Discovery of a glutathione utilization pathway in Francisella that shows functional divergence between environmental and pathogenic species

Abstract

Glutathione (GSH) is an abundant metabolite within eukaryotic cells that can act as a signal, a nutrient source, or serve in a redox capacity for intracellular bacterial pathogens. For Francisella, GSH is thought to be a critical in vivo source of cysteine; however, the cellular pathways permitting GSH utilization by Francisella differ between strains and have remained poorly understood. Using genetic screening, we discovered a unique pathway for GSH utilization in Francisella. Whereas prior work suggested GSH catabolism initiates in the periplasm, the pathway we define consists of a major facilitator superfamily (MFS) member that transports intact GSH and a previously unrecognized bacterial cytoplasmic enzyme that catalyzes the first step of GSH degradation. Interestingly, we find that the transporter gene for this pathway is pseudogenized in pathogenic Francisella, explaining phenotypic discrepancies in GSH utilization among Francisella spp. and revealing a critical role for GSH in the environmental niche of these bacteria.

Figure 1.. Tn-seq for discovery of F. novicida genes with synthetic phenotypes during growth on GSH.

(A) Schematic illustrating known and unidentified potential GSH catabolism pathways and their products (Glu, red circle; Cys-Gly, yellow and white circles) in the two genetic backgrounds employed in our screen. The heavy arrow depicted for the wild-type background emphasizes the primary conversion pathway (left, Ggt-mediated) while the dashed arrow indicates the residual GSH cleavage mediated by ChaC in the absence of Ggt (right). (B,C) Results of Tn-seq screen to identify genes required for growth of F. novicida on GSH medium in the wild-type (B) and Δggt backgrounds (C). Genes with the greatest difference in transposon insertion reads between growth in GSH and cysteine media in the Δggt background (purple) and other genes shown previously or in this study to participate in GSH uptake or catabolism (blue) are indicated. (D) Rank order depiction of the strength of the synthetic phenotype for genes important for growth of F. novicida Δggt in GSH medium. Rank order was calculated by dividing the ratio of transposon insertion frequency obtained for each gene during growth on GSH compared to growth on cysteine using the F. novicida Δggt background by the same ratio obtained using the wild-type background. See also Figure S1, Tables S1 and S2.

Figure 2.. NgtA is a major facilitator superfamily protein that transports intact GSH in a Ggt-independent manner.

(A) Normalized growth yield in GSH medium of the indicated F. novicida strains. (B) Quantification of the level of [Glycine-2-³H]-Glutathione (³H-GSH) uptake in the indicated strains of F. novicida after 45 min incubation. (C) Normalized growth yield of the indicated F. novicida strains after 36 hrs in defined medium containing Cys–Gly as a sole source of cysteine. (D) Neighbor-joining phylogeny of proteins from the Pht family of MFS transporters. Colored clades contain sequences identified in the original description of the family or subsequently characterized. Representative proteins from the Chen et al. study or other reports are indicated by their respective clades, and transport substrate are indicated in parentheses when known. Candidate NgtA homologs are shown in purple, and the region of the phylogeny amplified in (E) is indicated (shading). (E) Neighbor-joining phylogeny of NgtA homologs in Francisella and related genera. Species names indicate the source of the protein sequences. Data in (A-C) represent mean ± s.d. Asterisks indicate statistically significant differences (unpaired two-tailed student’s t- test.; *p<0.05, ns, not significant). See also Figure S2 and Table S3.

Figure 3.. CgaA is a cytoplasmic glutamine amidotransferase (GATase) family protein that degrades GSH.

(A) Alignment of the predicted structure of CgaA (orange) and the crystal structure of a characterized class I GATases, P. aeruginosa SpuA (blue, PDB: 7D4R, only one subunit of the SpuA homodimer is shown). The conserved catalytic triad is indicated (numbers correspond to amino acid positions in F. novicida CgaA). (B,C) Normalized 36 hrs growth yields of the indicated strains of F. novicida. (C) Coomassie stained SDS-PAGE analysis of purified CgaA and CgaA^C97A. (D) Glutamate released following 60 min incubation of purified CgaA or CgaA^C97A (1 μM protein) with GSH or Gln (10 mM substrate). Data in (B) (C) and (E) represent means ± s.d. Asterisks represent statistically significant differences (unpaired two-tailed student’s t- test.; *p<0.05, ns, not significant). See also Figure S3.

Figure 4.. NgtA contributes to intramacrophage replication of F. novicida but is mutationally inactivated in pathogenic Francisella strains.

(A) Normalized intracellular growth of the indicated strains of F. novicida in bone marrow-derived murine macrophages (24 hrs post-infection. (B) Bacterial burden in mouse lungs at 48 hrs post intranasal infection with ~100 CFU of the indicated strains of F. novicida. (C) Structural model of F. novicida NgtA highlighting differences with NgtA in other Francisella species. The terminal residues resulting from truncating frameshift mutations (FS) indicated in parentheses. FS-1, location of truncation resulting from frameshift in F. tularensis subsp. holarctica; FS-2, location of truncation resulting from frameshift in F. noatunensis; loop, poorly conserved region with many differences between species; IGS, three-residue deletion found in F. tularensis subsp. tularensis. The first six and the last nine residues in the NgtA structure are trimmed. (D-F) Normalized growth yields in GSH medium of the indicated strains of F. novicida (D,E) or F. tularensis LVS (F) . (G) Schematized phylogeny of Francisella species indicating predicted functionality of NgtA (functional, green; inactivated, solid grey; absent, dashed grey), animal association when known (mammal association indicated by rabbit schematic) and mutations present in ngtA (colors correspond to panel C). The ngtA sequence of P. persica contains many mutations (***) and ngtA appears to have been lost completely from F. endociliophora. Phylogenetic relationships derived from Vallesi et al.. Data shown in (A) (B) and (D-F) represent as means ± s.d. Data points in (A-B) indicate technical replicates from 3 (A) or 4 (B) biological replicates conducted. Asterisks indicate statistically significant differences (A and B, one-way ANOVA with Dunnett’s multiple comparison test comparing mutant strains to wild-type; D-F, unpaired two-tailed student’s t- test; *p<0.05, ns, not significant.) See also Figure S4.

Figure 5.. FupA is a porin required for GSH uptake in F. novicida.

(A) Normalized growth in GSH medium of the indicated strains of F. novicida. (B) Quantification of the level of ³H-GSH uptake in the indicated strains of F. novicida after 45 min incubation. (C) Survival of the indicated strains of F. novicida after incubation of mid-log phase cultures with 1.5 mM H₂O₂ for 30 min or 60 min. Data in (A-C) represent mean ± s.d. Asterisks indicate statistically significant differences (unpaired two-tailed student’s t- test.; *p<0.05, ns, not significant).

Figure 6.. Comprehensive model of GSH transport and catabolism in Francisella.

Both the pathways for import and cytosolic catabolism of GSH discovered in this study (left, green shading) and for periplasmic degradation of GSH and subsequent fate of imported Cys-Gly (right, grey shading) are indicated.