Constitutive Expression of the PhoP Regulon Attenuates S Virulence and Survival within Macrophages almonella SAMUEL I. MILLER ' 2 * AND JOHN J. MEKALANOS1 Department of Microbiology and Molecular Genetics , Harvard Medical School , Boston , Massachusetts 02115,1 and Infectious Disease Unit , Massachusetts General Hospital , Boston , Massachusetts 021142 * Received 1 December 1989/Accepted 15 February 1990 The phoP genetic locus is a two-component regulatory system ( phoP-phoQ ) that controls the expression of genes essential for SalmoneUa typhimurium virulence and survival within macrophages .
Strains with a phoP constitutive mutation ( phenotype PhoPC ) showed up to 10-fold greater expression ofphoP-activated genes ( pag loci ) than did strains with a wild-type phoP locus ( phenotype PhoP + ) .
While the phoP constitutive mutation resulted in increased expression ofpag loci , it also dramatically reduced the expression of other protein species .
Comparison of the protein content of PhoP + and PhoPC strains by two-dimensional protein gel electrophoresis demonstrated that at least 40 separate protein species were changed in expression as a result of this mutation .
The PhoPC S. typhimurium were found to be attenuated for virulence and survival within macrophages .
This finding suggests that a balanced PhoP-PhoQ regulatory response , which allows expression ofphoP-repressed as well as-activated genes , is required for full virulence of S. typhimurium .
We have further shown that small numbers of PhoPc bacteria can be used as .
a live attenuated vaccine to protect against mouse typhoid .
As few as 15 PhoPc bacteria protected mice against challenge with 10 50 % lethal doses of wild-type organisms , suggesting that important protective antigens are regulated by the PhoP-PhoQ virulence regulon .
The phoP regulatory locus is composed of two genes , phoP and phoQ , essential for full virulence , survival within macrophages , and defensin resistance of Salmonella typhi-murium ( 4 , 15 , 17 ; S. I. Miller , W. Pulkkinen , M. Selsted , and J. J. Mekalanos , submitted for publication ) .
The PhoP ( transcriptional activator ) and PhoQ ( sensor/kinase ) proteins have amino-acid similarity to other bacterial regulatory proteins ( two-component regulators ) that control the synthesis of many proteins in response to environmental signals ( 6 , 15 , 17 , 20 ) .
The PhoP and PhoQ gene products are essential for the transcriptional activation of a number of unlinked phoP-activated genes ( pag loci ) , suggesting that this system comprises a regulon ( 6 , 9 , 17 ) .
One pag locus is essential for full virulence and survival within macrophages ( pagC ) , whereas others ( pagA , pagB , and psiD ) , including the phoN gene that encodes a periplasmic acid phosphatase , do not as single mutations alter virulence in the BALB/c mouse model of typhoid fever ( 6 , 17 ) .
S. typhimurium strains with pagC mutations are not as attenuated in virulence as phoP or phoQ mutants , suggesting that other pag loci are essential for full virulence ( 17 ) .
We have previously hypothesized that the PhoQ protein recognizes specific environmental factors in the phagolysosome that lead to activation of the PhoP regulon ( 17 ) .
Previous work by Ames and co-workers , as well as past work in our laboratory , identified starvation for essential nutrients ( phosphate , carbon , sulfur , and nitrogen ) , as well as low pH , as signals for induction ofpag gene expression ( 8 , 17 ) .
Because of the difficulties involved in obtaining specific expression of pag gene products under starvation conditions , have characterized strains with S. typhimurium we a mutation in the phoP locus ( pho-24 [ 9 ] ) that renders these strains constitutive in the expression of loci in rich pag media ( phenotype PhoPC ) .
We here that in addition report to PhoPc an expected global change in protein expression , strains show a surprisingly pronounced defect in mouse virulence and survival within macrophages .
MATERIALS AND METHODS Bacterial strains and genetic methods .
American Type Culture Collection ( ATCC ) strain 14028 , a smooth virulent strain of S. typhimurium , was the parent strain for all virulence studies .
Strain TT13208 was a gift from Nang Zhu and John Roth .
Strain TA2367 was a generous , gift of Gigi Stortz and Bruce Ames ( 9 ) .
Bacteriophage P22HT int was used in transductional crosses to construct strains isogenic except for phoP locus mutations ( 2 ) .
Luria broth was used as rich-medium , and minimal-medium was M9 ( 2 ) .
The chromogenic phosphatase substrate 5-bromo-4-c hloro-3-indolyl phosphate ( XP ) was used to qualitatively access acid and alkaline phosphatase production in solid media .
Derivatives of S. typhimurium ATCC 14028 with the pho-24 mutation were constructed by use of strain TA2367 as a donor of the purB gene in a P22 transductional cross with strain CS003 AphoP ApurB ( 17 ) .
Colonies were then selected for the ability to grow on minimal-medium .
A transductant designated CS022 ( phenotype PhoPc ) that synthesized 1,750 U of acid phosphatase in rich-medium ( a ninefold increase over the wild-type level in rich-medium ; Table 1 and Fig. 1 ) was used in further studies .
Derivatives of strains CS022 and CS023 pho-24 phoN2 zxx : : 6251 Tnl0d-Cam , an acid phosphatase-negative derivative of CS022 , containing pag gene fusions were constructed by bacteriophage P22 transductional crosses , using selection of TnphoA-or Mu dJ-encoded kanamycin resistance .
Strains were checked for the intact pag gene fusion by demonstration of appropriate loss of fusion protein activity on introduction of a phoP105 : : TnlOd or phoP102 : : TnlOd-Cam allele .
Assays of acid phosphatase , alkaline phosphatase , and,3-galactosidase were performed as previously described ( 17 ) and are reported in units as defined by Miller ( 16 ) .
Mouse virulence and vaccination studies .
Bacteria grown overnight in Luria broth were washed and diluted in normal saline .
The wild-type parent strain of CA022 ( ATCC 10428 ) was used for all live vaccine challenge studies .
This strain has a 50 % lethal dose ( LD50 ) for naive adult BALB/c mice of less than 20 organisms when administered by intraperitoneal ( i.p. ) injection and 5 x 104 when administered orally in NaHCO3 .
Mice were purchased from Charles River Breeding Laboratories , Inc. ( Wilmington , Mass. ) and were 5 to 6 weeks of age at initial challenge .
All i.p. inoculations were performed as previously described ( 17 ) .
Oral challenge experiments were performed with bacteria grown in LB broth and concentrated by centrifugation .
The bacteria were resuspended in 0.1 M NaHCO3 , to neutralize stomach acid , and administered as a 0.5-ml bolus to animals under ether anesthesia .
Colony counts were performed to accurately access the number of organisms administered .
All challenge experiments were performed 1 month after i.p. inoculation and 6 weeks after oral challenge .
Challenge inocula were administered by the same route as vaccinations .
The care of all animals was under institutional guidelines as set by the animal care committees at the Massachusetts General Hos-and Harvard Medical pital School .
One-dimensional protein gel electrophoresis was performed by the method of Laemmli ( 11 ) on whole-cell protein extracts of stationary-phase cells grown overnight in Luria broth .
The gels were fixed and stained with Coomassie brilliant blue R250 in 10 % acetic acid-10 % methanol .
Two-dimensional protein gel electro-phoresis was performed by the method of O'Farrell ( 18 ) on the same whole-cell extracts .
Isoelectric focusing using 1.5 % pH 3.5 to 10 ampholines ( LKB Instruments , Baltimore , Md. ) was carried out for 9,600 V * h ( 700 V for 13 h 45 min ) .
The final tube gel pH gradient extended from pH 4.1 to pH 8.1 as measured by a surface pH electrode ( BioRad Laboratories , Richmond , Calif. ) and colored acetylated cytochrome pl markers ( Calbiochem-Behring , La Jolla , Calif. ) run in an adjacent tube .
The slab gels were silver stained ( 14 ) .
Experiments were performed as previously described ( 17 ) by the method of Buchmeiher and Heffron ( 1 ) as modified from the method of Lissner et al. ( 12 ) .
Stationary-phase cells were opsonized for 30 min in normal mouse serum before exposure to the cultured bone marrow-derived macrophages harvested from BALB/c mice .
One hour after infection , gentamicin sulfate ( 8 , ug/ml ) was added to kill extracellular bacteria .
All time points were done in triplicate and repeated on three separate occasions .
Bacterial strains and properties Enzyme Reference or activity ( U ) a source Strain Genotype 10428 Wild type 180 ( A ) ATCC ; 17 TA2367 pho-24 1,925 ( A ) 9 CS003 AphoP ApurB < 10 ( A ) 17 CS022 pho-24 1,750 ( A ) This work CS023 pho-24 phoN2 zxx : : 6251 TnlOd-25 ( A ) This work Cam CS012 pagAl : : Mu dJ 45 ( B ) 17 CS013 pagBl : : Mu dJ 120 ( B ) 17 CS119 pagCl : : TnphoA phoN2 85 ( C ) 17 zxx : : 6251 TnlOd-Cam SC024 pagAl : : Mu dJ pho-24 450 ( B ) This work CS025 pagBl : : Mu dJ pho-24 980 ( B ) This work CS026 pagCl : : TnphoA pho-24 phoN2 385 ( C ) This work zxx : : 6251 TnlOd-Cam CS015 phoP102 : : TnlOd-Cam < 10 ( A ) 17 TT13208 phoP105 : : TnlOd < 10 ( A ) b a A , Acid phosphatase ; B , , B-galactosidase ; C , alkaline phosphatase .
b Gift of Ning Zhu and John Roth .
RESULTS The phoP constitutive allele , pho-24 , results in derepression of pag loci .
Using diethyl sulfate mutagenesis of S. typhimu-rium LT-2 , Ames and co-workers isolated strain TA2367 pho-24 , which contained a phoP locus mutation that resulted in constitutive production of acid phosphatase in rich media ( 9 ) .
This phoP-regulated acid phosphatase is encoded by the phoN gene , a pag locus ( 9 , 17 ) .
We wished to analyze whether the pho-24 allele increased the expression of other pag loci .
This was done by scoring the effect of the pho -- 24 allele on the expression of other pag loci recently identified as transcriptional ( e.g. , pagA and pagB ) and translational ( e.g. , pagC ) fusion proteins that required phoP and phoQ for expression ( 17 ) .
Therefore , we constructed pag gene fusion strains , isogenic except for the pho-24 allele , and assayed fusion protein activity .
PhoPc derivatives of the pagA : : Mu dJ and pagB : : Mu dJ strains produced 480 and 980 U , respectively , of P-galactosidase in rich-medium , an increase of 9-to 10-fold over values for the fusion strains with a wild-type phoP locus ( Table 1 ) .
The pagC : : TnphoA gene fusion produced 350 U of alkaline phosphatase , an increase of three-to fourfold over that produced in strain CS119 , which is isogenic except for the pho-24 mutation ( 17 ) .
These results compare with a ninefold increase in the acid phosphatase activity in strain CS022 on introduction of the pho-24 allele .
Therefore , these available assays for pag gene expression document that the pho-24 mutation causes constitutive expression of other pag loci besides phoN .
Identification of protein species that are repressed as well as activated in the PhoPC mutant strain .
We wished to analyze whole-cell proteins of strain CS022 to estimate the number of protein species that could be potentially regulated by the PhoP regulon .
Remarkably , analysis by one-dimensional polyacrylamide gel electrophoresis of the proteins produce by strains with the PhoPc phenotype indicated that some protein species were decreased in expression when many presumptive pag gene products were fully induced by the pho-24 mutations ( Fig. 1 ) .
The proteins decreased in the PhoPc strain might represent products of genes that are repressed by the PhoP regulator ( see Discussion ) .
To simplify discussion , we have tentatively designated the genes encoding proteins decreased by the pho-24 allele prg loci , for phoP-repressed genes .
Comparison of wild-type , PhoP - , and PhoPc mutant strain proteins shows that growth in LB-medium at 37 °C represents repressing conditions for pag gene products and derepressing conditions for prg gene products .
To estimate the total number of potentially PhoP-regulated gene products , the total cell proteins of wild-type and PhoPc mutant strains grown in LB were analyzed by two-dimen-sional gel electrophoresis .
At least 40 species underwent major fluctuation in expression in response to the pho-24 mutation ( Fig. 2 ) , although we have indicated only 15 well-resolved species in this figure .
Virulence defects of the PhoPc strain .
The marked changes in protein expression of the PhoPc strain led us to investigate its virulence characteristics .
Remarkably , strains with the single pho-24 mutation were markedly attenuated for virulence in mice ( Table 2 ) .
The number of PhoPC organisms ( 2 x 105 ) that killed 50 % of BALB/c mice challenged ( LD50 ) by the i.p. route was near that ( 6 x 105 ) of PhoP-bacteria ( 17 ) .
The PhoPc strains had growth comparable to wild-type organisms in rich and minimal media ( 9 ; data not shown ) .
We also tested PhoPc mutants for alteration in lipopolysaccharide , which could also explain the virulence defect we observed .
Strain CS022 had normal sensitivity to phage P22 , normal group B reactivity to antibody to 0 antigen , and a lipopolysaccharide profile identical to that of the parent strain by polyacrylamide gel electrophoresis and staining ( data not shown ) .
10428 Wild type 180 ( A ) ATCC ; 17 TA2367 pho-24 1,925 ( A ) 9 CS003 AphoP ApurB < 10 ( A ) 17 CS022 pho-24 1,750 ( A ) This work CS023 pho-24 phoN2 zxx : : 6251 TnlOd-25 ( A ) This work Cam CS012 pagAl : : Mu dJ 45 ( B ) 17 CS013 pagBl : : Mu dJ 120 ( B ) 17 CS119 pagCl : : TnphoA phoN2 85 ( C ) 17 zxx : : 6251 TnlOd-Cam SC024 pagAl : : Mu dJ pho-24 450 ( B ) This work CS025 pagBl : : Mu dJ pho-24 980 ( B ) This work CS026 pagCl : : TnphoA pho-24 phoN2 385 ( C ) This work zxx : : 6251 TnlOd-Cam CS015 phoP102 : : TnlOd-Cam < 10 ( A ) 17 TT13208 phoP105 : : TnlOd < 10 ( A ) b a A , Acid phosphatase ; B , , B-galactosidase ; C , alkaline phosphatase .
b Gift of Ning Zhu and John Roth .
... 29 prg 's pag 's FIG. 1 .
Denaturing polyacrylamide gel electrophoresis of wholecell proteins of strains ATCC 10428 ( wild-type virulent parent ; PhoP + ) , CS015 ( PhoP - ) ( 17 ) , and CS022 ( PhoPC ) .
The gel ( 12.5 % polyacrylamide-sodium dodecyl sulfate ) was run in buffer conditions as described by Laemmli ( 11 ) .
The gel was fixed and stained with Coomassie brillant blue R250 in 10 % acetic acid-10 % metha-nol .
Arrows on the right indicate activated species ( pag 's ) ; arrows on the left indicate repressed species ( prg 's ) .
The numbers indicate the positions of molecular size standards ( in kilodaltons [ KD ] ) run in adjacent lanes .
The whole-cell protein extracts were from station-ary-phase cultures grown in Luria broth .
N , The 26-kilodalton nonspecific acid phosphatase that is the product of the phoN gene .
PhoPC Since the TA2367 pho-24 strain was constructed by chemical mutagenesis and could have another linked mutation responsible for its virulence defect , we wished to isolate revertants of the PhoPc phenotype to document that the pho-24 allele was responsible for the attenuation of virulence we observed .
Phenotype PhoPC revertants , identified by their normal levels of acid phosphatase in rich-medium , were isolated among the bacteria recovered from the livers of a Organisms were administered by the oral route .
In all other experiments , organisms were administered by i.p. challenge mice infected with strain CS022 .
Six separate phenotypic revertants , designated CS122 to CS128 , were found to be fully virulent ( LD50 of less than 20 organisms for BALB/c mice ) .
The locus responsible for the reversion phenotype was mapped in all six revertants tested for virulence by bacteriophage P22 cotransduction and had linkage characteristics consistent with the phoP locus ( greater than 90 % linkage to purB ) .
We therefore conclude that these reversion mutations are not extragenic suppressors but are intragenic suppressors or true revertants of the pho-24 mutation .
Thus , the virulence defect of PhoPc mutants is probably the result of a single revertable mutation in the phoP locus and not the result of a second unrelated mutation acquired during mutagenesis .
Reversion frequency of the PhoPC phenotype .
We investigated the reversion frequency of the PhoPC mutation in-vivo in mice to assess whether reversion could reduce the LD50 of this strain .
We tested for the presence of revertants of strain CS022 by administering 106 , 104 , and 102 challenge organisms to each of eight animals by i.p. injection .
On day 7 , three animals died that received 106 PhoPC organisms .
On that day , the livers and spleens of all animals were harvested and homogenized in saline .
After appropriate dilution , 10 % of the tissue was plated on LB plates containing the chromogenic phosphatase substrate XP .
Revertants were identified by their lighter blue colonies compared with PhoPc bacteria and were confirmed by quantitative acid phos-107 , 105 , phatase assays .
An estimated and 103 organisms per organ were recovered from animals at each of the three respective challenge doses .
Revertants were identified only at the highest dose and 105 comprised 0.5 to 1 % , or organisms per organ , at the time of death .
It is likely that revertants are able to compete more effectively for growth in these macrophage-containing organs , since strain CS022 is deficient in survival within macrophages ( see below ) .
However , revertants were not identified if fewer than 105 organisms were administered in the challenge dose , suggesting that the reversion frequency must be approximately 10-5 .
The reversion rate of the PhoPC phenotype for CS022 10 ' bacteria grown in LB is in fact 6 x when scored by the same colony phenotypes .
The percentage of revertants recovered from animals near death suggests that pressure is applied in-vivo that selects for revertants of the PhoPC phenotype and implies that the virulence defect we observed could be much greater quantitatively for a strain with a nonrevertable PhoPC mutation .
The PhoPC strain is deficient in survival within macro-phages .
Because of the importance of survival within mac-rophages to Salmonella virulence ( 5 ) , we tested PhoPC bacteria for this property .
Strain CS022 was defective in the ability to grow and persist in macrophages as compared with wild-type organisms ( Fig. 3 ) .
PhoP-bacteria seemed to have a macrophage survival defect qualitatively similar to that of PhoPC bacteria but survived consistently better by two-to threefold in side-by-side experiments ( data not shown ) .
The increased recovery of organisms that reverted the PhoPC phenotype in mouse organs rich in macrophage content is consistent with the reduced macrophage survival of PhoPC mutants in-vitro .
Use of the PhoPc strain as a live vaccine .
We previously reported that PhoP-strains were useful as live vaccines in protecting against mouse typhoid ( 17 ) .
Therefore , we wished to compare the PhoPC and PhoP-strains for immunogenicity when used as live attenuated vaccines in mice .
This was done by simultaneous determination of survival , after graded challenge doses with the wild-type strain ATCC 10428 , in mice previously immunized with graded doses of the two live vaccine strains , CS015 phoP : : TnJOd-Cam and CS022 pho-24 , as well as a saline control .
The results obtained ( Table 2 ) suggest the following conclusions : ( i ) small i.p. doses of the PhoPC strain ( e.g. , 15 organisms ) effectively protect mice from challenge doses as large as 5 x 105 bacteria ( a challenge dose that represents greater than 104 i.p. ( ii ) large LD50s ) , doses of PhoPc organisms given orally completely protect mice from an oral challenge consisting of 5 x 107 wild-type bacteria ( over 200 oral wild-type LD50s ) and ( iii ) by comparison , a large dose of PhoP-organisms ( 5 x 105 ) does not provide similar protection .
The reversion of the PhoPC mutation in-vivo somewhat complicates the analysis of the use of these strains as vaccines , since revertants of the CS022 strain ( i.e. , wild-type cells ) could increase immunogenicity .
However , we were unable to identify revertants by examining 10 % of the available spleen and liver tissue from those mice that received 104 or fewer organisms .
Two-dimensional electrophoresis performed on wild-type and PhoPC whole-cell protein samples according to the method of O'Farrell ( 18 ) The numbers indicate some of the most easily seen changes in species .
Odd numbers indicate species repressed in the PhoPC strain ( potential prg loci ) , and even numbers indicating species activated by the PhoPC mutation ( pag loci ) .
Vitamin D-dependent calcium-binding protein ( 40 ng ; pI 5.2 ) was added to the samples ; this standard is indicated by the arrow on the stained gel .
The final tube gel pH gradient extended from pH 4.1 to 8.1 ( from right to left ) .
Molecular weight standards are marked by bars and correspond ( from top to bottom ) to myosin ( 230,000 ) , phosphorylase A ( 94,000 ) , catalase ( 60,000 ) , actin ( 43,000 ) , and lysozyme ( 14,000 ) .
The 10 % o acrylamide slab gels are silver stained ( 14 ) .
Virulence and protective efficacy of PhoPC and PhoP-Salmonella strains No .
of survivors/total after wild-type initial challenge dose of : survivors/total 5 x 107a 5 x 105 5 x 104 5 x 103 Immunizing dose PhoPc organisms 15 50 1.5 x 102 5 x 102 1.5 x 103 5 x 103 1.5 x 104 S x 104 1.5 x 105 5 x 105 5 x 106 3 x 1o9a 3 x lo10a 1.5 x lolla PhoP-organisms 6 x 103 6 x 104 6 x 105 5 x lo10a 5/5 4/5 3/3 13/13 4/4 11/11 16/16 5/5 4/4 5/5 19/23 5/5 1/4 0/6 5/5 5/5 5/5 36/36 36/36 19/36 7/7 4/4 4/4 4/4 4/4 3/3 2/2 3/3 2/2 4/4 1/1 3/3 2/2 5/5 5/5 5/5 0/12 0/12 0/6 0/12 0/12 0/6 0/12 3/12 4 / 3/7 7T 61 0 5 .
3 + 2 + 0 w U * - ~ ~ ~ ~ ~ ~ ~ 0 I - J 0 24 4 8 12 16 20 HOURS AFTER INFECTION OF MACROPHAGES FIG. 3 .
Survival of strain CS022 ( PhoPC ) ( A ) in cultured macro-phages as compared with wild-type S. typhimurium ATCC 10428 ( A ) .
The experiment shown is a representative one .
The difference between the two strains at four and 24 h is significant ( P < 0.05 ) .
DISCUSSION In this report , we show that a phoP locus mutation ( pho-24 ) results in constitutive expression of genes activated by the PhoP-PhoQ two-component regulatory system and in attenuation of virulence and survival within cultured macro-phages ( Table 2 and Fig. 3 ) .
Previously , we ( 17 ) as well as others ( 4 ) have shown that strains with null mutations in phoP or phoQ have reduced survival within macrophages , increased sensitivity to the NP-1 defensins , and reduced lethality for mice ( 4 , 17 ; Miller et al. , submitted ) .
Null mutations in phoP and phoQ decrease the expression of the same genes whose expression is derepressed by the pho-24 mutation ( i.e. , phoN , pagA , pagB , and pagC ) .
Therefore , it is apparent that mutations that inactivate or activate the PhoP-PhoQ regulon can attenuate virulence .
Given that insertion mutations in at least one phoP-activated gene , pagC , result in decreased virulence and survival within macrophages , the attenuation of PhoP-and PhoQ-strains seems logical .
The molecular basis for attenuation of virulence of PhoPC strains may be related to the observation that in addition to the expected increase in expression of pag gene products , PhoPC strains show reduced expression of many other proteins detectable by two-dimensional gel electrophoresis For purposes of simplicity of discussion , we have termed the genes encoding these repressed proteins prg loci , for phoP-repressed genes .
If one or more of these prg gene products are essential to virulence , this may explain the virulence attenuation of PhoPc S. typhimurium .
Several possibilities exist for the mechanism of repression of protein species by the phoP constitutive mutation .
The PhoP protein may bind to DNA and function as a transcriptional repressor at prg loci .
The ability to function as a repressor as well as an activator is consistent with the alternate expression of the porin genes ompF and ompC by the analogous two-component regulator OmpR ( 7 ) , as well as the identification of both repressed and activated gene products controlled by the coordinate virulence regulators of Staphylococcus aureus ( 19 ) , Vibrio cholerae ( 21 ) , and Bordetella pertussis ( 10 ) .
Another possibility is that a pag locus encodes a transcriptional repressor of prg loci and the system thus functions as a cascade .
If the pho-24 mutation is inphoQ and results in full activation of PhoQ kinase activity , this could lead to activation , through phosphorylation , of another PhoP-like transcriptional regulator that functions as a repressor .
It is also possible that the activation of pag loci leads to posttranscriptional protein effects , such as degradation or modification , that cause a change in the protein expression pattern observed .
Identification of phoP-repressed gene fusions should help resolve the mechanism of these global changes in protein expression .
The demonstration that both PhoP-and PhoPc bacteria are defective in survival within macrophages suggests that both induction and repression of the PhoP regulon may be necessary for survival in the intracellular environment .
This demonstration , coupled with the fact that PhoP and PhoQ are environmentally responsive regulators , further suggests that the products of pag and prg loci may contribute to intracellular survival either in different environments or in a different temporal sequence .
The fact that PhoP-Salmo-nella strains are markedly sensitive to defensins ( 4 ) , microbicidal peptides of mammalian phagocytes ( 13 ) , may provide a clue to the intracellular environment in which pag gene products are required .
We have recently shown that unlike PhoP-strains , PhoPc bacteria are as resistant to defensins as are wild-type bacteria , indicating that defensin resistance is likely encoded by a pag gene product ( Miller et al. , submitted ) .
Defensins are active in-vitro only at neutral pH , and the killing effect of these molecules has been postulated to be most important in the immediate postphagocytic period , when the pH of the phagosome has been documented to be transiently neutral to alkaline ( pH 7.8 ) ( 13 ) .
It is possible that pag gene products , including those necessary for defensin resistance , are most important in the early stages of phagocytosis , perhaps before phagolysosome fusion .
Other factors encoded by prg loci may be more important to survival within macrophages at a time when defensins are not active ( i.e. , after acidification of the phagolysosome ) .
Thus , a switch in the PhoP regulon from pag to prg gene expression might be essential to efficient macrophage survival after salmonellae are phagocytosed .
The pathogenesis of mouse typhoid involves passage of bacteria through multiple environments , including gut lumen , endothelial cells , phagocytes , lymphatics , and bloodstream .
The macrophage survival defect of PhoPc bacteria be sufficient may to explain the attenuation we observed , and this may be due to the inability of PhoPc organisms to synthesize prg gene products .
However , it is also possible that the reduced virulence of the PhoPc phenotype demonstrates how important balanced regulation is to the pathophysiologic process .
Full virulence is dependent not simply on the synthesis of virulence-factors but also on the ability to regulate their expression in the appropriate environment ( 3 , 15 ) .
The attenuation observed after overexpression of virulence properties as a result of constitutive mutations in virulence regulators provides evidence that the ability to turn off the expression of loci encoding essential virulence-factors may be as important as the ability to turn them on .
Further analysis of the PhoP-PhoQ regulon should lead to greater understanding of the molecular basis of S. typhimu-rium virulence and survival within macrophages .
The marked immunogenicity of strains with the PhoPc phenotype suggests that pag genes may be important antigens for the induction of protective immunity against Salmonella species and that PhoPc strains with nonrevertable mutations could be useful as live vaccines in humans and other animals .
This work was supported by a grant from the Rockefeller Foundation and by Public Health Service grants AI-18045 and AI-26289 from the National Institutes of Health .
S.I.M. was a fellow of the Medical Foundation during the time this work was performed .
We thank Bruce Ames , Gigi Stortz , John Roth , and Nang Zhu for the generous gift of strains .
We thank all members of the Mekalanos laboratory for helpful discussions and support .
Buchmeier , N. A. , and F. Heffron .
Intracellular survival of wild-type Salmonella typhimurium and macrophage-sensitive mutants in diverse populations of macrophages .
Davis , R. W. , D. Botstein , and J. R. Roth .
Advanced bacterial genetics , p. 78 , 87 .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 3 .
Dirita , V. J. , and J. J. Mekalanos .
Genetic regulation of bacterial virulence .
Fields , P. I. , E. A. Groisman , and F. Heifron .
A Salmo-nella locus that controls resistance to microbicidal proteins from phagocytic cells .
Fields , P. I. , R. V. Swanson , C. G. Haidairis , and F. Heffron .
Mutants of Salmonella typhimurium that can not survive with the macrophage are avirulent .
Groisman , E. A. , E. Chiao , C. J. Lipps , and F. Heffron .
Salmonella typhimurium phoP virulence gene is a transcriptional regulator .
Hall , M. N. , and T. J. Silhavy .
Genetic analysis of the major outer membranes of Escherichia coli .
Kier , L. K. , R. Weppleman , and B. N. Ames .
Regulation of two phosphatases and a cyclic phosphodiesterance of Salmo-nella typhimurium .
Kier , L. K. , R. M. Weppleman , and B. N. Ames .
Regulation of nonspecific acid phosphatase in Salmonella : phoN and phoP genes .
Knapp , S. , and J. J. Mekalanos .
Two trans-acting regulatory genes ( vir and mod ) control antigenic modulation in Bordetella pertussis .
Cleavage of structural proteins during the assembly of the head of bacteriophage T4 .
Lissner , C. R. , R. Swanson , and A. O'Brien .
Genetic mice to Salmonella control of the innate resistance of typhimu-rium : expression of the Ity gene in peritoneal and splenic macrophages isolated in-vitro .
Lehrer , R. I. , M. E. Selsted , D. Szklarek , and J. Fleischmann .
Antibacterial activity of microbicidal cationic proteins 1 and 2 , natural peptide antibiotics of rabbit lung macrophages 14 .
Merril , C. R. , D. Goldman , and M. L. Van Keuren .
Gel protein strains : silver stain .
Miller , J. F. , J. J. Mekalanos , and S. Falkow .
Coordinate regulation and sensory transduction in the control of bacterial virulence .
Experiments in molecular genetics , p. 352-355 .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 17 .
Miller , S. I. , A. M. Kukral , and J. J. Mekalanos .
A two component regulatory system ( phoP and phoQ ) controls Salmo-nella typhimurium virulence .
High resolution two-dimensional electro-phoresis of proteins .
Recesi , P. , B. Kreiswirth , M. O'Reilly , P. Schlievert , and A. Gruss .
Regulation of exoprotein gene expression in Staph-ylococcus aureus by agr .
Ronson , C. W. , B. T. Nixon , and F. M. Ausubel .
Conserved domains in bacterial regulatory proteins that respond to environmental stimuli .
Taylor , R. K. , V. L. Miller , D. B. Furlong , and J. J. Mekalanos .
Use of phoA gene fusions to identify a pilus colonization factor coordinately regulated with cholera toxin .