Salmonella enterica serovar Typhimurium mutants completely lacking the F(0)F(1) ATPase are novel live attenuated vaccine strains

Abstract

The F(0)F(1) ATPase plays a central role in both the generation of ATP and the utilisation of ATP for cellular processes such as rotation of bacterial flagella. We have deleted the entire operon encoding the F(0)F(1) ATPase, as well as genes encoding individual F(0) or F(1) subunits, in Salmonella enteric serovar Typhimurium. These mutants were attenuated for virulence, as assessed by bacterial counts in the livers and spleens of intravenously infected mice. The attenuated in vivo growth of the entire atp operon mutant was complemented by the insertion of the atp operon into the malXY pseudogene region. Following clearance of the attenuated mutants from the organs, mice were protected against challenge with the virulent wild type parent strain. We have shown that the F(0)F(1) ATPase is important for bacterial growth in vivo and that atp mutants are effective live attenuated vaccines against Salmonella infection.

Fig. 1

Bacterial counts in spleens and livers following intravenous infection with various atp mutants or SL1344. Female BALB/c mice were infected intravenously with 10⁵ CFU of S. Typhimurium SL1344, SL1344 atp, SL1344 F₀ or SL1344 F₁. Bacterial numbers in the spleens (A) and livers (B) were enumerated. Data are presented as mean log₁₀ CFU ± SEM (n = 4), representative of two experiments giving similar results. (+) SL1344 infected mice were not included on day 7 as mice would not survive to this time point. *P ≤ 0.05 compared to SL1344.

Fig. 2

Protection following intravenous vaccination with various atp mutants or SL3261. Female BALB/c mice were immunised intravenously with 10⁵ CFU of SL1344 atp, SL1344 F₀, SL1344 F₁ or SL3261. After 13 weeks, following clearance of the immunising strains, immunised and age-matched unimmunised mice were challenged intravenously with 10⁴ CFU of SL1344. Bacterial numbers in the spleens (A) and livers (B) were enumerated. Data are presented as mean log₁₀ CFU ± SEM (n = 4), representative of two experiments giving similar results. (+) Unimmunised mice were not included on day 7 as mice would not survive to this time point in contrast to immunised mice which all did survive until then. *P < 0.05 compared to unimmunised mice.

Fig. 3

Bacterial counts following intravenous infection with SL1344, SL1344 atp, SL1344 atp (malXY atp operon⁺) or SL1344 atp operon (malXY Cm^R). Female BALB/c mice were infected intravenously with 10⁵ CFU of SL1344, SL1344 atp, SL1344 atp (malXY atp operon ⁺) or SL1344 atp (malXY Cm^R). Bacterial numbers in the spleens (A) and livers (B) were enumerated. Data are represented as mean log₁₀ CFU ± SEM (n = 4), representative of two experiments giving similar results. *P ≤ 0.05 significant compared to SL1344.

Fig. 4

Intracellular survival of SL1344 or SL1344 atp in RAW264.7 cells. Intracellular bacteria (A and C) and macrophage survival (B and D) were determined at 3 and 24 h postinfection, with MOIs of I (A and B) and 10 (C and D). Macrophage survival expressed as percentage of survival compared to uninfected cells. Data are represented as mean ± SEM from three experiments each performed in triplicate and giving similar results.

Fig. 5

Bacterial counts in gp91phox knock-out, IFNγR1 knock-out or wild type C57/BL6 mice. Mice were infected intravenously with 10⁵ CFU of SL1344 atp or SL3261. Bacterial numbers in the spleens and livers of gp91phox knock-out or wild type C57/BL6 mice were enumerated on day 4 postinfection (3–5 mice per point). Bacterial counts in the spleens and livers of IFNγR1 knock-out or wild type C57/BL6 mice were enumerated on day 9 postinfection (4–5 mice per point). Data are presented as mean log₁₀ CFU ± SEM. *P < 0.05 knockout mice compared to wild type mice for each bacterial strain.

Fig. 6

Protection following intravenous vaccination with SL1344 atp or SL3261. Female BALB/c mice were immunised intravenously with 10⁵ CFU of SL1344 atp or SL3261. After 13 weeks, following clearance of the immunising strains, immunised and age-matched unimmunised mice were challenged orally with 10⁹ CFU SL1344. Bacterial counts in the spleens (A) and livers (B) were enumerated. Data are presented as mean log₁₀ CFU ± SEM (n = 4), representative of two experiments giving similar results. (+) Unimmunised mice were not included on day 14 as mice would not survive to this time point in contrast to immunised mice which all did survive until then. *P < 0.05 immunised compared to unimmunised mice.

Fig. 7

Protection following oral vaccination with SL1344 atp or SL3261. Female BALB/c mice were immunised orally with 10⁹ CFU of SL1344 atp or SL3261. After 13 weeks, following clearance of the immunising strains, immunised and age-matched unimmunised mice were challenged either intravenously (A and B) with 10⁴ of CFU SL1344 or orally (C and D) with 10⁹ CFU of SL1344. Bacterial counts in spleens and livers were enumerated. Data are presented as mean log₁₀ CFU ± SEM (n = 4), representative of two experiments giving similar results. (+) Unimmunised mice were not included on days 7 and 14 during intravenous challenge and day 14 during oral challenge as mice would not survive to this time point in contrast to immunised mice which all did survive until these times. *P < 0.05 immunised compared to unimmunised mice.

Fig. 8

Anti-Salmonella antibody levels following vaccination with SL1344 atp or SL3261. Pooled sera from 3 to 4 mice taken 11 weeks after intravenous immunisation with 10⁵ CFU or oral immunisation with 10⁹ CFU of SL1344 atp or SL3261. Total Ig levels (A) following intravenous or oral immunisation. IgG subclasses IgG1 (B) and IgG2a (C) assayed following intravenous or oral immunisation. *P ≤ 0.05 comparing SL1344 atp and SL3261 with unimmunised controls.

Fig. 9

Spleen weight and bacterial counts following vaccination with SL1344 atp or SL3261. Female BALB/c mice were immunised intravenously with 10⁵ CFU of SL1344 atp or SL3261 and age-matched unimmunised controls were included for spleen weight comparisons. Spleens were weighted (A) and bacterial counts in spleens (B) and livers (C) were enumerated. Data are presented as mean grams ± SEM for spleen weight or mean log₁₀ CFU ± SEM for bacterial count (n = 4–8). Data are pooled from two experiments giving similar results. ⁺P ≤ 0.05 uninfected compared with both infected groups, *P ≤ 0.05 SL1344 atp infected mice compared to SL3261 infected mice.