Note added in proof: After this report was accepted for publication, a paper (27) con- cluded from a different assay that the EST1 protein is required for telomerase activity. Our results show that EST1 is neither an essential catalytic nor primer-binding com- ponent of telomerase. REFERENCES AND NOTES 1. E. H. Blackburn and J. W. Szostak, Annu. Rev. Bio- chem. 53, 163 (1984). 2. G.-L. Yu, J. D. Bradley, L. D. Attardi, E. H. Blackburn, Nature 344, 126 (1990). 3. E. H. Blackburn, Annu. Rev. Biochem. 61, 113 (1992). 4. M. S. Singer and D. E. Gottschling, Science 266,404 (1994). 5. M. J. McEachern and E. H. Blackburn, Nature, in press. 6. C. W. Greider and E. H. Blackburn, Cell 43, 405 (1985); A. M. Zahler and D. M. Prescott, Nucleic Acids Res. 16,6953 (1988); D. Shippen-Lentz and E. H. Blackburn, Mol. Cell. Biol. 9, 2761 (1989). 7. G. B. Morin, Cell 59, 521 (1989); K. R. Prowse, A. A. Avilion, C. W. Greider, Proc. Natl. Acad. Sci. U.S.A. 90, 1493 (1993); L. L. Mantell and C. W. Greider, EMBOJ. 13, 3211 (1994). 8. C. W. Greider and E. H. Blackburn, Nature 337, 331 (1989); G.-L. Yu and E. H. Blackburn, Cell 67, 823 (1991). 9. J. Shampay, J. W. Szostak, E. H. Blackburn, Nature 310, 154 (1984); S. S. Wang and V. A. Zakian, Mol. Cell. Biol. 10, 4415 (1990). 10. R. W. Walmsley, C. S. M. Chan, B.-K. Tye, T. D. Petes, Nature 310,157 (1984); S.-S. Wang and V. A. Zakian, ibid. 345, 456 (1990). 11. M. Cohn and E. H. Blackburn, unpublished data. 12. Whole cell extracts were prepared as follows: Cells (24) harvested in early log phase (optical density at 600 nm= 1) were resuspended in TMG buffer [10 mM tris-HCI (pH 8), 1.2 mM MgCI2, 15% (v/v) glyc- erol, 0.1 mM EDTA, 0.1 mM EGTA, 1.5 mM dithio- threitol, 1 mM pefabloc, 1 (iM pepstatin, 1 iM leu- peptin, 130 iM bestatin, and RNasin (40 units/ml)] and disrupted in a Bead-Beater (Biospec Products). S-100 extracts were prepared by ultracentrifugation at 100,000g for 90 min at 4°C, divided into portions, and stored at -80°C; the protein concentration was 22 mg/ml for S. caste/lli and 17 mg/ml for S. cerevi- sae S-100 extracts. The S-100 supernatants were fractionated on a DEAE-agarose column (Biogel; Bio-Rad) that had been equilibrated with TMG buffer (with 10% glycerol). The column was washed with four volumes of TMG buffer containing 0.5 M sodium acetate for S. caste/lli and 0.6 M sodium acetate for S. cerevisiae. After elution with 1.8 column volumes of 0.65 or 0.75 M sodium acetate in TMG buffer for S. caste/lli and S. cerevisiae, respectively, eluted fractions were desalted and concentrated on Micro- con-30 columns (Amicon). Protein concentrations were not measured after purification. It was antici- pated that the same relative amounts were still present in the extracts because they were treated identically. Saccharomyces cerevisiae extracts were diluted -100 times and S. caste/lli extracts -10 times during the Microcon-30 column desalting. For telomerase assays, equal volumes of desalted DEAE fractions (usually 10 or 20 xl) and 2 x reaction buffer were mixed and incubated at either 20° or 30°C for 30 min. Final concentrations for S. caste//lli were 50 mM tris-HCI (pH 8), 1 mM spermidine, 1 mM dithio- threitol, 100 mM potassium glutamate, 50 .LM dCTP, 50 p.M dTTP, 1.9 i.M [a-32P]dGTP (800 Ci/ mmol), and 1 p.M primer (25). Potassium glutamate and dCTP were omitted for S. cerevisiae reactions. Reaction products were resolved on denaturing 10% acrylamide gels containing 7 M urea as de- scribed (14). RNase treatment of extracts was per- formed with RNase A (6.3 pg/ml) at 25°C for 5 min, immediately before starting the reaction. RNase A was inhibited by incubating the extract with RNasin (2300 units/ml) at 25°C for 5 min before addition of RNase A (6.3 ig/ml) and incubation for another 5 min. 13. C. W. Greider and E. H. Blackburn, Cell 51, 887 (1987). 14. M. Lee and E. H. Blackburn, Mol. Cell. Biol. 13, 6586 (1993). 15. D. Shippen-Lentz and E. H. Blackburn, Science 247, 546 (1990); L. A. Harrington and C. W. Greider, Na- ture 353, 451 (1991); J. Lingner, L. L. Hendrick, T. R. Cech, Genes Dev. 8, 1984 (1994). 16. Saccharomyces caste//lli telomerase activity was also primed with 12- to 18-nt S. cerevisiae or Tetrahy- mena telomeric repeats. The S. cerevisiae activity was primed efficiently by oligonucleotides consisting of the Tetrahymena telomeric sequence (T2G4)3, or by permutations of the S. castellii telomeric repeats with T or G, but not C, residues at the 3' end. Non- telomeric oligonucleotides did not prime synthesis by either extract (11). 17. V. Lundblad and J. W. Szostak, Cell 57,633 (1989). 18. V. Lundblad and E. H. Blackburn, ibid. 73, 347 (1993). 19. ___ ibid. 60, 529 (1990). 20. D. Gilley, M. Lee, E. H. Blackburn, Genes Dev., in press. 21. K. Collins and C. W. Greider, ibid. 7, 1364 (1993). 22. Shortened labeled reaction products were not pro- duced by elongation of degraded input primer, as shown by 5' end labeling of the bulk primer (present in excess over telomerase) after it had been subjected to the telomerase reaction; >95% of the bulk primer remained full length. The short labeled products were not formed by initial telom- erase elongation and subsequent degradation of the labeled elongation products by a contaminating nuclease as shown by pulse-chase experiments, chased with either ddGTP or excess unlabeled dGTP, in the presence or absence of RNase. Dur- ing each chase there was no further production of shorter labeled products or breakdown of the products longer than input size. Moreover, the nu- cleolytic cleavage activity coeluted with the telom- erase elongation activity in step elutions from DEAE-agarose, heparin-agarose, and octyl- Sepharose (Pharmacia) columns (11). 23. E. H. Blackburn, Cell 77, 621 (1994). 24. Cell strains were S. caste//lli NRRL Y-12630 and S. cerevisiae YPH399 (MATo, ura3-52, lys2-801, ade2-101, his3 A200, leu2-A1, trp1-A63, ga12+/-). 25. All oligonucleotides were purified to a single species by separation on a denaturing gel. Markers were cre- ated by elongation of primers (TGTGGG)2 and (GGGTGTCT)2 with terminal deoxynucleotidyl trans- ferase (Boehringer Mannheim) and [a-32P]dGTP. The primary products were 13 and 17 nt, respectively. 26. The diploid S. cerevisiae strain TVL115 contains the estl -A3::HIS3 mutation, in which -80% of the EST1 gene is deleted. The heterozygous diploid was sporulated and the tetrads dissected. The haploid deletion strain showed the expected senescence and correct genomic band on Southern (DNA) blot analysis (11) with an EST1 riboprobe (17). The pro- tein concentration of the estl S-100 extract was 16 mg/ml. 27. J. Lin and V. Zakian, Cell 81, 1127 (1995). 28. We thank J. Li for providing S. cerevisiae YPH399; V. Lundblad for S. cerevisiae TVL115; D. Gottschling for S. cerevisiae UCC 3508; C. Gross, D. Gilley, A. Krauskopf, and K. Kirk for critical reading of the manuscript; and A. Bhattacharyya for valuable ad- vice. Supported by NIH grant GM26259 (to E.H.B.), the Wenner-Gren Center Foundation (M.C.), and the Lucille P. Markey Foundation. 30 December 1994; accepted 19 April 1995 Simultaneous Identification of Bacterial Virulence Genes by Negative Selection Michael Hensel, Jacqueline E. Shea, Colin Gleeson, Michael D. Jones, Emma Dalton, David W. Holden* An insertional mutagenesis system that uses transposons carrying unique DNA sequence tags was developed for the isolation of bacterial virulence genes. The tags from a mixed population of bacterial mutants representing the inoculum and bacteria recovered from infected hosts were detected by amplification, radiolabeling, and hybridization analysis. When applied to a murine model of typhoid fever caused by Salmonella typhimurium, mutants with attenuated virulence were revealed by use of tags that were present in the inoculum but not in bacteria recovered from infected mice. This approach resulted in the identification of new virulence genes, some of which are related to, but functionally distinct from, the inv/spa family of S. typhimurium. Several different approaches have been used to exploit transposon mutagenesis for the isolation of bacterial virulence genes, including screens for the loss of specific virulence-associated factors (1), survival within macrophages (2), and penetration of epithelial cells (3). Although these screens have identified many bacterial genes re- M. Hensel, J. E. Shea, C. Gleeson, D. W. Holden, Depart- ment of Infectious Diseases and Bacteriology, Royal Postgraduate Medical School, Hammersmith Hospital, Du Cane Road, London W12 ONN, UK. M. D. Jones and E. Dalton, Department of Virology, Royal Postgraduate Medical School, Hammersmith Hospital, Du Cane Road, London W12 ONN, UK. *To whom correspondence should be addressed. quired for virulence, they are restricted to certain stages of infection. Transposon mu- tants have also been tested individually for altered virulence in live animal models of infection (4), but comprehensive screening of bacterial genomes for virulence genes has not been possible because of the inability to identify mutants with attenuated virulence within pools of mutagenized bacteria and the impracticability of separately assessing the virulence of each of the several thou- sand mutants necessary to screen a bacterial genome. We have circumvented this prob- lem by developing a transposon mutagene- sis system, termed signature-tagged mu- tagenesis, in which each transposon mutant SCIENCE · VOL.269 · 21 JULY 1995 400 on February 8, 2011 www.sciencemag.org Downloaded from