Streptococcal Reporter Gene-Fusion Vector for Identification of in Vivo Expressed Genes

Plasmid 42, 67–72 (1999) Article ID plas.1999.1408, available online at http://www.idealibrary.com on

SHORT COMMUNICATION Streptococcal Reporter Gene-Fusion Vector for Identification of in Vivo Expressed Genes Ali O. Kilic¸,* Mark C. Herzberg,† Maurice W. Meyer,† Xuemei Zhao,‡ and Lin Tao‡ ,1 *Department of Microbiology and Clinical Microbiology, School of Medicine, Karadeniz Technical University, 61080 Trabzon, Turkey; †School of Dentistry, University of Minnesota, 17-164 Moos Tower, 515 Delaware Street SE, Minneapolis, Minnesota 55455; and ‡Department of Oral Biology, School of Dentistry, University of Missouri– Kansas City, 650 E. 25th Street, Kansas City, Missouri 64108 Received December 14, 1998 To study streptococcal genes that are specifically induced in the host during endocarditis, we have developed a novel plasmid for use in in vivo expression technology (IVET). This IVET uses an integration plasmid, pAK36, that carries dual (amy– cat) reporter genes. A gene-fusion strain library was constructed with the plasmid randomly inserted into the chromosome of Streptococcus gordonii V288 by insertion– duplication. The library was inoculated intravenously into a rabbit that had been prepared for experimental endocarditis. Beginning 6 h after the inoculation, the rabbit was given chloramphenicol (Cm) intravenously twice a day to a final serum level of 5 mg/ml and was euthanized 3 days later. The aortic valve vegetations containing Cm R S. gordonii clones were cultured. Colonies were screened in vitro for negative amylase activity and sensitivity to Cm. Forty-eight such colonies showed 13 different insertion patterns when Southern hybridization blots were probed with labeled pAK36. For each of the 13 isolates, the gene fragment proximal to the insertion of the reporter amy– cat was cloned, and its nucleotide sequence was determined. Functions of these genes were inferred by their homology to known genes. Therefore, this novel IVET vector can be useful for identification of in vivo induced genes in S. gordonii and other streptococcal species. © 1999 Academic Press

Viridans streptococci normally reside in the oral cavity where they are nonpathogenic, except for the mutans group, which causes dental caries (Hamada and Slade, 1980). At other anatomic sites, however, viridans streptococci may be pathogenic. For example, viridans streptococci are the most common organisms causing endocarditis (Bayliss et al., 1983; Manford et al., 1992). These bacteria can enter a person’s circulation during dental procedures and cause transient bacteremia. If predisposing factors, such as rheumatic disease or congenital heart diseases (valve defects) exist, these bacteria

can colonize the heart and cause infective endocarditis. It is intriguing that apparently avirulent viridans streptococci in the oral cavity could behave as pathogens once they colonize the heart. Apparently, certain virulence genes of these organisms may be induced in vivo. However, specific conditions that might occur in vivo to modify streptococcal gene expression during endocarditis are undefined. There have been two major approaches for studying infective endocarditis: in vivo animal models (Dewar et al., 1987; Durack and Beeson, 1972; Garrison and Freedman, 1970; Herzberg et al., 1992) and in vitro bioassays (Herzberg et al., 1990; Ramirez-Ronda, 1978). We do not know whether in vivo conditions are adequately simulated in vitro. To detect host-induced bacterial genes encoding potential virulence factors in

1 To whom correspondence should be addressed at present address: Department of Oral Biology, University of Illinois at Chicago, College of Dentistry, 801 South Paulina Street, Chicago, IL 60612-7213. Fax: (312) 996-6044. E-mail: [email protected].

67

0147-619X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

68

SHORT COMMUNICATION

Salmonella, Mahan et al. (1993) have reported a novel approach. In vivo expression technology (IVET) uses an integration plasmid vector carrying two promoterless reporter genes, purA and lacZ, for in vivo and in vitro screenings, respectively. Using this and similar IVET systems, many host-induced genes have been identified in several different bacterial species, including Salmonella typhimurium (Mahan et al., 1993, 1995; Heithoff et al., 1997; Valdivia and Falkow, 1997), Pseudomonas aeruginosa (Wang et al., 1996), Vibrio cholerae (Camilli and Mekalanos, 1995), and Staphylococcus aureus (Lowe et al., 1998). We now report an IVET system that can select in vivo induced (ivi) genes in the gram-positive organism, Streptococcus gordonii. To construct an IVET vector specifically for studying streptococci, two promoterless reporter genes of gram-positive bacterial origin, amy and cat, from pRQ200 (Lane et al., 1991) and pMH109 (Hudson and Curtiss, 1990), respectively, were cloned into the streptococcal integration vector pSF143 (Tao et al., 1992). The cat gene encoded a chloramphenicol acetyl transferase conferring resistance to chloramphenicol. The amy gene encoded a-amylase, which could be monitored by the hydrolysis of starch on agar medium (Lane et al., 1991). In this new plasmid construct, pAK36 (Fig. 1), the cat gene provided in vivo positive selection for inducible promoters in S. gordonii. Positive selections in vivo were obtained by inoculating an animal with transformed S. gordonii and treating the animal with chloramphenicol (Cm). To identify promoters that were active in vivo, expression of the amy gene was evaluated in vitro after the bacteria were isolated from the animal. The tetracycline resistance (Tc R) gene was originated from Streptococcus mutans chromosome (Tobian and Macrina, 1982). This gene is expressed in both Escherichia coli and Streptococcus (Tao et al., 1992). In some streptococcal species, such as Streptococcus pneumoniae, the expression of an antibiotic resistance gene on an integrated plasmid may depend on a promoter in the chromosome (Claverys et al., 1995). The Tc R gene of pAK36, however, is expressed autonomously in S. gordonii (Tobian and Mac-

FIG. 1. Construction of pAK36. First, the promoterless amy gene (BamHI and BclI fragment) from pRQ200 was inserted into the BamHI site of pBluescript SK(1) (Strategene) to form pAK2 (not shown). The promoterless cat gene (EcoRI and BamHI fragment) from pMH109 was inserted into pUC18, digested out with EcoRI and HindIII, and then inserted into pAK2 downstream from amy to form pAK22. The plasmid was cut open between the EcoNI site immediately after the amy stop codon and the SacI site at the beginning of cat. About 120 bp, including the amy stop codon, was deleted by exonuclease III and S1 nuclease digestion, and subsequently the ends were joined by ligation with T4 ligase to form pAK24. The in-frame fusion of amy– cat was confirmed by simultaneous expression of amy and cat in E. coli. Finally, the dual-reporter gene fragment between XbaI and XhoI was inserted into the XbaI and SalI sites of the streptococcal integration plasmid pSF143 to form pAK36.

rina, 1982). This feature was critical for constructing a plasmid-fusion strain library with unbiased representation. Normally, to construct a plasmid-fusion strain library in streptococci, E. coli is used to amplify the plasmid clones in order to achieve a higher yield of transformation (Tao et al. 1993b). However, this procedure could create a bias in clone representation because some plasmid clones may be preferentially amplified in the E. coli host. To circumvent this problem, S. gordonii V288 was transformed directly with the pAK36 –S. gordonii DNA ligation mixture (Tao et al., 1993a). As a result, a reporter genefusion strain library of over 10,000 clones was achieved. Upon in vitro testing of the reporter gene-fusion strain library, about 2% of clones

SHORT COMMUNICATION

FIG. 2. Isolation of S. gordonii ivi genes induced in endocarditis. First, S. gordonii V288 chromosomal DNA was digested with Sau3AI. The DNA fragments were inserted into the BamHI site of pAK36 by ligation. The ligation mixture was used to transform S. gordonii V288 to Tc R by insertion– duplication. Over 10,000 transformant colonies were collected with a sterile cotton swab and suspended in THB supplemented with Tc (10 mg/ml). After incubation for 4 h at 37°C, the culture was stored in multiple aliquots at 270°C in 15% glycerol as the S. gordonii reporter gene-fusion strain library. Then, a midexponential phase of the library in THB with Tc was harvested and resuspended in 2 ml of sterile saline (about 10 9 cells). The cell suspension was injected intravenously through an ear vein into a New Zealand White rabbit that had been prepared for experimental endocarditis by placing an indwell catheter as described previously (Herzberg et al., 1992). Beginning 6 h after the inoculation, the rabbit was given Cm intravenously twice a day to a final serum level of 5 mg/ml. The rabbit was euthanized 3 days later. The aortic valve vegetations were dissected out, immersed in sterile saline, and dispersed with a mortar and pestle to recover viable cells under aseptic conditions. The cells were plated onto THB agar supplemented with Tc and incubated in a candle jar at 37°C for 24 h. Single colonies were isolated and replica plated onto three plates: THB with Tc (master plate), THB with Cm, and THB with 0.5% starch (assay plates). The amylase activity was detected by flooding the starch plate with an iodine solution (0.2% I 2 and 2% KI). The bacterial colonies that displayed negative amylase activity and sensitivity to Cm were isolated from the master plate as ivi clones. Finally, to clone these ivi genes, chromosomal DNA from each unique S. gordonii clone was isolated as previously described (Tao et al., 1993b) and digested with the restriction enzyme XbaI or XbaI plus StyI (Promega). After self-ligation with T4 ligase (Promega), the ligated DNA was used to transform E. coli C600 to Tc R (Chung

69

were Cm R (5 mg/ml; the minimal inhibitory concentration for S. gordonii V288 was 0.5 mg/ml) and amylase positive. Hence, 2 of every 100 clones represented a possible fusion to a constitutively expressed promoter or promoterlike sequence. In contrast, 98% of the clones were sensitive to Cm and negative in amylase, indicating that the reporter genes were not expressed in vitro and thus might be fused to an inducible gene. To identify S. gordonii genes that were inducible in the host, we inoculated the reporter gene-fusion strain library into a rabbit prepared for experimental endocarditis as detailed in Fig. 2. The rabbit was treated with Cm to effect in vivo selection. After in vivo selection of S. gordonii clones, valvular vegetations were dissected free and aseptically homogenized. Single colonies were isolated and replica-plated onto three types of Todd–Hewitt broth (THB) agar plates for in vitro screening. These included THB agar supplemented with Tc (10 mg/ml) as master plates, THB agar supplemented with 5 mg/ml Cm, and THB agar supplemented with 0.5% starch. After incubation at 37°C for 24 h, about 98% of clones were Cm R and amylase positive. This suggested that the in vivo selection through the animal enriched the Cm R clones about 50-fold. The remaining 2% of clones represented by 48 isolates were negative for amylase and sensitive to Cm. In these isolates, the reporter-fused genes were presumably expressed in vivo in the host and therefore were candidates of host-induced clones. In no cases were colonies either amylase positive and Cm S or amylase negative and Cm R. From each of the host-induced clones, DNA was isolated and analyzed by Southern blot (Southern, 1975) with biotinylated pAK36 probe. Among these isolates, multiple identical hybridization patterns were noticed (data not shown), indicating sib-

et al., 1989). Plasmids were isolated from the Tc R colonies with the Qiagen plasmid isolation kit. All animal experiments were done under the aegis of Institutional Animal Care and Use Committee Protocol 9509A00003, University of Minnesota.

2 1

1 1 1

orf3-iviF orf4-iviK

iviA iviJ orf6-iviM

Sugar metabolism Exo-1,4-b-cellobiohydrolase Endo-1,3-b-glucanase Cellobiose phosphotransferase enzyme II Transporter Amino acid transporter ATP-binding transport protein Regulator AraC family of transcription regulator Rgg-like regulator DNA synthesis DNA polymerase III g and t Cell surface structure-associated ADP-L-glycero-D-mannoheptose-6-epimerase Peptide methionine sulfoxide reductase Other functions IS transposase (IS 1167) Serine/threonine kinase Glutathione reductase

Similar protein in the database b

Streptococcus pneumoniae Caenorhabditis elegans Burkholderia cepacia

Salmonella typhimurum S. pneumoniae

B. subtilis

B. subtilis S. gordonii

Arabidopsis thaliana Bacillus subtilis

Cellulomonas fimi Arthrobacter sp. Bacillus stearothermophilius

Organism

M36180 Z38016 U19883

U06472 U41735

X06803

AB001488 M89776

X77502 D83026

L38827 D23668 U07818

Accession No.

b

The number of identical ivi clones among the 48 clones analyzed. The proteins are grouped according to their putative functions; note that in the case of an orf-ivi locus, the homology data are for the orf. c The number of amino acid residues analyzed versus the size of the protein compared. d Percentage identity versus similarity.

a

22

orf2-iviC

12 1

iviD ivil

2 1

2 1 1

iviE iviH orf5-iviL

iviG orf1-iviB

Sib a

Locus

Homology Data of S. gordonii Loci Identified as Induced during Endocarditis

TABLE 1

199/199 47/856 114/449

44/310 77/312

60/422

43/291 61/297

66/533 161/567

55/1090 49/548 64/100

Aa/Protein c

58/68 42/58 34/53

27/50 80/89

64/76

32/51 34/59

50/76 25/39

42/63 49/58 52/64

Iden/Sim d

70 SHORT COMMUNICATION

71

SHORT COMMUNICATION

ling clones. A total of 13 unique S. gordonii clones were isolated. The chromosomal DNA from each unique clone was isolated. The inserted plasmid pAK36 and its flanking streptococcal genes were recovered by restriction enzyme (e.g., XbaI) digestion and self-ligation with T4 ligase and by transforming E. coli to Tc R. To determine the nucleotide sequence of the cloned genes, forward and reverse primers were designed according to the sequence of amy and ori genes, which flanked the cloned streptococcal gene fragment in the plasmid. The sequence of the upstream amy gene primer was 59-AGC GCA AAT AAC AGC GTC AGC AA-39, which was complementary to the 59 end of the amy coding region of Bacillus licheniformis (GenBank Accession No. A17930). The other primer sequence was 59-CAA GAG ATT ACG CGC AGA CC-39 and was derived from the plasmid ori sequence of pACYC184 (X06403). Nucleotide sequences were determined first by the dideoxy nucleotide chain-termination method (Sanger et al., 1977) and later by the DNA sequencing service at the University of Missouri–Kansas City with the Perkin–Elmer ABI Prism 377 Automated DNA Sequencer. However, due to deletions of S. gordonii genes in E. coli, some plasmids lost a portion of the amy gene. Sequence determination for these plasmids was carried out only with the forward primer. Sequences of about 500 bp were obtained in either or both directions from each of the plasmids. The products of their conceptual translation in all six reading frames were used to search the nonredundant sequence database (National Center for Biotechnology Information, National Institutes of Health) by the BLASTX program (Altschul et al., 1990). All of the 13 identified ivi genes were new for S. gordonii. Among them, 7 were homologous to a gene of known function (Table 1). Although the remaining 6 were unknown, each was adjacent to an open reading frame (orf) that had homology with a gene of known function. Several genes found in this study have not been reported among ivi genes of other bacteria. These included a cellobiose-related sugar metabolic system and a bacterial endogenous mu-

tagenesis system—the IS element. Unlike other bacteria, such as Salmonella, the growth of Streptococcus is strictly sugar-dependent. Therefore, the ability to utilize sugars in vivo is critical for the survival of streptococci. In fact, the heart valves of humans (Masuda, 1984) and rabbits (Sarphie, 1985) are rich in glycosaminoglycans. Once degraded, N,N9-diacetylchitobiose, (GlcNAc) 2, a structural analogue of cellobiose, is released. Recently, the cel operon of E. coli has been renamed the chb (N,N9-diacetylchitobiose) operon (Keyhani and Rosemann, 1997) because it not only metabolizes, but also is induced by (GlcNAc) 2. It is of interest that the three newly identified, putative S. gordonii cellobiose metabolic genes (iviE, iviH, and orf5iviL) may be involved in the breakdown of sugar associated with heart valves. To date, more than a hundred ivi genes have been identified in Sa. typhimurium (Heithoff et al., 1997). The 13 S. gordonii ivi genes identified in this study may represent only a small portion of the total ivi genes in this bacterium. To verify their role in virulence, strains of S. gordonii carrying mutations in each of these ivi loci are currently being constructed, and the effect of these mutations on bacterial virulence in endocarditis will be examined using the animal model. The availability of the new IVET system for S. gordonii has expanded the repertoire of genetic tools for identification of in vivo expressed microbial genes. The plasmid pAK36 may also be applicable to the identification of in vivo induced genes from other streptococci and closely related gram-positive bacterial species. ACKNOWLEDGMENTS We thank Dr. R. Yasbin for the gift of plasmid pRQ200, Dr. M. Hudson for the gift of plasmid pMH109, and Dr. G. Dunny for the gift of S. gordonii V288. This study was supported by NIH Grants DE11336, DE05501, and DE00270 and by a Grant-in-Aid (KS-96-GB-56) from the American Heart Association, Kansas Affiliate, Inc.

REFERENCES Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990). Basic local alignment search tool. J. Mol. Biol. 215, 403– 410. Bayliss, R., Clarke, C., Oakley, C. M., Somerville, W.,

72

SHORT COMMUNICATION

Whitfield, A. G. W., and Young, S. E. J. (1983). The microbiology and pathogenesis of infective endocarditis. Br. Heart J. 50, 513–519. Camilli, A., and Mekalanos, J. (1995). Use of recombinase gene fusions to identify Vibrio cholerae genes induced during infection. Mol. Microbiol. 18, 671– 683. Chung, C. T., Niemela, S. L., and Miller, R. H. (1989). One-step preparation of competent Escherichia coli: Transformation and storage of bacterial cells in the same solution. Proc. Natl. Acad. Sci. USA 86, 2172–2175. Claverys, J-P., Dintilhas, A., Pestova, E. V., Martin, B., and Morrison, D. A. (1995). Construction and evaluation of new drug-resistance cassettes for gene disruption mutagenesis in Streptococcus pneumoniae, using an ami test platform. Gene 164, 123–128. Dewar, H. A., Jones, M. R., Griffin, S. G., Oxley, A., and Marriner, J. (1987). A study of experimental endocarditis in pigs. J. Comp. Pathol. 97, 567–574. Durack, D. T., and Beeson, P. B. (1972). Experimental bacterial endocarditis. I. Colonization of a sterile vegetation. Br. J. Exp. Pathol. 53, 44 – 49. Garrison, P. K., and Freedman, L. R. (1970). Experimental endocarditis. I. Staphylococcal endocarditis in rabbits resulting from placement of a polyethylene catheter in the right side of the heart. Yale J. Biol. Med. 42, 394 – 410. Hamada, S., and Slade, H. D. (1980). Biology, immunology, and cariogenicity of Streptococcus mutans. Microbiol. Rev. 44, 331–384. Heithoff, E. M., Conner, C. P., Hanna, P. C., Julio, S. M., Hentschel, U., and Mahan, M. J. (1997). Bacterial infection as assessed by in vivo gene expression. Proc. Natl. Acad. Sci. USA 94, 934 –939. Herzberg, M. C., Gong, K., MacFarlane, G. D., Erickson, P. R., Soberay, A. H., Krebsbach, P. H., Manjula, G., Schilling, K., and Bowen, W. H. (1990). Phenotypic characterization of Streptococcus sanguis virulence factors associated with bacterial endocarditis. Infect. Immun. 58, 515–522. Herzberg, M. C., MacFarlane, G. D., Gong, K., Armstrong, N. N., Witt, A. R., Erickson, P. R., and Meyer, M. W. (1992). The platelet interactivity phenotype of Streptococcus sanguis influences the course of experimental endocarditis. Infect. Immun. 60, 4809 – 4818. Hudson, M. C., and Curtiss, R., III (1990). Regulation of expression of Streptococcus mutans genes important to virulence. Infect. Immun. 58, 464 – 470. Keyhani, N. O., and Rosemann, S. (1997). Wild-type Escherichia coli grows on the chitin disaccharide, N,N9-diacetylchitobiose, by expressing the cel operon. Proc. Natl. Acad. Sci. USA 94, 14367–14371. Lane, M. A., Bayles, K. W., and Yasbin, R. E. (1991). Identification and initial characterization of glucose-repressible promoters of Streptococcus mutans. Gene 100, 225–229.

Lowe, A. M., Beattie, D. T., and Deresiewicz, R. L. (1998). Identification of novel staphylococcal virulence genes by in vivo expression technology. Mol. Microbiol. 27, 967– 976. Mahan, M. J., Slauch, J. M., and Mekalanos, J. J. (1993). Selection of bacterial virulence genes that are specifically induced in host tissues. Science 259, 686 – 688. Mahan, M. J., Tobias, J. W., Slauch, J. M., Hanna, P. C., Collier, R. J., and Mekalanos, J. J. (1995). Antibioticbased selection for bacterial genes that are specifically induced during infection of a host. Proc. Natl. Acad. Sci. USA 92, 669 – 673. Manford, M., Matharu, J., and Farrington, K. (1992). Infective endocarditis in a district general hospital. J. R. Soc. Med. 85, 262–266. Masuda, H. (1984). The preparation of heparan sulfate from the mitral valve of the human heart. Int. J. Biochem. 16, 99 –103. Ramirez-Ronda, C. H. (1978). Adherence of glucan-positive and glucan-negative streptococcal strains to normal and damaged heart valves. J. Clin. Invest. 62, 805– 814. Sanger, F., Nicklen, S., and Coulson, A. R. (1977). DNA sequencing with chain termination inhibitor. Proc. Natl. Acad. Sci. USA 74, 5463–5467. Sarphie, T. G. (1985). Anionic surface properties of aortic and mitral valve endothelium from New Zealand white rabbits. Am. J. Anat. 174, 145–160. Southern, E. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98, 503–517. Tao, L., LeBlanc, D. J., and Ferretti, J. J. (1992). Novel streptococcal-integration shuttle vectors for gene cloning and inactivation. Gene 120, 105–110. Tao, L., Sutcliffe, I. S., Russell, R. R. B., and Ferretti, J. J. (1993a). Cloning and expression of the multiple sugar metabolism (msm) operon of Streptococcus mutans in heterologous streptococcal hosts. Infect. Immun. 61, 1121–1125. Tao, L., Tanzer, J. M., Kuramitsu, H. K., and Das, A. (1993b). Identification of several rod loci and cloning the rodD locus in Streptococcus mutans. Gene 126, 123–128. Tobian, J. A., and Macrina, F. L. (1982). Helper plasmid cloning in Streptococcus sauguis: Cloning of a tetracycline resistance determinant from the Streptococcus mutans chromosome. J. Bacteriol. 152, 215–222. Valdivia, R. H., and Falkow, S. (1997). Fluorescence-based isolation of bacterial genes expressed within host cells. Science 277, 2007–2011. Wang, J., Mushegian, A., Lory, S., and Jin, S. (1996). Large-scale isolation of candidate virulence genes of Pseudomonas aeruginosa by in vivo selection. Proc. Natl. Acad. Sci. USA 93, 10434 –10439. Communicated by F. Macrina