as described in Materials and Methods. We found one recombinant cosmid, CosFGR2, whose LPS core mobility (Fig. 1, lane 3) was altered compared with that of E. coli NM554 harboring vector Super cos1 (Fig. 1, lane 2). CosFGR2 conferred to E. coli NM554 a high level of bacteriocin 28b resistance when the level was measured quantitatively (data not shown). CosFGR2 was cured from E. coli NM554(CosFGR2) by serial growth in LB Miller broth without antibiotics and single-colony isolation on LB Miller agar. Colonies were tested for loss of kanamycin resistance and absence of CosFGR2. The cured strain showed an LPS core mobility (Fig. 1, lane 4) identical to that of the parent strain (Fig. 1, lane 1) and became bacteriocin sensitive. Similar results were found with E. coli DH5a as the host strain (Fig. 1, lanes 5 to 8). These results suggested that enzymatic activities encoded by CosFGR2 altered E. coli NM554 LPS core composition and thus its mobility on the SDS-tricine buffer system. A previous study has shown that E. coli LPS core is involved in bacteriocin 28b binding (6), suggesting that Cos FGR2 encoded genes from S. marcescens responsible for the bacteriocin 28b resistance phenotype and that at least some of these genes could be involved in S. marcescens LPS core biosynthesis. CosFGR2 contains DNA homologous to the rfa region of E. coli. CosFGR2 was digested with BamHI, and the DNA fragments were subcloned in pSK. Among these subclones, plasmid pSKF41 was found to confer bacteriocin 28b resistance (Fig. 2). In E. coli and S. typhimurium, genes involved in LPS core biosynthesis are clustered in one region of the chromosomal DNA and are organized in three operons (25, 28); furthermore, there is homology among most of the E. coli and S. typhimurium rfa genes (28). Plasmids pBA1767, pJK2252, and pCP2089, containing rfa genes from E. coli (14, 21), were digested with SalI-EcoRI, SalI-EcoRI, and SalI-SacI, respec-

FIG. 4. DNA nucleotide sequence of the kdtA-kdtX region. The deduced kdtA, kdtX, and truncated ORF3 gene products are indicated below the nucleotide sequence. Nucleotides defining a putative ribosomal binding site are shown in boldface.

FIG. 5. Amino acid alignment of deduced kadtA and kdtX gene products and similar proteins. (A) Amino acid alignment of the deduced S. marcescens kdtA gene product (Kdta-Sm) and KdtA proteins from E. coli (Kdta-Ec) (5), H. influenzae (Kdta-Hi) (9), C. trachomatis (Kdta-Ct) (3), C. psittaci (KdtA-Cps) (17), and C. pneumoniae (Kdta-Cpn) (16). (B) Amino acid alignment of the deduced S. marcescens kdtX gene product (Kdtx-Sm), KdtB protein from H. influenzae (Kdtb-Hi) (9), and putative glycosyltransferase from Y. enterocolitica (Trsb-Ye) (GenBank accession number X63827).

5744

S. MARCESCENS kdtA AND kdtB GENES

VOL. 178, 1996

5745

FIG. 5—Continued.

tively. From these digestions, DNA fragments of 3.4, 5.5, and 5.0 kb containing rfaQ and truncated kdtA, rfaZKLCF, and rfaGPSBI genes, respectively, were purified. These fragments were digoxigenin labelled and used as probes in Southern blot hybridization experiments with BamHI-digested pSKF41. The hybridization results suggested that genes homologous to rfaQkdtA region of E. coli genes are located in plasmid pSKF41 (data not shown). Sequencing of the DNA conferring bacteriocin resistance. To confirm that the isolated S. marcescens genes were involved in LPS core biosynthesis, recombinant plasmid pSKF41 was mutagenized by transposon Tn5tac1 insertion. Each insertion mutant was tested for bacteriocin 28b resistance, and the points of Tn5tac1 insertion were determined (Fig. 3). A nucleotide sequence of 2,433 bp was determined in both directions by using oligonucleotides I and O (Table 2) complementary to the ends of Tn5tac1 and the collection of pSKF41::Tn5tac1 insertions (Fig. 3). Other sequence-derived oligonucleotides were used to complete the nucleotide sequence (Table 2). Analysis of the sequence showed two potential ORFs separated by only the stop codon of the first ORF, suggesting that the two genes are cotranscribed (Fig. 4). The larger ORF, ORF1 (1,275 bp), encoded a protein of 425 amino acid residues with a theoretical molecular mass of 47,655 kDa. The smaller ORF, ORF2 (771 bp), encoded a putative protein of 257 amino acid residues with a predicted molecular mass of 29,240 kDa. The beginning of a third ORF was found divergently transcribed from upstream ORF1. The DNA region between the divergently transcribed ORF1 and ORF3 does not contain an E. coli strong promoter consensus sequence. Analysis of the ORFs’ deduced amino acid sequences. SDSpolyacrylamide gel electrophoresis of plasmid pSKF41 gene

products, by using an E. coli S30 coupled transcription-translation system (Amersham) and [35S]methionine (Amersham), revealed several polypeptides, among them two of about 48 and 24 kDa (data not shown), similar in size to the expected ORF1 and ORF2 products. Computer database searching showed similarities among the deduced 425-amino-acid ORF1 and the 3-deoxy-manno-octulosonic acid transferases (KdtA proteins) from E. coli (5), Haemophilus influenzae (9), Chlamydia trachomatis (3), Chlamydia psittaci (17), and Chlamydia pneumoniae (16). Amino acid alignment of these six proteins (Fig. 5A) shows 80% amino acid identity between the KdtA proteins from E. coli and S. marcescens, while a lower level of similarity was found among the enterobacterial KdtAs and those of H. influenzae and Chlamydia spp. The two enterobacterial KdtA proteins were nearly identical in estimated molecular mass and isoelectric point. Furthermore, the amount of 3-deoxy-manno-octulosonic acid per milligram of LPS was 40% higher in LPS obtained from E. coli NM554 (CosFGR2) than in LPS obtained from E. coli NM554 (Supercos 1) when assayed by previously described methods (19, 32). These results strongly suggested that ORF1 correspond to the kdtA gene from S. marcescens. A phylogenetic analysis of the KdtA proteins was performed by a protein sequence parsimony method (Fig. 6). Two clearly defined clusters, formed by the E. coli and S. marcescens KdtA proteins and by the C. psittaci and C. pneumoniae KdtA proteins, were found. The KdtA from H. influenzae appears to be evolutionarily more closely related to the C. psittaci-C. pneumoniae pair. On the other hand, the KdtA from C. trachomatis appears to be evolutionarily more related to that of H. influenzae than to those of the other two Chlamydia species.

5746

GUASCH ET AL.

J. BACTERIOL.

4. 5. 6.

7. FIG. 6. Estimated phylogeny of KdtAs proteins by analysis of the deduced amino acid sequences by the parsimony method. This unrooted tree was found to be the most parsimonious one after 10 iterations of the algorithm PROTPARS from the PHYLIP package (7). E.c., E. coli; S.m., S. marcescens; H.i., H. influenzae; C.t., C. trachomatis; C.pn., C. pneumoniae; C.ps., C. psittaci.

The deduced 257-amino-acid ORF2 (kdtX) protein showed 20% amino acid identity and 48% amino acid similarity to the protein encoded by the kdtB gene (9), of unknown function, found downstream the kdtA gene in H. influenzae (Fig. 5B). This protein has similar levels of amino acid identity and similarity to glycosyltransferase (TsrB) from Yersinia enterocolitica (GenBank accession number X63827). No similarity was found between the deduced 257-amino-acid protein and the 18-kDa protein encoded by a gene found downstream and cotranscribed with kdtA in E. coli (kdtB) (5, 25). Since kdtA and kdtX appear to be cotranscribed, results regarding the bacteriocin resistance phenotype associated with plasmid pSKF41::Tn5tac1 insertions (Fig. 3) suggest that the kdtX gene product modifies the E. coli LPS core to confer bacteriocin 28b resistance. The N-terminal 84 amino acids encoded by the beginning of ORF3 showed 34 and 32% amino acid identity to the RfaQ protein from E. coli (20) and to the 1,5-heptosyltransferase I (RfaC) from Neisseria meningitidis (GenBank accession number U33454). It has been speculated that RfaQ is involved in addition of partial substituents to the LPS core in E. coli (28), suggesting that the ORF3-encoded protein could play a similar role in S. marcescens. Our results suggest that expression in E. coli NM554 of S. marcescens genes putatively associated with LPS core biosynthesis modifies E. coli LPS core, accounting for altered LPS core mobility and conferring bacteriocin 28b resistance. Furthermore, it appears that the organization of the kdtA region in S. marcescens differs from that found in E. coli K-12, mainly by the presence of a kdtX gene coding for a protein similar to the KdtB of H. influenzae, instead of an E. coli-like KdtB protein, downstream the kdtA gene. Further work will be directed to elucidation of the functions of the reported genes in S. marcescens LPS core biosynthesis and searching for other rfa genes upstream of the kdtA region. ACKNOWLEDGMENTS This work was supported by grant PB 94-0906 from DGCICYT (Spain) and grant 2122 from the Generalitat de Catalunya. N.P. was supported by an FI fellowship from the Generalitat de Catalunya. X.R. was supported by an FPI fellowship from the Ministerio de Educacio ´n y Ciencia (Spain). REFERENCES 1. Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403–410. 2. Austin, E. A., J. F. Graves, L. A. Hite, C. T. Parker, and C. L. Schnaitman. 1990. Genetic analysis of lipopolysaccharide core biosynthesis by Escherichia coli K-12: insertion mutagenesis of the rfa locus. J. Bacteriol. 172:5312–5325. 3. Belunis, C. J., K. E. Mdluli, C. R. H. Raetz, and F. E. Nano. 1992. A novel 3-deoxy-D-manno-octulosonic acid transferase from Chlamydia trachomatis

8.

9.

10. 11. 12.

13. 14. 15. 16.

17.

18. 19.

20.

21.

22.

23.

24.

25.

26.

27. 28.

required for expression of the genus-specific epitope. J. Biol. Chem. 267: 18702–18707. Chow, W. Y., and D. E. Berg. 1988. Tn5tac1, a derivative of Tn5 that generates conditional mutations. Proc. Natl. Acad. Sci. USA 85:6468–6472. Clementz, T., and C. R. H. Raetz. 1991. A gene coding for 3-deoxy-D-mannooctulosonic-acid transferase in Escherichia coli. Identification, mapping, cloning, and sequencing. J. Biol. Chem. 266:9687–9696. Enfedaque, J., S. Ferrer, J. F. Guasch, J. Toma ´s, and M. Regue´. 1996. Bacteriocin 28b from Serratia marcescens N28b: identification of Escherichia coli surface components involved in bacteriocin binding and translocation. Can. J. Microbiol. 42:19–26. Felsenstein, J. 1989. PHYLIP: phylogeny inference package (version 3.2). Cladistics 5:164–166. Ferrer, S., M. B. Viejo, J. F. Guasch, J. Enfedaque, and M. Regue´. 1996. Genetic evidence for an activator required for induction of colicin-like bacteriocin 28b production in Serratia marcescens by DNA-damaging agents. J. Bacteriol. 178:951–960. Fleischmann, R. D., M. D. Adams, O. White, R. A. Clayton, E. F. Kirkness, A. R. Kerlavage, C. J. Bult, J.-F. Tomb, B. A. Dougherty, J. M. Merrick, K. MacKenney, G. Sutton, W. FitzHugh, C. Fields, J. D. Gocayne, J. Scott, R. Shirley, L.-I. Liu, A. Glodek, J. M. Kelley, J. F. Weidman, C. A. Philips, T. Spriggs, E. Hedblom, M. D. Cotton, T. R. Utterback, M. C. Hanna, D. T. Nguyen, D. M. Saudek, R. C. Brandon, L. D. Fine, J. L. Fritchman, J. L. Fuhrmann, N. S. M. Geoghagen, C. L. Gnehm, L. A. McDonald, K. V. Small, C. M. Fraser, H. O. Smith, and J. C. Venter. 1995. Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science 269:496– 512. Gargallo-Viola, D. V. 1989. Enzyme polymorphism, prodigiosin production, and plasmid fingerprints in clinical and naturally occurring isolates of Serratia marcescens. J. Clin. Microbiol. 27:860–868. Guasch, J. F., J. Enfedaque, S. Ferrer, D. Gargallo, and M. Regue´. 1995. Bacteriocin 28b, a chromosomally encoded bacteriocin produced by most Serratia marcescens biotypes. Res. Microbiol. 146:477–483. Guasch, J. F., S. Ferrer, J. Enfedaque, M. B. Viejo, and M. Regue´. 1995. A 17 kDa outer-membrane protein (Omp4) from Serratia marcescens confers partial resistance to bacteriocin 28b when expressed in Escherichia coli. Microbiology 141:2535–2542. Hanahan, D. 1983. Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166:557–580. Klena, J. D., E. Pradel, and C. A. Schnaitman. 1992. Comparison of lipopolysaccharide biosynthesis genes rfaK, rfaI, rfaY, and rfaZ of Escherichia coli K-12 and Salmonella typhimurium. J. Bacteriol. 174:4746–4752. Kyte, J., and R. F. Doolittle. 1982. A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157:105–132. Lo¨bau, S., U. Mamat, W. Brabetz, and H. Brade. 1995. Molecular cloning, sequence analysis, and functional characterization of the lipopolysaccharide biosynthetic gene kdtA encoding 3-deoxy-a-D-manno-octulosonic acid transferase of Chlamydia pneumoniae strain TW-183. Mol. Microbiol. 18:391–399. Mamat, U., M. Baumann, G. Schmidt, and H. Brade. 1993. The genusspecific lipopolysaccharide epitope of Chlamydia is assembled in C. psittaci and C. trachomatis by glycosyltransferases of low homology. Mol. Microbiol. 10:935–941. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Osborn, M. J. 1963. Studies of the Gram-negative cell wall. I. Evidence for the role of 2-keto-3-deoxyoctanoate in lipolysaccharide of Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 50:499–506. Parker, C. T., E. Pradel, and C. A. Schnaitman. 1992. Identification and sequences of the lipopolysaccharide core biosynthetic genes rfaQ, rfaP, and rfaG of Escherichia coli K-12. J. Bacteriol. 174:930–934. Pradel, E., C. T. Parker, and C. A. Schnaitman. 1992. Structure of the rfaB, rfaI, rfaJ, and rfaS genes of Escherichia coli K-12 and their roles in assembly of the lipopolysaccharide core. J. Bacteriol. 174:4736–4745. Pradel, E., and C. A. Schnaitman. 1991. Effect of the rfaH (sfrB) and temperature on the expression of rfa genes of Escherichia coli K-12. J. Bacteriol. 173:6428–6431. Pugsley, A. P., and B. Oudega. 1987. Methods for studying colicins and their plasmids, p. 105–161. In K. G. Hardy (ed.), Plasmids, a practical approach. IRL Press, Oxford. Radziejewska-Lebrecht, J., D. Krajewska-Pietrasik, and H. Mayer. 1990. Terminal and chain-linked residues of D-galacturonic acid: characteristic constituents of the R-core regions of Proteae and of Serratia marcescens. Syst. Appl. Microbiol. 13:214–219. Roncero, C., and M. J. Casadaban. 1992. Genetic analysis of the genes involved in synthesis of the lipopolysaccharide core in Escherichia coli K-12: three operons in the rfa locus. J. Bacteriol. 174:3250–3260. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463–5467. Schnaitman, C. L., and J. D. Klena. 1993. Genetics of lipopolysaccharide

VOL. 178, 1996 biosynthesis in enteric bacteria. Microbiol. Rev. 57:655–682. 29. Viejo, M. B., J. Enfedaque, J. F. Guasch, S. Ferrer, and M. Regue´. 1995. Protection against bacteriocin 28b in Serratia marcescens is apparently not related to the expression of an immunity gene. Can. J. Microbiol. 41:217– 226. 30. Viejo, M. B., S. Ferrer, J. Enfedaque, and M. Regue´. 1992. Cloning and DNA

S. MARCESCENS kdtA AND kdtB GENES

5747

sequence analysis of a bacteriocin gene from Serratia marcescens. J. Gen. Microbiol. 138:1737–1743. 31. Viljanen, P., and M. Vaara. 1984. Susceptibility of gram-negative bacteria to polymyxin B nonapeptide. Antimicrob. Agents Chemother. 25:701–705. 32. Weisbach, A., and J. Hurwitz. 1959. The formation of 2-keto-3-deoxyheptanoic acid in extracts of E. coli B. J. Biol. Chem. 234:705–712.