Identification of MrtAB, an ABC transporter specifically required for Yersinia pseudotuberculosis to colonize the mesenteric lymph nodes

Abstract

A highly conserved virulence plasmid encoding a type III secretion system is shared by the three Yersinia species most pathogenic for mammals. Although factors encoded on this plasmid enhance the ability of Yersinia to thrive in their mammalian hosts, the loss of this virulence plasmid does not eliminate growth or survival in host organs. Most notably, yields of viable plasmid-deficient Yersinia pseudotuberculosis (Yptb) are indistinguishable from wild-type Yptb within mesenteric lymph nodes. To identify chromosomal virulence factors that allow for plasmid-independent survival during systemic infection of mice, we generated transposon insertions in plasmid-deficient Yptb, and screened a library having over 20,000 sequence-identified insertions. Among the previously uncharacterized loci, insertions in mrtAB, an operon encoding an ABC family transporter, had the most profound phenotype in a plasmid-deficient background. The absence of MrtAB, however, had no effect on growth in the liver and spleen of a wild type strain having an intact virulence plasmid, but caused a severe defect in colonization of the mesenteric lymph nodes. Although this result is consistent with lack of expression of the type III secretion system by Wt Yptb in the mesenteric lymph nodes, a reporter for YopE indicated that expression of the system was robust. We demonstrate that the ATPase activity of MrtB is required for growth in mice, indicating that transport activity is required for virulence. Indeed, MrtAB appears to function as an efflux pump, as the ATPase activity enhances resistance to ethidium bromide while increasing sensitivity to pyocyanin, consistent with export across the inner membrane.

Figure 1. Plasmid deficient Yersinia pseudotuberculosis grow and persist in mouse deep tissue sites with little clonal loss.

A) Growth within the spleen (blue diamond) and liver (red square) after IV inoculation of Yptb(P⁻) in C57BL/6 mice. 1×10⁵ Yptb(P⁻) were IV inoculated, organs were collected 3 days post-infection, and bacterial numbers were determined by colony forming units (CFU) per organ. N = 3–6 mice, mean CFU is plotted, ± standard deviation. B) Average number of unique transposon insertion clones in the Input library, and per organ over 6 days post IV infection, +/− standard deviation, N = 3.

Figure 2. Genetic screen for chromosomal Yptb virulence factors.

A) Number of genes that were mutated in each input library, and the number of genes mutated in both libraries. Red = library #1, Blue = library #2. B) Quality control of 2 biological replicates (BR) of the Input Library #2, sequenced separately. Gene number is on the X axis (Yptb has 4250 genes), and log₂(ratio of each gene in BR 1/BR 2) is on the Y axis. Dashed line = 1 standard deviation, solid line = 2 standard deviations C) Histogram of 1977 genes mutated in both library #1 and library #2, X axis = log₂(Average ratio of Liver Output/Input), Y axis = number of genes that have a threshold X value. The X axis extends to include values for all genes.

Figure 3. mrtAB is required for growth of Yptb (P⁻) in liver and spleen.

A) In frame deletion of mrtAB in Yptb(P−) recapitulates the data from the screen. Mice were inoculated IV with 1×10⁵ bacteria, organs were collected 3 days post-infection, and bacterial number was determined by colony forming units (CFU) per organ. N = 4–6 mice. B) Deletion of mrtAB does not alter growth at 37° in 2XYT broth culture. Data are mean of 3 replicates, error bars = ± standard deviation. C) Rescue of Yptb(P⁻) ΔmrtAB in trans with pmrtAB. Mice were injected IV with 1×10⁵ Yptb (P⁻)/Vector, Yptb (P⁻)ΔmrtAB/Vector, or Yptb (P⁻) ΔmrtAB/pmrtAB, spleens were collected 3 days post-infection and analyzed as in A. N = 5 mice. D) Growth curve of Yptb (P⁻) and Yptb (P⁻) ΔmrtAB in liver and spleen over 3 days. Mice were injected IV with 1×10⁵ Yptb (P⁻) and Yptb (P⁻) ΔmrtAB, organs were collected between 4 hours and 3 days post-infection, and analyzed as in A. N = 3 mice, +/− standard deviation. *Statistical significance (P*) in Figure 3 was determined by nonparametric Mann–Whitney test.

Figure 4. In Wt Yptb, mrtAB is only required in the mesenteric lymph node.

A) Growth of Yptb (P⁺) in spleen and liver is unaffected by the absence of mrtAB. Mice were inoculated IV with 10³ Yptb (P⁺) derivatives, organs were collected 3 days post-infection, and bacterial number was determined by colony forming units (CFU) per organ. N = 4 or 6. B) Yptb (P⁺) requires MrtAB for optimal colonization of mesenteric lymph nodes. Mice were orally inoculated with 2×10⁹ Yptb(P⁺) or Yptb(P⁺)ΔmrtAB, organs were collected at 1 day post-infection, and analyzed as in A. The dashed line indicates the limit of detection. N = 10 mice. C) Defect in MLN colonization caused by absence of MrtAB is rescued in trans by intact mrtAB. Mice were orally inoculated with 2×10⁹ Yptb(P⁺)/vector, Yptb(P⁺)ΔmrtAB/vector, or Yptb(P⁺)ΔmrtAB/pmrtAB, organs were collected 1 day post-infection and analyzed as in A. N = 8 or 9 mice. D) The defect in MLN colonization caused by the absence of MrtAB can be recapitulated after intraperitoneal inoculation. IP inoculation was performed with 2×10⁵ Yptb(P⁺)or Yptb(P⁺)ΔmrtAB Yptb. Organs were collected 1 day post-infection, and analyzed as in A. N = 4 mice. *Statistical significance was determined by nonparametric Mann–Whitney test.

Figure 5. ATPase activity of MrtB is required for optimal growth in vivo.

A) Mice were infected IV with 1×10⁵ Yptb(P⁻)/vector, Yptb(P⁻)ΔmrtAB/vector, Yptb(P⁻)ΔmrtAB/pmrtA ⁺ mrtB ⁺-flag complementation vector, or Yptb(P⁻)ΔmrtAB/pmrtA ⁺ mrtB*-flag complementation vector with *K380A mutation in MrtB. Spleens were collected 3 days post infection, and bacterial number was determined by colony forming units (CFU) per organ. N = 6 or 8 mice. B) Bacteria were grown in vitro to examine the effect of disrupting the MrtB-FLAG Walker A box on MrtB-FLAG expression. Yptb(P⁻) ΔmrtAB/vector (Lane 1 = 26°C, Lane 2 = 37°C), Yptb(P⁻)ΔmrtAB/pmrtA ⁺ mrtB ⁺ -flag (Lane 3 = 26°C, Lane 4 = 37°C), or Yptb(P⁻)ΔmrtAB/pmrtA ⁺ mrtB*-flag (*K380A) (Lane 6 = 26°C, Lane 7 = 37°C), were grown in LB in vitro at 26° or 37°. Blots were stripped and re-probed with S2 antibody for a loading control. *P: Statistical significance was determined by nonparametric Mann–Whitney test.

Figure 6. MLN-localized Yptb(P⁺) expresses YopE and is in contact with neutrophils.

Mice were orally inoculated with 2×10⁹ (MLN) or injected IV with 10³ (Spleen) bacteria to approximately synchronize the infections, organs were collected 2 days post-inoculation, and tissue sections were stained for DNA (Hoechst). Displayed are representative images of Yptb(P⁺)-GFP/YopE-mCherry micro-colonies in the spleen (A) or MLN (B). (Median mCherry fluorescence intensity)/(median GFP fluorescence intensity) for Yptb(P⁺)/GFP/YopE-mCherry micro-colonies, or Yptb(P⁺)/GFP micro-colonies in the spleens and MLNs, is graphed in C. Each symbol in C represents a different focus of infection. N = 3 (GFP control) or N = 7–8 (Reporter). D) Spleens from mice infected IV with 1×10³ GFP-Wt Yptb, or E) MLN from mice orally inoculated with 2×10⁹ GFP-Wt Yptb, were isolated 2 days post-infection, and tissue sections were stained for neutrophils (Ly6G) and DNA (Hoechst). D and E are representative images.

Figure 7. The Yptb(P+) mrtAB mutant exhibits delayed growth in the MLN, but equal spleen colonization and lethality in mice.

A) MrtAB deficient Yptb(P+) exhibit a modest decrease on MLN colonization at 4 days post-infection. Mice were orally inoculated with 2×10⁹ Yptb(P⁺) or Yptb(P⁺)ΔmrtAB, and the small intestines, MLNs, and PPs were collected at 4 days post-infection, and bacterial number was determined by colony forming units (CFU) per organ. N = 4 mice (PP) or 9 mice (SI and MLN). *P: Statistical significance was determined by nonparametric Mann–Whitney test. B) Yptb(P+) and Yptb(P⁺)ΔmrtAB display equal early colonization of the spleen following oral infection. Mice were orally inoculated with 2×10⁹ Yptb(P⁺) or Yptb(P⁺)ΔmrtAB, and the small intestines and spleens were collected at 2 days post-infection, and analyzed as in A. N = 5 mice. C) Yptb(P+) and Yptb(P⁺)ΔmrtAB are equally lethal following oral infection with 10⁹ bacteria. Mice were orally inoculated with 10⁹ Yptb(P⁺) or Yptb(P⁺)ΔmrtAB, and the survival of infected mice was monitored over 12 days. N = 8 mice.