Inorganic polyphosphate is essential for long-term survival and virulence factors in Shigella and Salmonella spp. Kwang-Seo Kim*, Narayana N. Rao, Cresson D. Fraley, and Arthur Kornbergt Department of Biochemistry, Stanford University School of Medicine, Stanford, CA 94305-5307 Contributed by Arthur Kornberg, April 8, 2002 The importance of inorganic polyphosphate (poly P) and poly P kinase (PPK), the enzyme principally responsible for its synthesis, has been established previously for stationary-phase survival of Escherichia coli and virulence in Pseudomonas aeruginosa. The gene (ppk) that encodes PPK is highly conserved among many bacterial pathogens, including Shigelfa and Salmonella spp. In view of the phylogenetic similarity of the enteropathogens and the frequency with which virulence factors are expressed in stationary phase, the ppk gene of pathogenic Shigella flexneri, Salmonella enterica serovar Dublin, and Salmonella enterica serovar typhi- murium have been cloned and deleted. In some of these mutants lacking ppk, the phenotypes included features indicative of de- creased virulence such as: (ij) growth defects, (ii) defective re- sponses to stress and starvation, (ii/) loss of viability, (/v) polymyxin sensitivity, (v) intolerance to acid and heat, and (vi) diminished invasiveness in epithelial cells. Thus PPK may prove, as it has with P. aeruginosa, to be an attractive target for antibiotics, with low toxicity because PPK is not found in higher eukaryotes. norganic polyphosphate (poly P) is a chain of tens or many hundreds of phosphate residues linked by “high-energy” phos- phoanhydride bonds. Poly P is ubiquitous, having been found in every cell examined (1), and performs varied functions depending on the cell and circumstances (2). Escherichia coli mutants lacking polyphosphate kinase (PPK), the enzyme responsible for the syn- thesis of poly P from ATP, are deficient in responses to stresses and stringencies and fail to survive in stationary phase (3, 4). The gene ppk that encodes PPK is highly conserved in Gram-negative bac- teria (5), including some 20 pathogens. Mutation of ppk in six enteropathogens rendered them impaired in motility on a semisolid agar surface (6), indicative of a loss in ability to invade and establish systemic infections in host cells. A ppk mutant of Pseudomonas aeruginosa was also defective in quorum sensing and the dependent virulence factors, elastase and rhamnolipid; the mutant was also deficient in biofilm formation and was not lethal in a burned-mouse pathogenesis model (7). Vibrio cholerae ppk mutants also show defects in growth, motility, and surface attachment, features linked to virulence (8). Poly P is involved in the expression in E. coli of RpoS (9), the sigma factor responsible for activation of more than 50 genes required for survival during starvation, UV radiation, oxidative damage, and osmotic stress (10, 11). In addition to a decrease in long-term survival in the stationary phase, increased sensitivities to oxidative, osmotic, and heat stresses, and defects in adaptive growth in minimal media are among the phenotypic features exhibited by the ppk mutant (3); these can all be linked to a decreased expression of the rpoS gene (9). On the basis of all these factors, it would seem that poly P is likely needed for the virulence of Shigella and Salmonella spp. Shigella flexneri, a facultative intracellular pathogen, is the etiological agent of bacillary dysentery. The capacity of this bacterium to enter human epithelial cells depends on secreted proteins encoded by a regulon of virulence genes. Expression of the genes is controlled by multiple environmental signals (12). The ability of S. flexneri in stationary phase to survive for several www.pnas.org/cgi/doi/10.1073/pnas. 112210499 hours at pH 2.5 fikely accounts for the low infective dose in shigellosis. The acid resistance depends on expression of rpoS; a deletion mutant is highly acid sensitive (13). Interdependence with rpoS expression in £. coli (9) can be added to the similarity of S. flexneri in the expressions of several invasion operons (¢.g., viz, ipa, mxi, and spa) that are maximal in stationary-phase cultures (13, 14). Evidence that infective S. flexneri is in a stationary, nondividing state and that expression of stationary- phase-specific genes is essential for survival and virulence pro- vides support for a role for poly P in its pathogenesis. Salmonellae are Gram-negative facultative anaerobes and, when acquired by the ingestion of contaminated food or water, can cause a range of diseases depending on the serovar and host (15-17). Salmonella enterica serovar typhimurium (S. typhi- murium) causes inflammatory diarrhea in calves and also elicits a systemic disease in mice. It shows a broad host range, infecting cattle, pigs, sheep, horses, poultry, and rodents (18); similar diseases caused by Salmonella enterica serovar Dublin (S. dublin) are found principally in cattle. In salmonellae, as in S. flexneri, evidence exists for the regulation of virulence by RpoS. S. typhimurium harbors a large (80-100 kDa) plasmid (19), the absence of which results in attenuation or loss of virulence. The spvABCD (salmonella plasmid virulence) operon found on this plasmid greatly enhances the ability of the bacteria to proliferate in extraintestinal tissues and thus is required for the induction of systemic disease in mice (20). To investigate the role of poly P in the virulence of Shigella and Salmonella spp., null mutants of ppk were prepared and their phenotypes, with particular relation to virulence factors, were examined. Materials and Methods Reagents. ATP, creatine kinase, DNase I, and RNase I[la were from Roche Molecular Biochemicals. Creatine phosphate, 3-(N- morpholino)propanesulfonic acid (Mops), kanamycin, ampicil- lin, tetracycline, amino acids, and BSA were from Sigma. [y-?2P]ATP, Hybond-N™ nylon membranes, and carrier-free **P; were from Amersham Pharmacia. Polyethyleneimine-cellulose F TLC plates were from Merck. Restriction enzymes were from New England Biolabs. The DIG (digoxigenin) DNA labeling kit was from Boehringer Mannheim, and [°*P]poly P was as described (21). Strains, Plasmids, and Phages. Strains, plasmids, and phages are listed in Table 1. Plasmids were introduced into FE. coli by Abbreviations: poly P, inorganic polyphosphate; PPK, polyphosphate kinase; PPX, ex- opolyphosphatase; S$. typhimurium, Salmonella enterica serovar typhimurium; S. dublin, S. enterica serovar Dublin; WT, wild type; cfu, colony-forming unit. Data deposition: The sequence reported in this paper has been deposited in the GenBank database (accession no. AFO85682). *Present address: Department of Plant and Microbial Biology, University of California, Berkeley, CA 94720. *To whom reprint requests should be addressed. E-mail: akornberg@cmgm.stanford.edu. PNAS | May 28,2002 | vol.99 | no. 11 | 7675-7680 MICROBIOLOGY Table 1. Strains, phage, and plasmids Relevant characteristics Source or ref. Strains 5. typhimurium S. typhimurium CF 7519 S. typhimurium WRAY S. typhimurium FIRN SF11687 S$. dublin S$. dublin S. dublin N2 5. flexneri S. flexneri S. flexneri M1 E. coli $L7519 [F°] muth11t::Tnt0 SL7519 \ppk Appx:: kan WT Nal, WT Nal‘, S. typhimurium FIRN, ppk::kan SVA47 WT SVA47 Appk Appx::kan WT 2a Appk Appx:: kan B. A. D. Stocker This study B. A. D. Stocker This study This study B. A. D. Stocker This study A. T. Maurelli This study DH5a F’ endA? hsdR17 supE44 thi-1 recAl gyrA96 relAl AlargF-lacZYA) U169 (¢80dlacZAM15) Lab collection $17-1 (A pir) Tp" recA thi pro hsdR~ M* RP4:2-Tc:MuKm Tn7, a pir Lab collection CF5802 K-12 (MG1655) Appk Appx:: kan Lab collection KL16 Hfr CF16 KL16 Appk Appx:: kan This study Phage and Plasmids A DASHII A DASHII vector digested with BamHi Stratagene pBluescript tt KS (+) Ap", ColE1, high-copy-number cloning vector Stratagene pUuCc4k Pharmacia pKNG101 Sm’, oriR6K, mobRK2, sacB, suicide vector 22 pks? pBluescript Il SK (+) derivative harboring 5.7-kb fragment containing ppk and ppx genes of This study 5S. typhimurium FIRN pKS10 pBluescript Il SK (+) derivative harboring 2,072-bp ppk PCR fragment This study pKS10-1 pKS10, Apal, Sa/l, and EcoRI sites in pKS10 were removed This study pKS10-6 pKS10-1 derivative containing ppk coding region interrupted with Km cassette (ppk::km) This study pKSS1 pKNG101 containing ppk:: km cassette This study B. A. D. Stocker, Stanford University School of Medicine; A. T. Maurelli, Uniformed Services University of the Health Sciences, Bethesda, MD. transformation and into S. typhimurium strains by electrotrans- formation with a Bio-Rad Gene Pulser. DNA Manipulations and Analysis. DNA manipulations and analysis were as described by Sambrook ef al. (23). Preparation of Part of ppk Gene of S. typhimurium WRAY. Partial sequence of S. typhimurium ppk was obtained by a BLAST search of the TIGR (The Institute for Genomic Research) genome sequence database by using the E. coli ppk sequence. This sequence was amplified by PCR with two synthetic primers (Sa/I forward primer, 5'-CCGTGAATAAAACGGAGTATAGGTAG-3’; Sail reverse primer, 5’-AAAATGTCATCCAGGCAG-3’); genomic DNA of S. typhimurium was the template. Genomic Library Construction and Screening of S. typhimurium WRAY. Wild-type (WT) DNA was partially digested with Sau3AI. The fragments were ligated into BamHI-digested A DASHII and pack- aged by using Gigapack I] Gold (Stratagene). The library was screened by using the labeled PCR fragment (partial ppk gene, 659 bp) to obtain clones that carry the entire ppk gene. From the positive clones, a 5.7-kb Xbal-KpnI fragment was identified that contained the whole ppk gene. This fragment was cloned into pBluescript II KS (+) (Stratagene) that had been digested with Xbal and Kpni, yielding the plasmid pKS7. The 5.7-kb fragment in pKS7 was sequenced. Construction of ppk-Deletion Mutant of 5. typhimurium. A 2.072-kb PCR fragment containing the ppk coding sequence was generated with pKS7 as the template and the primers 5’-CGTGAATAAA- AACGGAGTAT-3' and 5'-ATGAAAGCTGTTTGAGCCG-3’. This fragment was ligated into pBluescript II KS (+) that had been 7676 | www.pnas.org/cgi/doi/10.1073/pnas. 112210499 digested with Smal to construct the plasmid:pKS10. It was further digested with Apal and EcoRI to remove the Sail site in the vector. The resulting 5.1-kb fragment was treated with Tz, DNA polymerase to create blunt ends, and the ends were then self-ligated. The plasmid pKS10-1 thus prepared was digested with Sa/I to remove a 800-bp fragment from the midportion of the ppk gene to obtain a 4.3-kb fragment. It was ligated to a kanamycin-resistance gene cassette contained within a Sa/I restriction fragment of pUC4K. The resulting plasmid, pKS10-6, was digested with KpnI and Xbal; the fragment was blunt-ended by using Tz DNA polymerase and cloned into pXNG101, yielding pKSS1. The plasmid pKSS1 was introduced into the WT S. typhimurium FIRN from E. coli $17-1 A pir by conjugal transfer. Cointegrant conjugants containing no plasmid and representing a single homologous recombination were isolated by plating on LB agar plates supplemented with nalidixic acid (50 g/ml) and carbenicillin (60 ug/ml), and the genotype was confirmed by PCR. The strains were subjected to sucrose selection; clones with a double-homologous recombination event (streptomycin-sensitive and kanamycin-resistant) were iden- tified and further tested by PCR and assay of PPK. Construction of S. enterica Serovar Dublin Appk Appx::kan Mutants. Plyir lysate of £. coli strain CF5802 was used to transduce the Appk Appx::kan mutation into £. coli Hfr strain KL16. Conjugation was performed between KLi6 Appk Appx::kan (donor) and S. typhi- murium strain SL7519 [F7] mutL111:Tnl0 (recipient). A P22 lysate prepared from the resultant SL7519 Appk Appx::kan strain was then used to transduce the mutation into S. dublin SVA47. Three independent S. dublin Appk Appxvkan mutants were isolated and verified by PCR. Construction of 5. flexneri Appk Appx::kan Mutants. P1,i; lysate of £. coli strain CF5802 was used to transduce the Appk Appx::kan mutation into WT S. flexneri strain 2a. Kim et al. Table 2. PPK, PPX, and poly P in cell lysates PPK, PPX, units/mg protein —_ units/mg protein Poly P, nmol/mg Strain Membrane Soluble Membrane Soluble protein S. flexneri WT 3,520 <50 2,950 <50 21 Mutant 90 <10 <10 <10 <0.1 5S. typhimurium FIRN WT 5,889 488 7,050 200 25 Mutant (SF11687) <60 <10 213 200 <0.1 5S. dublin WT 5,530 <50 4,750 <100 18 Mutant <100 <10 120 <10 <0.2 E. coli* WT 1,800 120 1,610 <40 25 Mutant <50 <10 <30 <10 <0.1 Values are the averages of three independent experiments. *Data calculated from Crooke et al. (4). Biochemical Assays. PPK and exopolyphosphatase (PPX) activities were assayed as described (24, 25). Poly P levels were determined by the nonradioactive method (26). Survival. Survival of stationary-phase cells. See legend to Fig. 3. Heat-shock survival. See legend to Fig. 4. Acid tolerance. Stationary-phase cells grown overnight (~20 h) in LB were washed in saline, resuspended in acidified LB (pH 3.0 for S. flexneri and 3.3 for Salmonella spp.), and incubated aerobically at 37°C; cultures were diluted and plated on LB to measure viability. Polymyxin B resistance. See legend to Fig. 5 for details. Surface attachment. As described (8), 96-well assay plates of polyviny! chloride (Falcon 3911 Microtester II flexible) were from Becton Dickinson. Gentamycin-protection assays. See legend of Fig. 6. Nucleotide Sequence Accession Number. The GenBank accession number for the sequence of S. typhimurium WRAY reported in this study is AF085682. Results PPK, PPX, and poly P in WT and Mutant Pathogens. Levels of PPK, PPX, and poly P in the enteric pathogens S. flexneri, S. typhimurium, and S. dublin were similar to those observed for E. coli (4) (Table 2). PPK of the pathogens was localized, as in E. coli, in the membrane fraction but at 2 to 3 times the level. PPX downstream in the ppk operon resembled E. coli PPK in levels and membrane localization. The null ppk mutants of all these enteric strains exhibited barely detectable levels of PPK, PPX, and poly P (Table 2). Growth in Rich Media. Of the several enterics, only the S. flexneri ppk mutants show a defect in growth and short-term loss of survival (Fig. 1). By contrast, the Salmonella spp. ppk mutants showed relatively little loss of viability even after 2 days of incubation, but as with E. coli (3), suffered profound losses in survival after many days of incubation (see below). In LB medium, S. flexneri grew to an ODsap of 4.and a cell density of 4 x 10° colony-forming units (cfu)/ml. Unlike WT, the ppk mutant, after a normal rate of growth for 3 h, declined in the next 2 h to an ODsqy of 0.3 and a cell density of 8 x 10° cfu/ml. Thereafter, the mutant cultures recovered to near WT levels by about 16 h. The increase in mutant cell density after the abrupt drop coincided with appearance of a smaill-colony phenotype; these variants represented about 60% of the number of typical large WT colonies present in overnight (~16-h) cultures. The ppk mutant of S. flexneri also differed from WT; when the Kim et al. 0 5 10 18 2 2 10! 10° CFU/mI 1 10" Mutant 10° o 5 10 1% 20 23 Time (h) Fig. 1. Growth studies. WT and mutant S. flexneri were grown aerobically at 37°C in LB medium. Samples were analyzed for growth (ODsao) and viability (cfu/ml). cells were grown aerobically in nutrient broth (Difco) the growth rate was significantly slower up to 6 h, and a lower cell density was reached after 16 h of growth. Remarkably, mutants grown in either LB or nutrient broth aggregated and settled out within a few minutes, unlike WT, which remained uniformly dispersed in the growth medium (data not shown). Mutant+PPK §WT+vector N x “ Mutant +vector OD 549 0.01 0 8 16 24 10 wT. ' \ 0.1 N Mutant 0.01 0 2 4 6 8 10 Time (h) Fig. 2. Growth after downshift. (A) 5. typhimurium WT complemented with vector, and mutant with either vector or vector containing ppk gene, were grown aerobically at 37°C in LB for about 20 h. The cultures were then diluted (1:100) into prewarmed (37°C) Mops-buffered minimal medium with 2 mM P; (26) and incu- bated aerobically at 37°C; growth was monitored by ODsaqo. (B) The S. dublin WT and mutant diluted in Mops medium in a similar way were compared. PNAS | May 28, 2002 | vol.99 | no.11 | 7677 MICROBIOLOGY Table 3. Survival of S. flexneri and Salmonella spp. in acidic medium cfu/ml Survival, Strain Before After % S$. flexneri WT 1.98 x 109 2.38 « 105 0.012 Mutant 1.85 x 10? 0.04 « 105 0.0002 S. typhimurium WT 2.7 x 109 0.26 x 109 9.6 Mutant 3.4 x 10° 0.12 x 109 3.5 S. dublin WT 1.4 x 109 0.83 x 109 59.0 Mutant 2.65 x 10° 0.27 x 109 10.0 Stationary-phase cells (20 h in LB medium) of S. flexneri were incubated aerobically at 37°C for 1h in LB medium adjusted to pH 3.0 with HCI. After incubation, cultures were diluted and plated to measure viability. In the case of Salmonella spp., the stationary-phase cells were incubated for 1.5 hin LB medium adjusted to pH 3.3. Values are the averages of three independent experiments. Growth in Minimal Media. PPK mutants of Sa/monella spp. could be adapted to grow in Mops-buffered minimal medium (27), but the ppk mutant of S. flexneri grew poorly. Supplementation of the medium with amino acids restored the growth of the mutant to WT levels (data not shown). Growth After Nutrient Downshift. 8. typhimurium WT grown in LB medium and diluted (1:100) into Mops-buffered minimal medium suffered a lag of about 4 h before the start of exponential growth with a generation time of 76 min. After a similar lag, the ppk mutant grew more slowly, with a generation time of about 120 min (Fig. 24). WT growth was restored when the mutant was complemented with a plasmid containing ppk. S. dublin WT subjected to a similar nutrient downshift grew with no lag and a doubling time of 60 min; the ppk mutant grew slowly to about 3 h before the onset of WT exponential growth (Fig. 28). Effect of pH on Growth. Many bacteria endure transient encounters with very low or high pH, well outside the growth range (13, 14). The Enterobacteriaceae must cope with low-pH stress during their passage through the stomach to the intestine (29, 30). Growth of WT and the ppk mutant were compared in LB medium supple- mented with 0.4% glucose and adjusted to pH values ranging from 3.0 to 8.0. S. flexneri WT cells grew well across a broad pH range of 5 to 8 and even at pH 4.3 after a prolonged lag period; the ppk mutant also grew well from pH 5 to 8 but failed to grow at all at pH 4,3; neither WT nor mutant cells could grow at pH 3.3 (data not shown). The WT and mutant of S. typhimurium did not differ in growth at pH values between 5.0 and 8; the density of the WT was about 2-fold higher after 7 h of growth between pH 5.0 and 7.0. Both WT and mutant did not grow at pH 3.0; growth of the WT was better than the mutant at pH 4.0; WT reached a density of 0.8 after 7 h compared with 0.3 for the mutant. The growth rate of both WT and mutant S. dublin, like S. typhimurium, did not differ at pH values between 5.0 and 8.0; but the density of the WT was about 3-fold higher after 7 h at pH 5.0-7.0. Both WT and mutant did not grow at pH 3.0. Growth of the WT and the mutant at pH 4.0 commenced after a 4-h lag, but no significant difference occurred in the growth curves (data not shown). Virulence Factors. Acid tolerance. Stationary-phase cells grown in rich medium were exposed to pH 3.0 (for 1 h at 37°C). Survival of WT S. flexneri was reduced by 10*-fold, but was still 50-fold better than that of the mutant (Table 3). Unlike S. flexneri, less difference occurred in acid tolerance in Salmonella spp. between WT and 7678 | www.pnas.org/cgi/doi/10.1073/pnas.112210499 A. S. typhimurium 100 10 1 01 Mutant 0.01 0.001 0.0001 4 8 12 16 Survival (%) B. S. dublin 100 10 Mutant 4 8 12 16 Time (days) Fig.3. Long-term survival. WT and mutants of S. typhimurium (A) and S. dublin (8) were tested. Long-term survival in LB was assayed as described (3). mutants. The S. typhimurium mutant showed a 3-fold lesser acid tolerance than WT; the S. dublin mutant showed a 6-fold lower acid tolerance (Table 3). Long-term survival. In LB medium, survival of the S. typhimurium mutant declined to about 0.001% after 8 days of incubation and still further after 10 days, whereas the WT remained at about 6% of the initial value (Fig. 3). At 10 days, the colony size of the mutant was like that of the WT, but in the next 4 days the culture grew back to about 1% of the initial number with two-thirds of the colonies being variants of small size. When stationary-phase cultures of the WT and mutant were present at a 1:1 ratio, the survival of the mutant increased to about 20% of the initial value until 8 days; thereafter, viability of both WT and mutant in the mixed culture dropped drastically to about 0.001% (data not shown). The S. dublin mutant showed only a modest loss of viability to about 3% after 7 days of incubation, whereas the WT remained at about 20% of the initial value. When stationary-phase cultures of the WT and mutant were present at a 1:1 ratio, the viability of the mutant decreased rapidly to about 0.2% of the initial value after 3 days of incubation compared with 20% for the WT. Thereafter, the mutant population decreased another 10-fold after 9 days, whereas the WT remained near the same level (data not shown). Heat resistance. On entry into stationary phase, the £. coli WT develops a tolerance to heat at 55-57°C (3). Stationary-phase WT S. flexneri exposed to 55°C for 3 min retained 70% of their viability compared with less than 0.5% for the mutant (Fig. 44). S. typhi- murium WT survival after 2 min was 30% compared with less than 4% for the mutant (Fig. 4B). The heat resistance of S. dublin was relatively greater than that of S. typhimurium; little loss of viability occurred after 4 min and, even after 10 min, 9% of the WT survived compared with 2% for the mutant (Fig. 4C). Polymyxin B resistance. Survival of mutant S. typhimurium in the stationary phase in polymyxin B at 10 ug/ml was 65% compared with 100% for WT; at 20 pg/ml only 15% of the mutant survived compared with 65% for WT (N.N.R. and A.K., unpublished data). In a repetition of this experiment with a wider range of polymyxin B levels, the resistance of the mutant was again significantly less Kim et al. A. Shigella flexneri Survival 10 (%) 0 2 4 6 C. S. dublin 100 10 4 Mutant J } L 0 2 4 6 8 10 12 1 Time (min) Fig. 4. Heat-shock survival. WT and mutants of Salmonella spp. and S. flexneri were grown overnight (16 h) in LB. The stationary-phase cells were washed and diluted in 0.9% NaCl to a cell density of about 5 = 103 per ml. Samples (2 ml) in glass tubes were prewarmed to 55°C and 0.1-m/ samples were plated directly on LB for viable cell numbers. than that of the WT and could be restored to WT levels by complementation with ppk (Fig. 5). Similar results were obtained with S. dublin (data not shown). 100 10 Mutant+PPK Survival 1 (%) 0.1 7 Mutant+vector 0.01 1 1 1} 0 20 40 60 Polymyxin B (ug/ml) Fig. 5. Polymyxin B resistance. In ppk mutant of S. typhimurium, resistance is restored after complementation with PPK. For induction of PPK, cells were grown to an ODsap of 0.8 in LB to which 1 mM isopropyl 8-p-thiogalactopyranoside (IPTG) was added and incubated further aerobically at 37°C for 3 h. Cells were washed, diluted in sterile 0.85% saline to about 5 x 10? celis per ml, and exposed for 1h at 37°C to various concentrations of polymyxin B. Survivors were measured by plating on LB. Kim et al. Motility and surface attachment. The mutants of Salmonella spp. were impaired in their swimming motility as observed with E. coli and other bacteria (6). WT S. flexneri is nonmotile. With regard to adherence to an abiotic surface, predictive of a capacity to form biofilms, the Sa/monella spp. mutants showed a 20-35% decrease relative to WT in their adherence to polystyrene (data not shown). invasion and growth in epithelial cells and survival in macro- phages. The S. typhimurium mutant was only half as invasive in HEp-2 epithelial cells measured at 2 h and was still at half the WT level after 24 h (Fig. 64), as measured by a gentamycin-protection assay. Survival in the macrophage RAW 264.7 was unaffected for the WT, but the mutant declined progressively when assayed after 4 and 24 h (Fig. 6B). Discussion The purpose of this study was to confirm and extend the hypothesis that poly P and PPK, the enzyme that makes it, are needed for virulence in the important enteric pathogens S. flexneri, S. typhimurium, and S. dublin. The hypothesis was based on several facts: (/) poly P participates in activation of stationary- CFU (109) 5.0 CFU (104) 2.5 2 4 24 2 4 2 Time (h) Fig.6. Invasion and growth in epithelial cells and survival in macrophages. The gentamycin-protection assays were as described (28). (8) WT and mutant of 5. typhimurium were grown overnight in LB to stationary phase. The cells were diluted in PBS and opsonized for 30 min in PBS containing 20% normal mouse serum. They were added to 2.5 x 105 macrophage RAW 264.7 cells (ATCC TIB-71) seeded in 24-well tissue culture plates at a multiplicity of infection of 10:1 and incubated at 37°C for 30 min. Infected monolayers were then treated with gentamycin (100 1g/ml) for 90 min and were lysed or further incubated for 4 and 24 h in the presence of 10 g/ml gentamycin. Infected cells were then washed twice with PBS and lysed with 1% Triton X-100 in PBS and plated on LB for measurement of cfu. (A) HEp-2 (ATCC CCL-23) epithelial cells were maintained in RPMI medium 1640 supplemented with 10% FCS. Gentamycin protection assays were as described above except that WT and mutant S. typhimurium were grown under an invasive condition [overnight in LB static culture, diluted 1:50 in LB- high-salt (0.3 M NaCi) medium followed by static incubation at 37°C for 4.5 h]. Bacteria were inoculated into 8 x 104 HEp-2 cells seeded in 24-well tissue-cuiture plates at a multiplicity of infection of 10:1 for 30 min. PNAS | May 28, 2002 | vol.99 [| no.11 | 7679 MICROBIOLOGY phase responses and starvation in E. coli and so is essential for its survival (2, 3); (i) virulence factors of many pathogens are expressed in stationary phase (1, 4, 31); (ii) the ppk gene is highly conserved among these and many other pathogens (2, 5); and (7v) virulence in mice of P. aeruginosa (7) and virulence factors of Neisseria meningitidis (32), V. cholerae (8), and Helicobacter pylori (C.D.F., C.-M. Tzeng, and A.K., unpublished results) all depend on the intact ppk gene. An additional incentive in this study was to widen the range of comparisons and relationships in the enzymology, metabolism, genetics, and physiology of poly P. Organization of the ppk operon and the levels and membrane location of PPK were similar in the three enteric pathogens and resembled those of E. coli (Table 2). Although knockout of ppk reduced PPK and poly P to levels below detection, it should be noted that another PPK activity (PPK2) has been discovered in ppk mutants of P. aeruginosa that was not detected by the standard assays for PPK; this pathway is likely responsible for significant accumulation of poly P in the ppk mutants of this organism (H. Zhang and A.K., unpublished results). Of the three enteric pathogens, only S. flexneri showed a strong defect in the extent of growth in both rich and minimal media (Fig. 1), which was accompanied by a profound loss in viability in a few hours. By contrast, the Salmonella spp. showed only a modest diminution of growth rate after a nutrient downshift (Fig. 2) and lost viability only after many days (Fig. 3). With the loss of viability, the large WT colony type was succeeded by the emergence of a small-colony variant, much as had been observed with E. coli (3). When mutant and WT Salmonella spp. were cultured together in a 1:1 ratio, both the WT and mutant S. typhimurium lost viability, whereas under similar conditions the S. dublin WT was spared. The basis for these results remains to be studied. The phenotypes of the enteric pathogens include defects in several factors (Table 4), some of which have been related to virulence. These virulence factors include diminished capacity to withstand low pH or elevated temperature or polymyxin B, im- pairment in motility, attachment to an abiotic surface, and inva- siveness in eukaryotic cells in culture. On the basis of these i vitro criteria of virulence, tests of the mutants in an animal host are now clearly indicated. Cloning of the ppk gene of the three enteric pathogens makes their overexpression feasible, as well as making their PPKs available for structural and functional studies. These should complement the comparative studies of the PPKs of E. coli, P. aeruginosa, V. . Kulaev, 1. S. (1979) The Biochemistry of Inorganic Polyphosphates (Wiley, New York). . Kornberg, A., Rao, N.N. & Ault-Riché, D. (1999) Annu. Rev. Biochem, 68, 89-125. . Rao, NN. & Kornberg, A. (1996) J. Bacteriol. 178, 1394-1400, . Crooke, E., Akiyama M., Rao, N. N. & Kornberg, A. (1994) J. Biol. Chem. 269, 6290-6295, . Tzeng, C.-M. & Kornberg, A. (1998) Mol. Microbiol. 29, 381-382. . Rashid, M. H., Rao, N. N. & Kornberg, A. (1999) J. Bacteriol. 182, 225-227. . Rashid, M. H., Rumbaugh, K., Passador, L., Davies, D. G., Hamood, A. N., Iglewski, B. H. & Kornberg, A. (2000) Proc. Natl. -tcad. Sci. USA 97, 9636-9641. 8. Ogawa, N., Tzeng, C.-M., Fraley, C. D. & Kornberg, A. (2000) /. Bacteriol. 182, 6687-6693. 9. Shiba, T., Tsutsumi, K., Yano H., thara, Y., Kameda A., Tanaka K., Takahashi, H., Munckata, M., Rao N. N. & Kornberg, A. (1997) Proc. Natl. Acad. Sci. USA 94, 11210-11215. 10. Hengge-Aronis, R. (1993) Cell 72, 165-168. 11, Loewen, P. C. & Hengge-Aronis, R. (1994) Annu. Rev. Biochem. 48, 53-80. 12, Dorman, C. J., McKenna, S. & Beloin, C. (2001) Int. J. Med. Microbiol. 291, 89-96, 13, Small, P., Blankenhorn, D., Welty, D., Zinser, E. & Slonezewski, J. L. (1994) J. Bacteriol. 176, 1729-1737. 14, Lin, J., Lee, L.S., Frey, J., Slonczewski, J. L. & Foster, J. W. (1995) /. Bacteriol. 177, 4097-4104. wp DAH 7680 | www.pnas.org/cgi/doi/10.1073/pnas.112210499 Table 4. Defects in PPK mutants Defect relative to WT Assay S. flexneri S$. typhimurium $. dublin E. coli Loss of PPK tet +4 tt abe Loss of PPX +44 +44 tee Lack of poly P tot tnt +++ +44 Growth in LB bf - + + Loss of survival Short-term +++ ~ + + Long-term NA +t hot sep Growth at low pH +++ +> +4 ND Resistance to Acid bet _ ++ ND Heat hat ps { tt Polymyxin B resistance ND + + ++ ND Surface attachment ND ~ ++ +4 Epithelial cell invasion ND t+ ND ND Survival in macrophage ND ++ ND ND Motility NA tat tt ++e Defect has been graded from high (+++) to low (+), based on data presented in Results. NA, not applicable; 5. flexneri loses viability in the short term, and the WT is nonmotile. ND, not determined. cholerae, and H. pylori, which have revealed striking differences in their kinetics and specificities (5). PPK is attractive as a target for antibiotics, because the absence of any similar enzyme in higher eukaryotic species makes toxicity less likely. Large-scale screening for inhibitors of E. coli and P. aeruginosa PPKs have produced candidates, unique among known kinases, and active at low concentration (S. Lee, ICOS Corp., Bothell, WA, personal communication). Such compounds may prove useful not only as drugs but also as reagents for studies in which prompt inhibition of PPK can be achieved and for which conditional mutants are not yet available. We thank the undergraduate students Amit Prasad and Ted Su for excellent technical assistance and Drs. B. A. D. Stocker and A. T. Maurelli for some strains and phages. The PAN Facility at Stanford University provided primer synthesis and nucleotide sequencing. We also thank Leroy Bertsch for critical reading of the manuscript. This work was supported by a grant from the National Institute of General Medical Sciences, National Institutes of Health. 15, Lucas, R. L. & Lee, C. A. (2000) Mol. Microbiol. 36, 1024-1033. 16. Schechter, L. M. & Lee, C. A. (2000) Subcell. Biochem. 33, 289-320. 17. Darwin, K. H. & Miller, V, L. (1999) Clin. Microbiol. Rev. 12, 405-428. 18. Kingsley, R. A. & Baumler, A. J. (2000) Mol. Microbiol. 36, 1006-1014. 19. Jones, G. W., Robert, D. K., Svinarich, D. M. & Whitfield, H. J. (1982) Infect. Immun. 38, 476-486. 20. Gulig, P. A., Danbara, H., Guiney, D. G,, Lax, A. J., Norel, F. & Rhen, M. (1993) Mol. Microbiol. 7, 825-830. 21. Wurst, H., Shiba, T. & Kornberg, A. (1995) J. Bacteriol. 177, 898-906. 22. Quandt, J. & Hynes, M. F. (1993) Gene 427, 15-21. 23. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Lab. Press, Plainview, NY), 2nd Ed. 24, Ahn, K. & Kornberg, A. (1990) J. Biol. Chem. 265, 11734-11739. 25. Akiyama, M., Crooke, E. & Kornberg, A. (1993) J. Biol. Chem. 268, 633-639. 26. Ault-Riché, D., Fraley, C. D., Tzeng, C.-M. & Kornberg, A. (1998) J. Bacteriol. 180, 1841-1847. 27. Neidhardt, F. C., Bloch, P. L. & Smith, D. F. (1974) J. Bacteriol. 119, 736-747. 28. Cirillo, D. M., Valdivia, R. H., Monack, D. M. & Falkow, S. (1998) Mol. Microbiol. 30, 175-188. 29, Gianella, R. A., Broitman, S. A. & Zamcheck, N. (1972) Gut 13, 251-256. 30. Gorden, J. & Small, P. L. C. (1993) Infect. Immun. 61, 1623-1630. 31, McLeod, G. I. & Spector, M. P. (1996) J. Bacteriol. 178, 3683-3688. 32. Tinsley, C. R. & Gotschlich, E. C. (1995) Infect. Immun. 63, 1624-1630. Kim et al.