Tryptophan biosynthesis protects mycobacteria from CD4 T-cell-mediated killing

Abstract

Bacteria that cause disease rely on their ability to counteract and overcome host defenses. Here, we present a genome-scale study of Mycobacterium tuberculosis (Mtb) that uncovers the bacterial determinants of surviving host immunity, sets of genes we term "counteractomes." Through this analysis, we found that CD4 T cells attempt to contain Mtb growth by starving it of tryptophan--a mechanism that successfully limits infections by Chlamydia and Leishmania, natural tryptophan auxotrophs. Mtb, however, can synthesize tryptophan under stress conditions, and thus, starvation fails as an Mtb-killing mechanism. We then identify a small-molecule inhibitor of Mtb tryptophan synthesis, which converts Mtb into a tryptophan auxotroph and restores the efficacy of a failed host defense. Together, our findings demonstrate that the Mtb immune counteractomes serve as probes of host immunity, uncovering immune-mediated stresses that can be leveraged for therapeutic discovery.

Figure 1

A. MHC Class II KO and wild type mice were infected with 10⁶ bacilli from our transposon library. At 10 days and 45 days after infection, 4 mice in each group were sacrificed and spleen homogenates were plated to recover surviving bacteria. To discover genes required for surviving CD4 T Cell immunity, we searched for genes that were required for growth in a wild type mouse but not required in the MHC Class II mice. For both the d10 (B.) and d45 (C.) timepoints, we calculated for each gene the insertion count differences between the input library and the recovered libraries from wild type mice, and a p-value expressing the significance of this differences. We then log-transformed the p-values and gave each gene with a loss of insertions a negative log-p-value and each gene with a gain of insertions a positive log-p-value. After ordering the genes based on their p-values, we plotted each gene as a dot with log-p-values on the y-axis and the size of the circles representing the fold-change of insertion count differences.

Figure 2

A. To search for genes required for surviving the CD4 T cell response, we identified genes that had a statistically validated increase in insertion counts (above the axis). Transposon insertion counts in the regions containing trpE (B) and trpD (C). Histograms represent the number of times insertions were found at each potential insertion site. Both trpE and trpD sustained insertions in our library, but whereas we were unable to retrieve insertions in these genes from wild type mice, we were able to recover trpE and trpD mutants from MHC Class II KO mice. D. We infected wild type, trpE KO, and complemented strain Mtb into wild type and MHC Class II KO mice. Growth of the three strains was determined, confirming the results of our transposon screen. E. By comparing gene requirement signatures, we profiled the similarity of CD4-mediated stress to in vitro models of potential immune-mediated stresses. Each box represents a pairwise comparison between two gene sets, where the larger gene set was ordered by p-value and ratio and the smaller gene set was used by the GSEA pre-ranked tool to search for enrichment of second set genes in the first

Figure 3

A. We made a tryptophan auxotroph by replacing trpE with a hygromycin resistance cassette. B. Tryptophan biosynthesis starts with the conversion of chorismate to anthranilate by TrpE, followed by ribosylation by TrpD. C. Tryptophan auxotrophs, complemented strains, and wild type Mtb were grown in the presence and absence of tryptophan in 7H9 media. No tryptophan was required for wild type and complement growth, but 1 mM tryptophan was required to restore normal growth of the tryptophan auxotroph. Auxotrophs grown to mid-log (D.) and stationary (E.) phase were washed, starved of tryptophan and plated for CFU. Starvation of the auxotroph resulted in rapid mycobacterial death.

Figure 4

A. Wild type, tryptophan auxotroph, and complemented strains were used to infect macrophages. One day after infection, macrophages were either stimulated with CD4 T cells in co-culture or remained unstimulated. The tryptophan auxotroph grew poorly in unstimulated macrophages and were hypersuscpetible to the effects of CD4 T cells (B). C. Macrophages were stimulated by either immune CD4 T cells (CD4 T cells isolated from spleens of Mtb-infected mice) or IFN-γ and TNF-α. To test whether the hypersusceptibility of the auxotroph was dependent on IDO, we also inhibited IDO with its small molecule inhibitor, 1-MT, or used macrophages isolated from IDO KO mice (D). E. Human monocyte derived macrophages were infected with wild type and tryptophan auxotroph bacteria. After stimulation with IFN-γ, we showed that the tryptophan auxotroph was also hypersusceptible to the effects of human IFN-γ

Figure 5

A. 6-FABA inhibits the growth of Mtb in 7H9. B. To test the bactericidal potential of 6-FABA, wild type Mtb was treated with 6-FABA in 7H9, and cultures were plated for CFU at various time points. C. Wild type (blue) and mutant (black) TrpE (F68I) were isolated and enzymatic activity was assessed by measuring chorismate concentration. D. The inhibitory affect of tryptophan on both wild type (blue) and mutant (black) TrpE was measured as % initial velocity of the no-inhibitor control reaction. E. Mtb TrpE structure (blue) was modeled based on the homologous enzyme from S. macrcescens (beige), showing F68 in the allosteric binding pocket of tryptophan. C. Wild type Mtb was used to infect macrophages. On day 1 after infection, macrophages were treated with 6-FABA, and on day 5, cells were lysed and bacteria plated, showing the 6-FABA was bactericidal in macrophages. D. To test the synergy between 6-FABA and IFN-γ, we added 6-FABA and IFN-γ at concentrations that inhibited approximately half of bacterial growth. E. Together, their effect was greater than the estimated additive effect of the two molecules in the absence of synergy, showing that they indeed work synergistically to kill Mtb in macrophages.

Figure 6

A. Wild type Mtb was used to infect macrophages. On day 1 after infection, macrophages were treated with 6-FABA, and on day 5, cells were lysed and bacteria plated, showing the 6-FABA was bactericidal in macrophages. B. To test the synergy between 6-FABA and IFN-γ, we added 6-FABA and IFN-γ at concentrations that inhibited approximately half of bacterial growth. C. Together, their effect was greater than the estimated additive effect of the two molecules in the absence of synergy, showing that they indeed work synergistically to kill Mtb in macrophages. Mice were infected with 10² aerosolized Mtb bacilli. After 8 days of infection, mice were treated with INH (25 mg/kg), 6-FABA (200 mg/kg) or with the ester derivative of 6-FABA (200 mg/kg). At 2 weeks (D and E) and 4 weeks (F and G) after infection, lungs and spleens were homogenized and plated for CFU. *: P-val < 0.05. **: P-val < 0.01. Insets: linear scale CFU.

Figure 7

A. Upon infection, unstimulated macrophages have a lot amount of tryptophan, which together with tryptophan synthesized by the bacterium is able to support Mtb growth. B. CD4 T cells, through IFN-γ and the induction of IDO, decreases the amount of intracellular tryptophan available to Mtb, demanding the need for mycobacterial tryptophan biosynthesis for bacterial survival. C. Treatment with 6-FABA chemically induces Mtb tryptophan auxotrophy. Together with immune-mediated tryptophan starvation, this results in mycobacterial death.