How Chlamydia trachomatis conquered gut microbiome-derived antimicrobial compounds and found a new home in the eye

Different variants of the obligate intracellular bacterial pathogen Chlamydia trachomatis cause the diseases trachoma and chlamydia. The trachoma strains cause chronic infections of the conjunctival epithelium and an intense inflammatory response that can lead to corneal damage, and trachoma is the most common cause of infectious blindness. The Chlamydia strains preferentially infect columnar epithelial cells of the cervix or urethra, and Chlamydia is the most common bacterial sexually transmitted infection. Chlamydia strains often elicit no symptoms but sometimes ascend the reproductive tract where they can elicit chronic inflammation and severe consequences, including pelvic inflammatory disease. Despite their distinct tissue tropisms, the genomes of the trachoma and Chlamydia strains are nearly identical (1). Chlamydia genital tropism has been linked to tryptophan synthase (TS), an enzyme that synthesizes tryptophan from indole (Fig. 1) (2, 3). Expression of the α-subunit (TrpA) and β-subunit (TrpB) of the TS holoenzyme is tightly regulated by an apo-repressor (TrpR) (4–7). In the presence of the corepressor tryptophan, TrpR binds an operator sequence (trpO) and blocks trpBA transcription. Low tryptophan levels promote TrpR release from trpO and allow TS expression. Chlamydia strains likely use TS to circumvent immune-regulated tryptophan catabolism, mediated by the host enzyme indoleamine 2,3′ dioxygenase (IDO1) (2). In contrast, most trachoma isolates do not express a functional TS (2). A study in PNAS by Sherchand and Aiyar (8) presents compelling evidence that the gut microbiome-derived indole metabolites indole-3-propionic acid (IPA) and indole-β-acrylic acid (IAA) compete with tryptophan for binding chlamydial TrpR and block TrpR binding of trpO (Fig. 1). This permits the expression of TS in tryptophan-replete conditions and the production of the bactericidal ammonia from serine by an alternate β-replacement reaction of TS. Beyond C. trachomatis, these observations may be relevant to the pathogenesis of a variety of important intracellular human pathogens.

Fig. 1.

De-repression of trpBA operon by indole derivatives in tryptophan-replete conditions. (A) Transcription of trpBA is blocked by TrpR bound to tryptophan corepressor. (B) In tryptophan-deficient conditions, indole is salvaged by tryptophan synthase holoenzyme to produce tryptophan. (C) Indole metabolites (IPA, IAA) replace tryptophan bound to TrpR, causing operon de-repression of trpBA expression in the absence of indole, leading to generation of ammonia from serine by an alternate β-elimination reaction. P, promoter.

Neither C. trachomatis nor human host cells can synthesize tryptophan de novo from chorismate (3). In human epithelial cells, C. trachomatis can transition into a viable, but not cultivable and reversible, state (persistence) when deprived of tryptophan or exposed to some inflammatory cytokines, including IFN-γ (9). Byrne and colleagues showed that IFN-γ–induced persistence requires IDO1 and can be reversed by excess tryptophan (10). Chlamydia isolates encode a highly conserved trpBA operon, which expresses a functional TS, whereas most trachoma isolates encode mutant trpBA operons that produce a truncated TrpA that cannot form a functional TS with TrpB (2). Chlamydia TS can synthesize tryptophan from serine and indole in Escherichia coli but cannot synthesize tryptophan from serine and indole-3-glycerol-phosphate like orthologous TS enzymes of some other bacteria (3). Indole can also rescue Chlamydia strains, but not trachoma strains, from IFN-γ– and tryptophan-induced persistence (2, 11), and genetic knockout and complementation have confirmed that Chlamydia TS is necessary (12) and sufficient (13) for indole rescue of C. trachomatis from IFN-γ–induced persistence.

Chlamydia likely uses TS to synthesize tryptophan inside the chlamydial vacuole, which is inaccessible to IDO1, using indole scavenged from other genital microorganisms (2, 3). Consistent with this hypothesis, Chlamydia prevalence is higher in women with bacterial vaginosis (14), a syndrome associated with overgrowth of indole-producing bacterial taxa in the vagina (15). Trachoma and Chlamydia strains diverged relatively recently (16), and why TS was inactivated in the trachoma strains is unclear. One hypothesis is that TS decayed in the trachoma strains because indole is unavailable in the ocular environment and TS provides no fitness advantage there (17). Alternately, ammonia produced by an alternate β-elimination reaction of TS may have selected against the retention of this enzyme by the trachoma strains (17).

Based on protein structure models that suggested that IPA and IAA could compete with tryptophan for binding to TrpR and that IPA- and IAA-bound TrpR proteins would be unable to bind trpO (Fig. 1), Sherchand and Aiyar (8) tested the effects of these compounds on TS expression using a representative Chlamydia isolate (TS+ Chlamydia). IPA and IAA de-repressed trpBA expression and compromised the growth and progeny production of TS+ Chlamydia in HeLa cells. In contrast, these compounds had no effect on the growth of a Chlamydia trpB nonsense mutant (TS− Chlamydia) or a trachoma isolate. Indole blocked the toxicity of IPA and IAA for TS+ Chlamydia, suggesting that the antichlamydial effects of these molecules were caused by the production of ammonia by TS. Consistent with this hypothesis, addition of IPA or IAA increased ammonia levels in the supernatants of cells infected with TS+ Chlamydia, but not in the supernatants of uninfected cells or in cells infected with TS− Chlamydia. Addition of ammonia also phenocopied the bactericidal effects of IPA and IAA on Chlamydia. Antibacterial effects of IPA and IAA have been described previously, but mostly at doses much higher than those normally found in human serum. Critically, Sherchand and Aiyar show that physiologically relevant concentrations of IPA and IAA have potent antichlamydial activities in a hypoxia model in which the oxygen tension is similar to what Chlamydia encounters in human tissues.

Overall, Sherchand and Aiyar’s (8) findings support the hypothesis that trachoma strains acquire inactivating mutations in TS (because this enzyme is detrimental in the ocular environment) but also pose some new questions. For example, how does TS impact the fitness of Chlamydia strains in indole-poor environments? The cervicovaginal microbiomes of many women contain few or no indole-producing taxa (18). Similarly, Chlamydia strains primarily infect the urethra in men, and although indole-producing bacteria sometimes colonize the male urethra, bacterial loads are low there (19). This suggests that other sources of tryptophan or indole may be available to Chlamydia strains in some niches, or that Chlamydia strains may have adaptations that help them counter IPA and IAA when indole is unavailable. Second, many trachoma strains still express functional TrpR and TrpB proteins (2, 3). Are these vestigial factors that impart negligible fitness costs, or were these proteins retained because they have cryptic activities that increase trachoma fitness? Finally, serum levels of IPA and IAA can vary between individuals and over time within an individual. Thus, Sherchand and Aiyar’s (8) results suggest that the composition of an individual’s gut microbiome could be a key determinant of their susceptibility to Chlamydia and other IAA- and/or IPA-sensitive pathogens, including Mycoplasma tuberculosis and Legionella pneumophila.