All accessible epitopes in the Salmonella lipopolysaccharide core are associated with branch residues.

Antisera generated against each of the nine known chemotypes of Salmonella lipopolysaccharide (LPS) core were characterized in order to delineate cross-reactive epitopes and define the bases for their accessibility. Strongly cross-reactive epitopes were associated with three chemotypes: Ra and Rb4, which recognized alpha-GlcNAc-1-->2-alpha-Glc, and Rd1, which recognized L-alpha-D-heptose-1-->7-L-alpha-D-heptose. Both these disaccharides and the more weakly cross-reactive alpha-Gal-1-->6-alpha-Glc terminal in Rb3 LPS represent branch points along the core oligosaccharide. Therefore, branch points in endotoxin core oligosaccharides may generally be cross-reactive.

alpha-GlcNAc-1-->2-alpha-glc, the Salmonella homologue of a conserved lipopolysaccharide motif in the Enterobacteriaceae, elicits broadly cross-reactive antibodies.

To define cross-reactive epitopes in Salmonella lipopolysaccharide (LPS), antisera designated anti-S, anti-Ra, and anti-Re were generated against smooth (S), complete-core (Ra), and deep-core mutant (Re) strains, respectively, and characterized immunochemically. The reactivities of anti-Ra and anti-S with rough LPS (rLPS) chemotypes in enzyme-linked immunosorbent assays (ELISA) decreased progressively with increasing truncation of the complete-core oligosaccharide (e.g., Ra > Rb1 >.Re), while that of anti-Re increased (Ra < Rb1 <.Re). Anti-Ra was relatively more reactive with nonhomologous smooth LPS (sLPS) than anti-S, which in turn was more reactive than anti-Re. This order reflected the relative reactivities of these sera with outer-core rLPS but not those with inner-core rLPS, which suggests that the cross-reactivities of all three sera with sLPS were mediated by antibodies which bind outer-core determinants. Anti-Ra, but not anti-S or anti-Re, reacted with molecules substituted by O chains in immunoblots and revealed ladder-like patterns in sLPSs of various serospecificities. Anti-Ra, however, did not react with O-antigen-specific neoglycoconjugates in ELISA, thus demonstrating specificity for core epitopes. Ra and Rb1 but not other Salmonella core chemotypes inhibited the reactivity of anti-Ra with sLPS in ELISA, which showed that the terminal outer-core disaccharide, alpha-GlcNAc-1-->2-alpha-Glc (GlcNAc-->Glc), was the major epitope of cross-reactive antibodies in the serum. GlcNAc-->Glc represents the conserved motif alpha-hexose-1-->2-alpha-hexose in cores of the Enterobacteriaceae, other homologues of which should likewise be cross-reactive. These results demonstrate that S or Re strains do not elicit cross-reactive antibodies and indicate that immunization with Ra strains may represent a general strategy for eliciting cross-reactive antibodies against LPSs from enteric bacteria.

Cross-reactivity between the mannan of Candida species, Klebsiella K24 polysaccharide and Salmonella C1 and E O-antigens is mediated by a terminal non-reducing beta-mannosyl residue.

Rat monoclonal antibody MASC1-MR9 (MR9) binds to a mannan of Candida species and the O-antigenic polysaccharides of Salmonella bacteria of serogroups C1 (CO) and E (EO). Mannan and glycoconjugates comprising BSA and O-antigen polysaccharides, decasaccharide-BSA (CO-BSA) or trisaccharide-BSA (EO-BSA), inhibited each other's reactivity with MR9. The saccharides beta-D-Manp-(1-->6)-alpha-D-Manp-1-OMe, beta-D-Manp(1-->3)-alpha-D-Manp-1-OMe, beta-D-Manp(1-->2)-alpha-D-Manp-1-OMe (corresponds to the terminal non-reducing end of Salmonella serogroup C1 O-antigen) and beta-D-Manp(1-->4)-alpha-L-Rhap(1-->3)-alpha-D-Galp-1-O-p-++ +trifluoroacetamido aniline (corresponds to the backbone of Salmonella serogroup E O-antigen) inhibited the binding of MR9 to these antigens whereas alpha-D-Manp(1-->3)-alpha-D-Manp-1-OMe and alpha-D-Manp(1-->4)-alpha-L- Rhap-1-O-p-nitrophenyl did not. Saccharides (3-10 residues) of mammalian origin with terminal and internal Manp alpha-1-->2, Manp alpha-1-->3 and Manp alpha-1-->6 residues also failed to inhibit at any concentration. None of the saccharides with internal beta-mannosyl residue was able to inhibit the MR9 antibody. Monosaccharides D-mannose, beta-D-Manp-1-OMe and 1,5 anhydro-D-mannitol inhibited the MR9 monoclonal antibody whereas alpha-D-Manp-1-OMe, beta-D-Glcp-1-OMe, and beta-D-Galp-1-OMe did not. In addition a Klebsiella K24 capsular polysaccharide containing a beta-D-Manp(1-->4)-alpha-D- GlcA (GlcA, glucuronic acid) as a structural element possessed an inhibitory activity. MR9 therefore recognizes an epitope within beta-mannose monosaccharide residues at the terminal non-reducing ends of carbohydrate chains in mannan, and polysaccharides in Salmonella serogroups CO and EO and Klebsiella K24.

Clinicopathological assessment of gastric biopsy samples of patients with Helicobacter pylori infection--metronidazole resistance and compliance problems in the United Arab Emirates.

The significance of Helicobacter pylori (HP) infection was assessed prospectively in forty-two patients with dyspepsia using histological, bacteriological and biopsy urease techniques. Thirty-eight patients (90.5%) were positive for HP infection and were treated with bismuth subcitrate (De Nol), tinidazole and doxycycline. HP was present in the antrum, corpus, fundus, duodenum and gastric juice in 36, 26, 23, 2 and 2 patients respectively (p < 0.01, X2 test). Histological assessment yielded more positive identifications of HP than the urease test (36 vs 28 positive cases, p < 0.01, McNemar's X2 test), while histology and bacteriology were virtually identical (38 vs 37 of 41 pairs, p > 0.5, X2 test). There was a good correlation between bacterial and polymorphonuclear leucocyte (PMNL) counts per high power field (r = 0.8; p < 0.001; n = 34 pairs). There was resistance to metronidazole in 10 out of 16 isolates, but no resistance was recorded against tetracycline (p < 0.001, X2 test). Among the sixteen patients who attended follow-up endoscopy, there was clinical improvement and no evidence of HP in 5 individuals (31.25%). One patient had amelioration of his symptoms, 5 experienced no change and in 5 their symptoms became worse. Metronidazole resistance may be one of the important factors in the United Arab Emirates and elsewhere.

The disaccharide L-alpha-D-heptose1-->7-L-alpha-D-heptose1-->of the inner core domain of Salmonella lipopolysaccharide is accessible to antibody and is the epitope of a broadly reactive monoclonal antibody.

We generated a panel of mAb containing at least one specificity against each of the known chemotypes of the Salmonella LPS core domain and used them to investigate the accessibility of core determinants in smooth LPS. Most of the mAb were reactive with at the most three chemotypes of the core as determined by enzyme immunoassay and failed to bind smooth LPS or any of the complete cores of E. coli. One mAb, MASC1-MM3 (MM3), reacted with six different Salmonella core chemotypes, the R2 core of Escherichia coli and a variety of smooth LPS. This mAb reacted equally well with live and heat-killed bacteria. It bound to 123 of 126 clinical isolates of Salmonella and 11 of 73 E. coli strains in a dot-immunoblot assay. Typical ladder-like patterns of bands were observed after immunoblotting of this mAb against electrophoretically resolved smooth LPS from the five major serogroups of Salmonella species (A, B, C1, D1, and E). MM3 had no reactivity with BSA conjugates of O-Ag polysaccharides from the above serogroups confirming specificity for a core epitope. Polysaccharides derived from or synthetic saccharides representative of the various chemotypes of Salmonella LPS core were tested as competitive inhibitors of the binding of MM3 to LPS. The results led to a conclusion that MM3 recognizes the structure, L-alpha-D-Heptose1-->7-L-alpha-D-Heptose1-->disaccharide present as a branch in the Ra, Rb1, Rb2, Rb3 and Rc but lacking in the Rd1, Rd2, and Re chemotypes of the Salmonella LPS core. This disaccharide seems free and accessible on the basis of the previously calculated conformations of the Salmonella (Ra) and E. coli complete cores (R1, R2, R3, R4, and K12). It therefore defines or contains an epitope within the inner core subdomain of Salmonella LPS that is accessible to antibody in long-chained LPS and in intact bacteria with complete LPS.

The role of O-antigen polysaccharide in the activation of neutrophils by lipopolysaccharides of Salmonella species.

Activation of neutrophils by lipid A, O-antigen polysaccharides (PS) and smooth lipopolysaccharides (LPS) isolated from Salmonella choleraesuis (O-6,7) and Salmonella typhimurium (O-4,5,12) was investigated. The methods used were assays for lysozyme release and for nitroblue tetrazolium (NBT) reduction which measures the level of oxidative metabolism of neutrophils. LPS from both species stimulated neutrophils to the same extent in the presence of autologous plasma. In the absence of plasma only the O-6,7 LPS activated neutrophils. Lipid A or PS isolated from both LPS either did not activate neutrophils or did so only at very high concentrations when tested in the presence of plasma; in the absence of plasma no activation occurred. The data indicate that both PS and lipid A segments of LPS are required for activation of neutrophils by LPS. We also deduce that plasma, probably complement, is required for the interaction of some LPS, e.g. O-4,5,12 with neutrophils whereas other LPS, e.g. O-6,7 can interact directly and activate neutrophils.

Salmonella choleraesuis and Salmonella typhimurium associated with liver cells after intravenous inoculation of rats are localized mainly in Kupffer cells and multiply intracellularly.

Male Sprague-Dawley rats were inoculated intravenously with Salmonella choleraesuis or Salmonella typhimurium and used over 3 consecutive days to produce highly enriched (greater than 95% homogenous) preparations of Kupffer and mononuclear cells (KC), liver endothelial cells (LEC), and hepatocytes. The methods involved collagenase perfusion of the liver in situ, differential centrifugation of liver cells over a Percoll gradient, and selective attachment of the cells to plastic or to culture dishes coated with collagen. The different cell preparations were then assayed for the number and location, intracellular or extracellular, of associated viable bacteria. Most of the viable bacteria recovered were associated with KC and were mainly intracellular. The intracellular bacteria in KC from rats infected with either bacterial strain increased about 20- to 50-fold over 2 days. Some of the bacteria associated with LEC and in some experiments with hepatocytes also survived treatment with gentamicin and increased in number with time. Intracellular bacteria were readily visualized in KC by light microscopy and transmission electron microscopy. On rare occasions, bacteria were seen within LEC from rats infected with S. choleraesuis but not from those infected with S. typhimurium. Microcolonies of S. typhimurium but not of S. choleraesuis were occasionally found on the surface of some LEC. Bacteria were not seen within or on the surface of hepatocytes by transmission or scanning electron microscopy. The integration of microscopic and viability data suggested that most intracellular S. choleraesuis organisms in KC had been killed whereas most intracellular S. typhimurium organisms were viable.

Relevance of inoculation route to virulence of three Salmonella spp. strains in mice.

The virulence of three Salmonella species strains was compared by the i.p. and i.v. routes in BALB/c mice. Salmonella choleraesuis, SL2824 (serogroup C1, O-6,7), was more virulent by the i.v. than i.p. route. A strain of S. typhimurium, SL1260 (serogroup B; O-1,4,12) was more virulent i.p. than i.v. while another strain, SL3201 (O-4,5,12) was equally virulent i.p. or i.v. The LD50 of each strain by either route correlated with the number of bacteria in the liver and spleen on day one after inoculation and thus seems determined mainly by initial bactericidal mechanisms. The rate of bacterial growth in the liver and spleen was independent of inoculation route but differed between the three strains. Salmonella choleraesuis multiplied faster than either strain of S. typhimurium. Non-virulent aromatic-dependent (aro) derivatives of these strains were tested, instead of their virulent ancestors, for survival within peritoneal macrophages in vitro. Salmonella choleraesuis SL 2824 aro and S. typhimurium SL1260 aro were much more readily killed intracellularly than S. typhimurium SL3201 aro. The data indicate that the survival and multiplication of different Salmonella serotypes or strains in vivo may depend on different critical properties or mechanisms to overcome host defenses.

Properties of a rat monoclonal antibody reactive with both the mannan of Candida species and the O-antigen 6,7 polysaccharide of serogroup C1 salmonellae.

Monoclonal antibody MASC1-MR9 of isotype immunoglobulin M was generated in LOU/C rats by immunization with heat-killed Salmonella thompson (serogroup C1, O-antigen 6,7 lipopolysaccharide) bacteria which had been further enriched for O-antigen by coating with homologous lipopolysaccharide. Eight monoclonal antibodies were selected by screening against the lipopolysaccharide of S. thompson. In a subsequent test against Candida mannan, only one antibody, MASC1-MR9, was reactive. Immunofluorescence microscopy of various yeasts and bacteria with MASC1-MR9 showed that this antibody bound to the surfaces of 142 of 148 Candida strains of 10 different species tested but failed to bind to the surface of Saccharomyces cerevisiae. The Candida strains which failed to bind MASC1-MR9 were all five strains of Candida krusei and the single strain of Candida utilis tested. Several (11 of 33) Salmonella strains belonging to serogroup C1 reacted with this antibody, as expected; however, 1 of 11 strains of Morganella morganii and 4 of 64 strains of Escherichia coli were also reactive. Serum-free supernatants of MASC1-MR9 agglutinated S. thompson, one strain of Salmonella choleraesuis, and 109 of 110 Candida strains tested. The immunochemical properties of MASC1-MR9 as studied by enzyme-linked immunosorbent inhibition assays show that it recognizes an epitope present in the mannan of Candida species as well as in the O-6,7 antigen of Salmonella species.

Salmonella choleraesuis strains deficient in O antigen remain fully virulent for mice by parenteral inoculation but are avirulent by oral administration.

O-antigen-deficient derivatives of two mouse-virulent strains of Salmonella choleraesuis (serogroup C1; O-6,7) were constructed by transduction of a long deletion of the rfb operon. Strains SN36 and SN57 were derived from the smooth ancestor SL2824, while SN37 was derived from the smooth ancestor SL2840. These rfb deletion derivatives (rfb strains) had typical bacteriophage sensitivity patterns of "rough" Salmonella strains and were at least 200,000 times more sensitive to serum than their smooth ancestors. Lipopolysaccharides (LPS) extracted from them consisted only of two low-molecular-weight bands and lacked the ladderlike pattern of bands seen in the LPS of their smooth ancestors. The LPS from the rfb strains did not react in an enzyme immunoassay with any of three monoclonal antibodies directed against different epitopes of the O-6,7 antigen but reacted well with at least one of three monoclonal antibodies specific for core epitopes. The data were consistent with inability of these strains to synthesize O-specific chains and showed that the LPS extracted from SN57 was of chemotype Ra and that from SN36 was of chemotype Rb1, while that of SN37 consisted of a mixture of the two chemotypes. The virulence of these strains was tested by various routes in BALB/c mice. All three O-antigen-deficient derivatives were about as virulent as their "smooth" ancestors by the intraperitoneal and intravenous routes (50% lethal dose, 20 to 700 bacteria) but, unlike their ancestors, were avirulent by the oral route (50% lethal dose, greater than or equal to 5 x 10(9) bacteria). This suggests that the major role of smooth LPS in the mouse virulence of S. choleraesuis is to facilitate survival in the gastrointestinal tract and eventual entry into deeper tissues.

Lysogenization of Salmonella choleraesuis by phage 14 increases average length of O-antigen chains, serum resistance and intraperitoneal mouse virulence.

Three clones from a strain of Salmonella choleraesuis (serogroup C1) were lysogenized with phage 14 (P14) which converts the O-antigen of serogroup C1 salmonellae from O-6,7 to O-6,7,14. The lysogens were compared with their parental non-lysogenic clones with respect to the following properties: average length of O-antigen polysaccharide chains, sensitivity to normal human serum, and mouse-virulence. SDS-polyacrylamide gel electrophoresis of lipopolysaccharides extracted from these bacteria showed that samples from lysogens consisted mainly of long-chained molecules whereas those from non-lysogens contained mainly short-chained molecules. The O-antigen polysaccharide from a lysogen was estimated by chemical analysis to be six times as long as that from a non-lysogen. Lysogens were serum-resistant whereas non-lysogens were serum-sensitive. About 10 times more colony forming units of a lysogen than of a non-lysogen were recovered from the livers and spleens of mice on day 1 and 3 after intraperitoneal inoculation of equal doses. By comparison with S. choleraesuis, lysogenization of S. typhimurium with phage P22 or phage A4 did not affect the chain-length distribution of O-antigen polysaccharide. Our data suggest that phage 14-coded determinants increase efficiency of O-antigen biosynthesis in S. choleraesuis leading to increase in average length of O-polysaccharide chains. Increased serum resistance and mouse virulence are logical consequences of increase in average length of O-polysaccharide chains and represent phage-conferred selective advantage not previously described in Salmonella.

Efficient production of mouse and rat monoclonal antibodies against the O antigens of Salmonella serogroup C1, using LPS-coated bacteria as immunogen.

A variety of different immunogens and immunisation schemes were investigated for the production of monoclonal antibodies directed against the O antigenic polysaccharide of the lipopolysaccharide (LPS) of Salmonella serogroup C1 (O:6,7). Of 12 fusions performed, higher yields of stable, LPS-reactive hybridomas producing monoclonal antibodies were achieved in both the mouse and rat when using O:6,7 LPS-coated S. thompson bacteria as immunogens than with live and heat-killed bacteria, or O:6,7-BSA glycoconjugate as immunogens. All of the 17 hybrid clones obtained were shown to bind the O antigens of Salmonella serogroup C1 when tested in ELISA against a set of chemically defined LPS from Salmonella smooth and rough strains. The results are discussed with a view to bettering the immunisation strategy for production of monoclonal antibodies against the LPS antigens of bacteria.

Mice vaccinated with a non-virulent, aromatic-dependent mutant of Salmonella choleraesuis die from challenge with its virulent parent but survive challenge with Salmonella typhimurium.

BALB/c mice given a live vaccine of an aroA mutant of Salmonella choleraesuis by intraperitoneal (i.p.) injection were not protected against i.p. challenge with its virulent parental strain but were protected against i.p. challenge with either of two virulent strains of Salmonella typhimurium (O [1], 4, [5], 12). Vaccination with a live vaccine of S. typhimurium aroA protected against challenge with S. typhimurium but not with S. choleraesuis. Intraperitoneal administration of either aroA strain evoked high levels of serum antibody against the homologous lipopolysacharide (LPS) as determined by an enzyme immunoassay. Sera from vaccinated mice also reacted with heterologous LPS but at dilutions at least seven-fold lower than homologous endpoint titres. The vaccination schedule employed with either live-vaccine strain conferred an equal degree of resistance to challenge with Listeria monocytogenes. After mixed infection of mice with equal numbers of virulent S. typhimurium and S. choleraesuis by the i.p. route, the former was isolated in numbers at least 50,000 times greater than the latter from the liver and spleen between days 1 and 5. When mice were vaccinated i.p. with S. choleraesuis aroA, L. monocytogenes or P. multocida before mixed infection, neither serotype showed more than a slight predominance in the liver and spleen during the same period. Thus, in relative terms, vaccination with S. choleraesuis aroA or inoculation with unrelated bacteria suppressed the growth of virulent S. typhimurium in mice but allowed virulent S. choleraesuis to multiply. These results clearly show that S. choleraesuis 38(1) can multiply to kill immunised BALB/c mice.

Tn7 inserts in both orientations at a single chromosomal location and apparently forms cointegrates in Pasteurella multocida.

Pasteurella multocida transconjugants isolated after mating with Escherichia coli strains that carry one or the other of two Tn7-containing suicide plasmids, pRKTV5 and pUW964 (pRKTV5::Tn5), were analysed. These plasmids have the ColE1 replication origin and were thus expected to deliver transposons but not be maintained as free replicons in Pasteurella. Five out of six transconjugants selected for acquisition of Tn7 from E. coli (pRKTV5) had simple insertions of the transposon, in either orientation, at a single chromosomal location, while the sixth had pRKTV5 integrated at the same location. By contrast, all of 27 transconjugants selected for acquisition of either Tn7 or Tn5 from E. coli (pUW964) maintained pUW964. Of seven subsequently examined at the molecular level, all had pUW964 (in one case, a deletion derivative) integrated at the same location as the Tn7 insertions obtained with pRKTV5. A copy of Tn7 was present at each boundary between the integrated plasmids (pRKTV5 or pUW964) and the chromosome in each strain. The two copies of Tn7 at either end of an integrated plasmid were either in the same (six cases) or in opposite (two cases) orientations with respect to each other. These seem to be products of replicative transposition by Tn7 but can also derive from conservative mechanisms.

Transfer and properties of some natural and suicide replicons in Pasteurella multocida.

We tested the transfer of several plasmids and transposons from Escherichia coli to Pasteurella multocida by filter mating. Two plasmids, pRKTV5 (pRK2013::Tn7) and pUW964 (pRKTV5::Tn5), were derived from pRK2013--a narrow-host-range plasmid with the broad-host-range IncP conjugation genes. Most P. multocida transconjugants obtained with pRKTV5 had Tn7 insertions in the chromosome but some had insertions of the whole plasmid. By contrast, all the transconjugants obtained with pUW964 had insertions of this plasmid or a deleted variant. pUW964 mediated low-frequency transfer of Tn7 or chromosomal markers between P. multocida strains. Broad-host-range IncP plasmid RP4 (RK2) did not yield selectable transconjugants in P. multocida but two plasmids derived by Tn5 insertion into a kanamycin-sensitive derivative of RP4 did yield transconjugants. pSUP1011, a narrow-host-range p15A replicon with the RP4 mob region allowing mobilization by the IncP conjugation genes also yielded transconjugants while several other plasmids tested did not transfer markers to P. multocida.

Vaccination route, infectivity and thioglycollate broth administration: effects on live vaccine efficacy of auxotrophic derivatives of Salmonella choleraesuis.

An aromatic-dependent, therefore non-virulent, derivative of a mouse-virulent strain of Salmonella choleraesuis previously shown not to be effective as a live vaccine when given intraperitoneally (i.p.) to Itys mice, was administered to BALB/c mice. Two doses given i.p. or by feeding did not protect against i.p. or oral challenge with 50 to 5000 LD50 of the virulent ancestor strain. By contrast two doses given intravenously (i.v.) gave almost complete protection against i.p. or oral challenge with 500 LD50 and some protection against larger doses. The number of live bacteria (cfu) in the liver and spleen 24 h after administration of the live vaccine was less than 1% of the number inoculated i.p., but c. 25% of the number injected i.v. The number of cfu in the gut 24 h after oral vaccine administration was only c 10(-5) of the number fed. Administration of thioglycollate broth i.p. 5 days before i.p. vaccination increased recovery of live vaccine cfu in the liver and spleen and its protective efficacy. In each case the live vaccine did not multiply extensively in vivo. We have previously shown that a purine- and a thymine-requiring derivative of S. choleraesuis were each considerably attenuated but unlike the aro derivative were effective as i.p. live vaccines in mice. Doses of these strains (c. 10(4) cfu) found protective were administered i.p. to BALB/c mice. Each strain multiplied extensively in the liver and spleen to c. 10(7) cfu by day 6. All these results are in agreement with a correlation of protective efficacy of a live vaccine with the persistence of a large number of the vaccine bacteria in the liver and spleen for several days.

Evidence for a neutrophil-mediated protective response in malaria.

Zymosan-activated and non-activated human polymorphonuclear neutrophils (PMN) were added to in-vitro cultures of the human malaria parasite Plasmodium falciparum in microtitre wells. Microscopic counting of parasites in Giemsa-stained smears showed that at a PMN:RBC ratio of 1:150, the same as occurs in human malaria, parasites in wells with zymosan-activated neutrophils were suppressed 65%. Determination of parasite nucleic acid synthesis by 3H-hypoxanthine incorporation showed that in wells with PMN:RBC ratio of 1:150 parasite viability was only 22% of control. Various oxygen scavengers were tested for ability to reverse the effects of activated neutrophils on parasite development. Superoxide dismutase (20 mg/ml) and catalase (50 mg/ml) had no effect; tryptophan protected the parasites to a moderate degree while histidine alleviated suppression of parasite development to the greatest extent. This suggests that singlet oxygen is the most effective neutrophil product in killing or suppressing the growth of parasites. We also observed that non-activated neutrophils were activated by parasites and/or their products resulting in killing of newly-released parasites.

The effects of O-antigen character and enterobacterial common antigen content on the in vivo persistence of aromatic-dependent Salmonella sp. live-vaccine strains.

Aromatic-dependent (aro) derivatives of Salmonella choleraesuis like aro S. typhimurium are non-virulent but, unlike them, are ineffective as live vaccines in mice, given i.p. An aro derivative of S. choleraesuis did not persist in the liver and spleen (RES) of mice after i.p. inoculation whereas a similar derivative of S. typhimurium persisted. S. choleraesuis (O group C1; O-6,7) and S. typhimurium [O group B; O-(1),4(5),12] differ in O antigen of LPS, determined by chromosomal locus, rfb. Three pairs of nearly-isogenic aro derivatives, one member O-6,7 and the other O-(1),4,(5),12, were constructed in two lines of S. typhimurium by replacement of their B-rfb genes with the C1-rfb genes of S. choleraesuis. In tests for persistence after mixed or separate i.p. inoculation of equal doses into BALB/c mice the O-(1),4,(5),12 member of each pair was recovered as CFU in the RES at ca. 100-fold greater number than the O-6,7 member at 24 hours post-inoculation and subsequently. O-6,7 derivatives of S. typhimurium constructed as described above by a simple replacement of group B with group C-rfb locus synthesise only trace (tr) amounts of enterobacterial common antigen (ECA). An ECA+ (able to make normal levels of ECA) derivative of one aro, O-6,7 S. typhimurium strain was constructed by replacement of its B-rfe locus with the C-rfe locus of S. choleraesuis. Tested by mixed inoculation i.p. this strain persisted in the RES in numbers 10-fold greater than its O-6,7 ECAtr but 5-10-fold lesser than its O-(1),4,(5),12 cousins. Thus both O-specificity and ECA contribute to the survival of salmonella species in mice as determined by in vivo persistence of non-multiplying aro derivatives.

Test of the virulence and live-vaccine efficacy of auxotrophic and galE derivatives of Salmonella choleraesuis.

Aromatic compound-dependent (aro) derivatives of three mouse-virulent strains of Salmonella choleraesuis (Salmonella cholerae-suis) were constructed and shown to be nonvirulent for mice (intraperitoneal [i.p.] 50% lethal dose [LD50], greater than 5 X 10(6) CFU). A pur derivative, and a thy derivative, each of a different virulent parent, remained moderately virulent (i.p. LD50S for BALB/c mice, ca. 10(5) and 5 X 10(4) CFU, respectively). Tested as live vaccines i.p., the aro strains were ineffective in salmonella-susceptible BALB/c and C57BL/6 mice but were somewhat effective in salmonella-resistant CBA/J mice and in outbred CD-1 mice. The pur and thy strains were effective as live vaccines in BALB/c mice when given in sublethal doses. Two previously isolated nonvirulent galE derivatives of S. choleraesuis (i.p. LD50 in BALB/c mice, greater than 10(6) CFU) were also ineffective as live vaccines in BALB/c and C57BL/6 mice. The main antigenic difference between S. choleraesuis (O-6,7) and S. typhimurium (O-4,12) is in O-antigen character, thought to largely determine the specificity of protection in salmonellosis. Paired, nearly isogenic O-6,7 and O-4,12 derivatives were constructed from an aro S. typhimurium strain of proven efficacy as a live vaccine. Used as live vaccines, the O-4,12 member protected BALB/c mice against challenge with virulent S. typhimurium, whereas the O-6,7 member did not protect against virulent S. choleraesuis. However, BALB/c mice vaccinated with the O-6,7 member and mice vaccinated with an aro S. choleraesuis strain were protected against challenge with a moderately virulent (LD50, 5 X 10(4) CFU) O-6,7 derivative of an S. typhimurium strain.

Some galE mutants of Salmonella choleraesuis retain virulence.

galE mutants were isolated from three mouse-virulent strains of Salmonella choleraesuis (of group C1, O antigen 6,7) by selection for resistance to 2-deoxygalactose. The galE derivative of strain 381 comprised two components: galactose sensitive, thought to be the original mutant; and galactose resistant, presumably by a second mutation reducing galK or galT function or both. The galactose-sensitive component had an intraperitoneal 50% lethal dose for BALB/c mice of ca. 4 X 10(6) CFU, whereas the galactose-resistant component was about as virulent as its gal+ parent, with a 50% lethal dose of ca. 100 CFU. The galE mutant of strain 110 was somewhat sensitive to galactose, as shown by retardation of growth; its 50% lethal dose, ca. 500 CFU, was not much greater than the ca. 200 CFU value for its parent. The galE mutant of strain 117 showed the same partial sensitivity to galactose as strain 110 galE, but was nonvirulent (50% lethal dose of ca. 10(6) CFU versus ca. 400 CFU for its parent). Growth on galactose-supplemented medium restored the smooth phenotype, as indicated by phage sensitivity to three of the four galE strains, but only partially so for the strain 117 galE mutant. The retention of parental virulence by galE mutants of S. choleraesuis which are galactose resistant or somewhat galactose sensitive contrasts with the greatly reduced virulence of galactose-resistant galE mutants of Salmonella typhimurium and Salmonella typhi; this difference may result from the absence of galactose from the O repeat unit in the lipopolysaccharide of group C1 salmonellae.