Tn-Seq reveals hidden complexity in the utilization of host-derived glutathione in Francisella tularensis

Abstract

Host-derived glutathione (GSH) is an essential source of cysteine for the intracellular pathogen Francisella tularensis. In a comprehensive transposon insertion sequencing screen, we identified several F. tularensis genes that play central and previously unappreciated roles in the utilization of GSH during the growth of the bacterium in macrophages. We show that one of these, a gene we named dptA, encodes a proton-dependent oligopeptide transporter that enables growth of the organism on the dipeptide Cys-Gly, a key breakdown product of GSH generated by the enzyme γ-glutamyltranspeptidase (GGT). Although GGT was thought to be the principal enzyme involved in GSH breakdown in F. tularensis, our screen identified a second enzyme, referred to as ChaC, that is also involved in the utilization of exogenous GSH. However, unlike GGT and DptA, we show that the importance of ChaC in supporting intramacrophage growth extends beyond cysteine acquisition. Taken together, our findings provide a compendium of F. tularensis genes required for intracellular growth and identify new players in the metabolism of GSH that could be attractive targets for therapeutic intervention.

Fig 1. Highly saturated Tn-Seq screen in F. tularensis LVS to identify genes required for intramacrophage growth.

(A) Map of transposon insertions across the F. tularensis genome. Of all the potential transposon insertion sites (TA dinucleotides) genome-wide, 75% contain a transposon insertion. Each line corresponds to one mutant with a single insertion; the line height corresponds to the relative abundance of sequencing reads identified for each individual insertion. Blue inner circle represents genes. (B) EL-ARTIST and Con-ARTIST analysis of Tn-Seq screen reveals genes that are non-essential under any tested condition (lavender), critical in vitro (grey), critical in vitro and in macrophages (maroon), and those that are non-essential in vitro but critical in macrophages (blue). (C) The relative abundance of mutations in each gene (average read count per gene) in the mutant pool used to infect J774A.1 cells (pre-infection) and in mutants recovered after 24 hours of infection. Genes significantly more or less abundant after infection (at least 2-fold changed, with 4 or more informative insertion sites, p < 0.05) are colored in blue. Genes involved in GSH metabolism are highlighted in purple. (D-E) Transposon insertion profiles for the indicated genetic regions from the input library (grey) and 24 hr post infection (green). Line height represents the relative abundance of sequencing reads at that position on a log scale.

Fig 2. DptA is a POT that contributes to GSH utilization by LVS.

(A) Transposon insertion profiles for the genetic region encompassing dptA from the input library (grey) and 24 hr post infection (green). Line height represents the relative abundance of sequencing reads at that position on a log scale. (B) Graphical representation of the predicted topology of DptA with conserved sequences motifs found in POT proteins highlighted (proton binding motif, orange; PTR2 domain 1, green; PTR2 domain 2, pink; peptide binding region, light blue). Sequence logo indicates motifs conserved in the POT family derived from an alignment of representative POT proteins. (C) Quantification of the bacterial burdens in the lungs of mice infected via aerosolization with the indicated strains of LVS. (D) Quantification of the bacterial dissemination to, and burdens within the spleens of mice infected via aerosolization with the indicated strains of LVS. (E-G) OD₆₀₀ measurements of the indicated strains of LVS after 16 hrs of growth in CDM with glutathione (E), γGlu-Cys (F) or Cys-Gly (G). Data in (C-G) are shown as mean ± s.d. Asterisks represent statistically significant differences (Student’s t test; ***p≤0.0005, **p≤0.005, *p≤0.05, ns, not significant).

Fig 3. DptA is the sole Cys-Gly transporter in LVS.

(A-B) Quantification of the intracellular levels of ³H-GSH byproducts after 45 min incubation with the labeled substrate in the indicated strains of LVS (A) or WT LVS with and without CCCP (B). (C) Autoradiograph of SDS-PAGE separated proteins produced by growing the indicated strains of LVS in CDM lacking cysteine supplemented with ³⁵S-Cysteine containing GSH. (D) Coomassie-blue staining of proteins generated as described in (C). (E) Percent of ³H-GSH uptake that was inhibited by addition 100 μM of the indicated di- or tripeptides. Data in (A, B and E) are shown as the mean ± s.d. Asterisks represent statistically significant differences (Student’s t test; ***p≤0.0005, **p≤0.005, ns, not significant).

Fig 4. DptA and GGT are essential for Francisella growth in vivo in the absence of exogenous cysteine.

(A) Model depicting the localization and role of the indicated enzymes and transporters in the metabolism of GSH by F. tularensis. The color of the name text indicates the essentiality of that protein based on our Tn-Seq screen in J774 macrophages; essential for growth in MH broth, red; essential for growth in J774 macrophages, green; non-essential in the tested conditions, grey. (B) Growth of the indicated strains of LVS in J774 cells, normalized to bacterial counts at 2 hrs post infection. J774 cells were either left untreated or treated with 5 mM cysteine prior to and during the infection. Data are shown as the mean ± s.d. Asterisks represent statistically significant differences (Student’s t test; **p≤0.005, *p≤0.05).

Fig 5. Identification of a second GGT-independent pathway in Francisella for GSH catabolism mediated by ChaC.

(A-B) OD₆₀₀ measurements of the indicated strains of U112 after 36 hrs of growth in CDM with GSH (A) and γGlu-Cys (B). Data are shown as the mean ± s.d. (C) GGT activity in the indicated strains of U112 and LVS determined via enzyme activity on the substrate L-γ-glutamyl-p-nitroanilide. Data are shown as the mean ± s.d. (D) Transposon insertion profiles from the Tn-Seq screen in J774 cells for the genetic region encompassing chaC from the input library (grey) and 24hr post infection (green). Line height represents the relative abundance of sequencing reads at that position on a log scale. (E) Growth of the indicated strains of LVS in J774 cells, normalized to bacterial counts at 2 hrs post infection. J774 cells were either left untreated or treated with 5 mM cysteine prior to and during the infection. Data are shown as the mean ± s.d. Asterisks represent statistically significant differences (Student’s t test; ***p≤0.0005, **p≤0.005, ns, not significant).