Competing Substrates for the Bifunctional Diaminopimelic Acid Epimerase/Glutamate Racemase Modulate Peptidoglycan Synthesis in Chlamydia trachomatis

The Chlamydia trachomatis genome encodes multiple bifunctional enzymes, such as DapF, which is capable of both diaminopimelic acid (DAP) epimerase and glutamate racemase activity. Our previous work demonstrated the bifunctional activity of chlamydial DapF in vitro and in a heterologous system (Escherichia coli). In the present study, we employed a substrate competition strategy to demonstrate DapF_Ct function in vivo in C. trachomatis.

ABSTRACT

The Chlamydia trachomatis genome encodes multiple bifunctional enzymes, such as DapF, which is capable of both diaminopimelic acid (DAP) epimerase and glutamate racemase activity. Our previous work demonstrated the bifunctional activity of chlamydial DapF in vitro and in a heterologous system (Escherichia coli). In the present study, we employed a substrate competition strategy to demonstrate DapF_Ct function in vivo in C. trachomatis. We reasoned that, because DapF_Ct utilizes a shared substrate-binding site for both racemase and epimerase activities, only one activity can occur at a time. Therefore, an excess of one substrate relative to another must determine which activity is favored. We show that the addition of excess l-glutamate or meso-DAP (mDAP) to C. trachomatis resulted in 90% reduction in bacterial titers, compared to untreated controls. Excess l-glutamate reduced in vivo synthesis of mDAP by C. trachomatis to undetectable levels, thus confirming that excess racemase substrate led to inhibition of DapF_Ct DAP epimerase activity. We previously showed that expression of dapF_Ct in a murI (racemase) ΔdapF (epimerase) double mutant of E. coli rescues the d-glutamate auxotrophic defect. Addition of excess mDAP inhibited growth of this strain, but overexpression of dapF_Ct allowed the mutant to overcome growth inhibition. These results confirm that DapF_Ct is the primary target of these mDAP and l-glutamate treatments. Our findings demonstrate that suppression of either the glutamate racemase or epimerase activity of DapF compromises the growth of C. trachomatis. Thus, a substrate competition strategy can be a useful tool for in vivo validation of an essential bifunctional enzyme.

INTRODUCTION

Chlamydia trachomatis is an obligate intracellular bacterial pathogen that causes both ocular infections and sexually transmitted infections. C. trachomatis utilizes a biphasic life cycle composed of two distinct developmental forms, the elementary body (EB) and the reticulate body (RB). During the infection cycle, the rigid, infectious, and metabolically inactive EBs initiate infection by adhering to susceptible host cells. Upon internalization, EBs differentiate into larger, metabolically active, and osmotically fragile vegetative RBs. RBs remain protected in a host cell vacuole called an inclusion, where RBs grow and divide. After multiple rounds of cell division (about 48 to 72 h postinfection [hpi]), the RBs differentiate into EBs and exit the host cell via extrusion or host cell lysis to initiate a new cycle of infection (1, 2).

Peptidoglycan is a polymer of glycan chains connected by short peptides that is unique to bacteria. It is composed of alternating N-acetylglucosamine (GlcNAc) and N-acetylmuramic acid (MurNAc) residues linked through a β1,4 linkage (3). A pentapeptide stem consisting of d- and l-amino acids is attached to the lactyl moiety of the MurNAc saccharide. Crosslinking of glycan strands in Gram-negative bacteria is between a meso-diaminopimelic acid (mDAP) in the third position of the pentapeptide stem and a d-alanine in the fourth position on an adjacent stem peptide. The cross-linking imparts strength to the fully assembled peptidoglycan layer and provides protection against osmotic shock and lysis (4).

Chlamydiae lack a classic bacterial cell wall sacculus, but peptidoglycan is present in RBs at the septal division plane (5, 6). One feature of peptidoglycan that makes the enzymes involved in its synthesis and assembly excellent antibiotic targets is the presence in the stem peptide of three unique d-amino acids, i.e., d-alanine, d-glutamate, and d,l-diaminopimelic acid, (DAP) (7–10). The peptidoglycan of Chlamydia contains all three of these d-amino acids in its stem peptide (11).

DapF, an enzyme that epimerizes l,l-DAP into l,d-DAP or mDAP and vice versa, has gained attention as a drug target because mDAP is an essential building block of the bacterial peptidoglycan as well as a precursor to the essential amino acid l-lysine (12). Inhibitors of DapF would have the additional advantage of selective toxicity, since mammals lack this metabolic pathway and require l-lysine in their diet. Although an alternative pathway for replacement of mDAP with cystathionine or lanthionine in the stem peptide exists in Escherichia coli (13, 14), a potent inhibitor of DapF, an N-hydroxy-DAP compound 10 analog in which the amine group is replaced by a hydroxyl group, has been developed (15).

Another potential drug target among the d-amino acid-synthesizing enzymes is MurI, which racemizes the glutamate isomers l- and d-glutamate. While l-glutamate is a normal component of proteins, d-glutamate is uniquely found in the stem peptide of peptidoglycan. MurI is essential in most bacteria (16, 17). However, an alternative enzyme, d-amino acid aminotransferase, can also synthesize d-glutamate in some bacteria (18, 19). Given the crucial role of MurI for bacterial growth, it has been proposed as a narrow-spectrum drug target for antibacterial development (20, 21), with a focus on its dependence on two Cys residues for catalytic activity (22). MurI inhibitors such as pyrazolopyrimidinediones for Helicobacter pylori (23) and the novel cyclic substrate–product analogue (R,S)-1-hydroxy-1-oxo-4-amino-4-carboxyphosphorinane for Fusobacterium nucleatum have been developed for extracellular pathogens (24).

While DapF and MurI may not be ideal drug targets for E. coli and other pathogens due to either their monofunctionality or the synthesis of mDAP and d-glutamate by alternative routes, C. trachomatis has only one bifunctional enzyme, DapF, that performs both glutamate racemase and DAP epimerase reactions (25). A complete understanding of how substrates of a bifunctional enzyme control catalysis could be used to favor one enzymatic activity while suppressing the other enzymatic activity. Inhibitor-based inactivation studies of the bifunctional chlamydial DapF are not feasible because no inhibitors are commercially available. Moreover, because dapF_Ct is likely essential, construction of deletion mutants to observe the effects on C. trachomatis is challenging. We previously demonstrated that DapF is subject to substrate competition between d-glutamate and l,l-DAP in vitro (25), and we hypothesize that the concentration balance between glutamate isomers and DAP isomers dictates the function of DapF_Ct in a similar fashion in vivo. Creation of a substrate imbalance in vivo would be a useful strategy to determine whether forced enzyme monofunctionality in C. trachomatis inhibits bacterial growth.

In the present study, an in vivo substrate competition was established in C. trachomatis, and its consequences on the survival of C. trachomatis were determined. Direct evidence of altered DapF enzymatic activity was supported by measuring the internal pool of either d-glutamate or mDAP. Finally, we showed that DapF was the only target in all treatments because its overexpression overcame growth inhibition. These studies provide a proof of principle to show how dual functioning enzymes can be targeted by substrate competition, an approach that may be applicable to other intracellular pathogens for which genetic and pharmacological tools are limited, to design novel inhibitors.

RESULTS

Growth of C. trachomatis in the presence of excess mDAP results in reduction of growth and titer.

d-Glutamate and mDAP are critical components of the stem pentapeptide of C. trachomatis peptidoglycan, and both are synthesized by the bifunctional DapF_Ct enzyme (25). Because the glutamate racemase activity (l-glutamate ↔ d-glutamate) of DapF_Ct is significantly reduced in vitro in the presence of excess l,l-DAP (the competing DAP epimerase substrate) (25), we wished to test whether the glutamate racemase activity of DapF_Ct could be reduced in vivo by exogenously supplying excess DAP epimerase substrates (mDAP or l,l-DAP). HeLa cells were preloaded with mDAP prior to infection with C. trachomatis because mDAP is taken up very slowly by mammalian cells (26). In principle, entry of excess mDAP into C. trachomatis is expected to favor the epimerization reaction (l,l-DAP ↔ mDAP), thus reducing the glutamate racemase activity of DapF_Ct (Fig. 1A). The resulting depletion of the d-glutamate pool should then cause peptidoglycan synthesis to stall and should lead to the arrest of cell division. Infected cells grown in medium containing exogenous mDAP (40 mM and 100 mM) contained smaller inclusions than untreated controls (Fig. 1B). The titer assay confirmed the inhibitory effect on treated C. trachomatis, as 40 mM mDAP treatment reduced titers by 32% and 100 mM mDAP reduced growth by 92%, compared to untreated samples (Fig. 1C). Both treatments significantly reduced inclusion sizes (Fig. 1D) and, to some extent, bacterial infectivity (Fig. 1E). With 100 mM mDAP, the number of C. trachomatis inclusion-forming units (IFU) was similar to the initial inoculum, indicating little or no bacterial growth. While it is formally possible that excess mDAP prevents Chlamydia adherence or invasion, we are not aware of any evidence in the primary literature that indicates that a surplus of any amino acid blocks either of these steps. Thus, these results strongly suggest that excess mDAP outcompetes glutamate isomers for DapF_Ct in vivo, resulting in a decrease in C. trachomatis growth.

FIG 1.

Treatment with excess mDAP inhibits growth of C. trachomatis. (A) Model of altered DapF_Ct activity favoring mDAP epimerization over glutamate isomer racemization in the presence of excess mDAP. (B) HeLa cells infected with C. trachomatis at an MOI of 0.5 and exposed to excess mDAP were harvested for analysis at 40 hpi. Representative fluorescence microscopy images (magnification, ×400) of chlamydial inclusions (green) stained with anti-chlamydial LPS and HeLa cell nuclei (blue) counterstained with DAPI are shown. Scale bars = 20 μm. (C) Production of infectious progeny was determined via chlamydial titer assays. (D) Sizes of 150 random inclusions per triplicate sample for each treatment were determined using the spline contour tool in the Zeiss Zen Blue software package. (E) Percent infectivity from triplicate samples was calculated by dividing the number of chlamydial inclusions (green) by HeLa cell nuclei (blue) in 10 fields per coverslip and multiplying by 100. ***, Significant difference (P ≤ 0.05) from untreated samples. Error bars show SEMs.

Growth of C. trachomatis in the presence of excess l-glutamate results in a reversible reduction of growth and titer.

Both activities of DapF_Ct (racemase and epimerase) are needed to construct the peptidoglycan stem peptide. We previously demonstrated that an excess of one of the preferred substrates outcompetes the second substrate in vitro (25). Thus, we reasoned that replicating this in vivo would result in a mDAP deficiency, leading to incomplete peptidoglycan stem peptide synthesis and stalled C. trachomatis replication. We added excess l-glutamate to the growth medium as a strategy to favor the glutamate racemase reaction (l- glutamate ↔ d- glutamate) at the expense of the competing epimerization reaction (l,l-DAP ↔ mDAP) (Fig. 2A). Treatment of C. trachomatis-infected HeLa cells for 40 h with various amounts of l-glutamate resulted in smaller than normal inclusions, indicative of reduced C. trachomatis replication (27, 28). Treatment with 75 mM l-glutamate significantly reduced inclusion size (Fig. 2B and D), and both 50 mM and 75 mM l-glutamate significantly lowered Chlamydia infectivity (Fig. 2E). Growth was also measured by titer assays, which showed that 50 mM l-glutamate reduced production of C. trachomatis infectious progeny by 62% and 75 mM l-glutamate reduced IFU production by 91%, compared to untreated samples (Fig. 2C). We conclude that growth reduction caused by the l-glutamate treatment was due to substrate competition-induced inhibition of DapF_Ct epimerase activity leading to insufficient production of mDAP.

FIG 2.

Treatment with excess l-glutamate inhibits growth of C. trachomatis in a reversible manner. (A) Model of altered DapF_Ct activity favoring d-glutamate racemization over DAP epimerization in the presence of excess l-glutamate. (B) HeLa cells infected with C. trachomatis at an MOI of 0.5 and exposed to excess l-glutamate were harvested for analysis at 40 hpi. Representative fluorescence microscopy images (magnification, ×400) of chlamydial inclusions (green) stained with anti-chlamydial LPS and HeLa cell nuclei (blue) counterstained with DAPI are shown. Scale bars = 20 μm. (C) Production of infectious progeny was determined via chlamydial titer assays. (D) Sizes of 150 random inclusions per triplicate sample for each treatment were determined using the spline contour tool in the Zeiss Zen Blue software package. (E) Percent infectivity from triplicate samples was calculated by dividing the number of chlamydial inclusions (green) by HeLa cell nuclei (blue) in 10 fields per coverslip and multiplying by 100. ***, Significant difference (P ≤ 0.05) from untreated samples. (F) At 40 h, cultures were washed once with complete medium and refed with medium either containing or lacking l-glutamate. Samples were incubated an additional 40 h, and titers were examined at 80 h. Error bars show SEMs.

While continuous treatment with l-glutamate for 40 h significantly reduced the growth of C. trachomatis, we sought to determine whether growth could be restored after removal of l-glutamate. C. trachomatis-infected HeLa cells treated with 75 mM l-glutamate for 40 h showed an increase in IFU when l-glutamate treatment was discontinued and the bacteria were incubated for an additional 40 h (total of 80 h). IFU remained suppressed when l-glutamate treatment was maintained for an additional 40 h (total of 80 h) (Fig. 2F). These data confirm that C. trachomatis can resume growth after removal of l-glutamate treatment, whereas growth of C. trachomatis remains suppressed as long as the 75 mM l-glutamate treatment is continued.

When chlamydial peptidoglycan synthesis is interrupted by peptidoglycan-targeting antibiotics such as penicillin, cell division is inhibited but cell growth often continues, resulting in enlarged, aberrant forms of Chlamydia referred to as aberrant bodies (ABs) (29). To determine whether l-glutamate supplementation resulted in AB formation, we imaged infected cells grown in the presence of exogenous l-glutamate for 24 h via structured illumination microscopy. RBs examined from treatment groups appeared to be similar in size and shape to untreated controls (data not shown). AB formation does not appear to occur during l-glutamate supplementation, indicating that this treatment condition reduced the rate of chlamydial cell division rather than halting it entirely.

Production of mDAP is reduced in C. trachomatis treated with excess l-glutamate.

The experiments described above suggest that suppression of C. trachomatis growth by l-glutamate treatment is due to the excess substrate favoring the glutamate racemase activity of DapF to synthesize d-glutamate over the epimerase reaction that yields mDAP. To quantify the amount of mDAP synthesized by DapF_Ct under conditions of l-glutamate excess, lysates of the l-glutamate-treated Chlamydia-infected cells were used as a nutrient source of mDAP for an E. coli ΔdapD reporter strain that requires exogenous mDAP for growth. We established a standard curve for the concentration of mDAP using the growth response of the E. coli ΔdapD mutant in the presence of various amounts of exogenous mDAP (Fig. 3A). These results were used to determine the amount of mDAP present in the lysates of C. trachomatis-infected HeLa cells. Lysates from C. trachomatis-infected cells treated with 50 mM or 75 mM l-glutamate did not produce sufficient mDAP to support growth of the E. coli ΔdapD mutant, whereas lysates of C. trachomatis that were not treated with l-glutamate supported growth of the mutant to an optical density at 600 nm (OD₆₀₀) of 0.5 (Fig. 3B). Based on the standard curve, the untreated C. trachomatis lysates were estimated to contain 2 μg/ml mDAP and the lysates of 50 mM and 75 mM l-glutamate-treated samples contained undetectable (<0.01 μg/ml) amounts of mDAP, insufficient to support growth of the E. coli ΔdapD mutant. Lysates of l-glutamate-treated C. trachomatis did not possess growth inhibitors, since growth of the ΔdapD mutant was rescued by supplementation with 10 μg/ml mDAP (Fig. 3C). These results demonstrate that excess l-glutamate blocks the in vivo epimerization activity of DapF_Ct, leading to reduced mDAP production and reduced chlamydial growth.

FIG 3.

Measurement of mDAP present in lysates of C. trachomatis grown with excess l-glutamate. A ΔdapD reporter strain of E. coli that requires mDAP for growth was used to estimate levels of mDAP in lysates of C. trachomatis-infected cells. (A) Growth of the ΔdapD mutant after 18 h in the presence of different amounts of mDAP. A linear regression equation was used to quantify the amount of mDAP in the lysates of treated C. trachomatis. (B) Amounts of mDAP present in 0 mM, 50 mM, and 75 mM l-glutamate-treated C. trachomatis lysates, as measured by growth of the reporter strain at 37°C for 18 h (P < 0.05). (C) Supplementation of the lysates of 50 mM and 75 mM l-glutamate-treated C. trachomatis with mDAP restored growth of the E. coli ΔdapD mutant.

Excess mDAP targets DapF_Ct in vivo and overexpression of dapF_Ct can overcome growth inhibition.

To demonstrate that growth inhibition of C. trachomatis by treatment with mDAP is due to saturation of DapF_Ct by excess mDAP, we employed ATM1465, an E. coli ΔdapF murI double mutant whose growth is solely dependent on the glutamate racemase activity of the plasmid-encoded dapF_Ct. An E. coli ΔdapF mutant can incorporate l,l-DAP into its peptidoglycan in place of mDAP and form colonies on rich medium (30). Thus, growth of the E. coli ΔdapF murI double mutant is supported by dapF_Ct through its glutamate racemase activity expressed under the control of an arabinose-inducible promoter in strain ATM1465. We reasoned that excess mDAP would be likely to block the glutamate racemase activity of DapF_Ct and the E. coli ΔdapF murI mutant would not grow. On the other hand, overexpression of dapF_Ct would restore growth as more DapF_Ct would titrate out the excess mDAP and shift the equilibrium to reestablish glutamate racemase activity.

As expected, when expression of dapF_Ct was not induced (absence of l-arabinose) or was repressed by the addition of glucose, the E. coli ΔdapF murI mutant failed to grow (Fig. 4A). The mutant grew only when the medium was supplemented with d-glutamate. Induction of dapF_Ct by 0.1%, 0.2%, or 1% l-arabinose supported growth of the mutant in the absence of mDAP (Fig. 4B). However, growth inhibition of the mutant was observed when 1 mM mDAP was added to the medium, and inhibition was reversed by overexpression of dapF_Ct induced by addition of 0.2% or 1% l-arabinose (Fig. 4C). Inhibition of growth was more pronounced with the addition of 10 mM mDAP and could be reversed only by robust induction of the cloned dapF_Ct with 1% l-arabinose (Fig. 4D). Thus, strong overexpression of DapF_Ct outcompeted the growth inhibitory effect caused by excess mDAP.

FIG 4.

Growth response of E. coli ΔdapF murI mutant complemented with dapF_Ct to excess mDAP and reversal of competitive inhibition by overexpression of dapF_Ct. (A) Growth of E. coli ΔdapF murI complemented with dapF_Ct in the presence of d-glutamate, under glucose repression of dapF_Ct, and no induction of dapF_Ct in the absence of l-arabinose. (B to D) Overexpression of dapF_Ct was induced by 0.1%, 0.2%, and 1% l-arabinose and growth was measured in the absence of mDAP (B), in the presence of 1 mM mDAP (C), and in the presence of 10 mM mDAP (D). All experiments were performed three times, and the error bars represent the standard deviations among experiments performed three times.

Cytotoxicity to HeLa cells of l-glutamate and mDAP treatments.

To observe any cytotoxic effect of l-glutamate and mDAP treatments on the host HeLa cells, the 3-(4,5-dimethyithiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was performed as described previously (28). l-Glutamate at up to 75 mM did not show any cytotoxicity to HeLa cells (P > 0.05) (Fig. 5A). At a concentration of 100 mM, mDAP began to show cell cytotoxicity (∼20 to 30%) (Fig. 5B); however, this effect does not appear to be large enough to explain the levels of chlamydial growth reduction we observed (Fig. 1E).

FIG 5.

Cell viability assay to evaluate the toxicity of excess l-glutamate and mDAP in HeLa cells. (A) MTT assay with different amounts of l-glutamate (P > 0.05). (B) Viability of HeLa cells after treatment with mDAP (P < 0.05). The error bars show SEMs.

DISCUSSION

Peptidoglycan is unique to bacteria and moss chloroplasts, and its interweaving and cross-linked glycan strands provide the strength to withstand internal turgor pressure and external osmotic pressure while acting as a division plane to generate two daughter cells (31, 32). The enzymes that synthesize d-Glu, mDAP, and d-Ala, the unique amino acids of the pentapeptide, are long-established drug targets in pathogens. However, multiple enzymes or pathways can synthesize many of the unique amino acids of the pentapeptide. For example, d-glutamate is typically synthesized by a glutamate racemase (MurI); however, its synthesis via a d-amino acid aminotransferase is also possible (19, 33). Similarly, d-alanine is primarily synthesized by alanine racemase (Alr); however, a serine hydroxymethyltransferase encoded by glyA can also synthesize it (34). C. trachomatis is unique in that DapF is the prime enzyme for synthesis of both d-glutamate and mDAP and no alternative pathways exist (25). Although inhibitors of MurI and DapF have been developed (24, 35, 36), the rationale that was used to create those inhibitors is not applicable to target the bifunctional DapF of C. trachomatis. For example, the MurI inhibitors are based on the two cysteine residues that are critical for its racemase activity. In contrast, the glutamate racemase activity of DapF_Ct is independent of those cysteine residues, because DapF is a pyridoxal phosphate (PLP)-dependent enzyme (19). Similarly, two cysteine residues are the basis for the epimerase activity of a typical DapF (35), and the corresponding cysteine residues in DapF of C. trachomatis are not required due to its PLP dependence. Because of these differences in mechanism, standard inhibitors are unlikely to inhibit the glutamate racemase and epimerase activity of chlamydial DapF. More importantly, commercial unavailability and the requirement that these compounds cross eukaryotic membranes to reach an intracellular pathogen hamper such studies in C. trachomatis. While a gene deletion of dapF_Ct is not feasible due to its dual essentiality, substrate competition is a plausible strategy to block one of its functions.

The use of ethanol to mitigate the toxic effects of methanol and ethylene glycol ingestion is an example of the competitive substrate inhibition strategy in the medical field. Ethanol acts as a competitor for alcohol dehydrogenase and thus blocks the metabolism of methanol and ethylene glycol to their toxic byproducts (37). Similarly, substrate analogs that have low toxicity to the host but have similarity to a natural substrate for a critical pathogen enzyme have been used in antibacterial, antiviral, and antitumor therapeutics (38–40). In modern antibacterial chemotherapy, sulfanilamide acts as a structural analog of p-aminobenzoic acid (PABA) to inhibit bacterial growth (41). Because d-alanine is generated from l-alanine by racemization, a structural analog of l-alanine, l-fluoroalanine, has been tested as an antibacterial drug to weaken peptidoglycan of extracellular bacteria (42). In the current study, we used the two characterized substrates of chlamydial DapF to carry out in vivo competition studies to inhibit bacterial growth.

We used excess l-glutamate to outcompete mDAP (to shut off the epimerase activity) in vivo, which forced DapF_Ct to be phenotypically monofunctional. The consequences of such monofunctionality (glutamate racemase activity only) negatively affected the growth of C. trachomatis. The inclusion formation titer, a measure of Chlamydia growth (27, 28), was reduced by 90%. Because very little or no mDAP synthesis was expected with the excess l-glutamate treatment, we used a highly efficient biological reporter system (43) (a mDAP auxotroph of E. coli) that can detect <10 ng/ml biologically available mDAP. The nearly complete absence of intracellular mDAP (<10 ng/ml) during this treatment further demonstrates that monofunctionality of DapF_Ct is achievable and that mDAP synthesis is crucial for normal growth of C. trachomatis. Our data indicated that mDAP supplementation may affect host cells and thus indirectly affect Chlamydia infectivity and growth. However, no such cytotoxic effects were observed for even the highest l-glutamate treatment condition. These data more firmly support our conclusion that the observed excess amino acid supplementation effects on infectivity and EB propagation are direct, and they further indicate that growth cessation of C. trachomatis with excess l-glutamate is primarily due to the absence of DapF epimerase activity (resulting in mDAP deficiency). Inhibition of C. trachomatis growth in the absence of mDAP is similar to findings observed in other bacteria that do not divide when they cannot synthesize sufficient mDAP (44). Altogether, these observations corroborate that the epimerase activity is required to synthesize sufficient mDAP for proper synthesis and cross-linking of peptidoglycan and lack of this activity prevents the growth of C. trachomatis.

While both excess l-glutamate and excess mDAP suppressed the growth of C. trachomatis, additional experiments provide evidence that DapF is the primary target of these treatments. For this, we employed an E. coli mutant whose growth is dependent on the glutamate racemase activity of DapF_Ct, whose expression is controlled by l-arabinose. High concentrations of mDAP, which promoted only DapF_Ct epimerase activity, suppressed the glutamate racemase activity and inhibited growth of the E. coli mutant. Growth was restored when dapF_Ct gene expression was increased by the addition of 1% l-arabinose. This establishes that mDAP primarily blocks the glutamate racemase activity and the block can be overcome by production of more DapF_Ct.

When exposed to β-lactam antibiotics and fosmidomycin, C. trachomatis is known to enter a persistent state that is characterized by the arrest of cell division, resulting in enlarged ABs that can later reenter the normal developmental cycle and generate infectious EBs (29). We determined that, upon removal of exogenous l-glutamate, C. trachomatis development was effectively restored to levels comparable to those of untreated controls, suggesting that a short-term deficiency of mDAP is not lethal to the pathogen. Interestingly, we did not observe AB formation in any of our treatment groups, suggesting that the effects on cell division, while present, are not as robust as those resulting from the inhibition of peptidoglycan cross-linking via β-lactam antibiotics.

In conclusion, this study demonstrates that the DapF of C. trachomatis is critical for normal intracellular functioning and compromising one of its in vivo enzymatic activities severely alters peptidoglycan cross-linking, pausing growth of C. trachomatis. The substrate competition strategy served as a novel method to demonstrate the significance of DapF in vivo. Further studies can use this approach to identify new inhibitors that could force DapF_Ct to behave as a monofunctional enzyme.

MATERIALS AND METHODS

Bacterial strains, media, and culturing conditions.

C. trachomatis L2 434/Bu was provided by Harlan Caldwell (Rocky Mountain Laboratories, Hamilton, MT). E. coli ΔdapD (ATM759) (43) was used as a reporter to measure the amount of mDAP in lysates of C. trachomatis treated with and without l-glutamate. ATM759 was pregrown at 37°C overnight in LB supplemented with 100 μg/ml mDAP. ATM1465, an E. coli ΔdapF murI double mutant transformed with pBAD18::dapF_Ct (25), was pregrown at 37°C overnight in LB supplemented with 100 μg/ml d-glutamate and 100 μg/ml thymine. l-Arabinose was used as an inducer, and glucose was used as a repressor. mDAP and l-glutamate were purchased from Sigma. HeLa cells were routinely cultured in Dulbecco’s modified Eagle’s medium (DMEM) (catalog number 1210046; Gibco) with 10% heat-inactivated fetal bovine serum (FBS) in appropriate tissue culture flasks.

Blocking DAP epimerase and glutamate racemase functions of chlamydial DapF in vivo in the presence of excess l-glutamate and mDAP, respectively.

HeLa cells were plated at 1.3 × 10⁵ cells/ml in triplicate wells of 24-well tissue culture plates containing 1 ml DMEM plus 10% FBS, with or without glass coverslips, for use in immunofluorescence and titer (IFU) assays, respectively. After 24 h of incubation, monolayers were infected at a multiplicity of infection (MOI) of 0.5 with C. trachomatis L2 in 100 μl sucrose-phosphate-glutamic acid buffer per well. Cultures were incubated at 37°C in 5% CO₂ for 2 h, with agitation every 30 min, to allow for adsorption. After infection, at time zero, inocula were removed and C. trachomatis-infected HeLa cells were incubated with fresh DMEM plus 10% FBS containing various amounts of mDAP and 1 μg/ml cycloheximide. To reduce the in vivo glutamate racemase activity of DapF_Ct and since the solubility of mDAP in water is low, mDAP powder was dissolved directly in DMEM and the pH was adjusted to 7.0 prior to use. Additionally, because of slow transport of mDAP, HeLa cells were preloaded with mDAP for 12 h prior to infection with C. trachomatis. mDAP treatments were maintained for 40 h in tissue culture plates with or without glass coverslips, for use in immunofluorescence and titer assays, respectively.

To impair the in vivo DAP epimerase activity of DapF_Ct, C. trachomatis-infected HeLa cells, as described above, were refed with DMEM plus 10% FBS containing different concentrations of l-glutamate and 1 μg/ml cycloheximide for the immunofluorescence and titer assays. At 40 hpi, cultures seeded on coverslips were fixed with 0.5 ml cold methanol for 20 to 30 min and stored in wash buffer for staining. After 40 h, infected monolayers lacking coverslips were scraped using sterile 200-μl large-orifice pipette tips, collected in the preexisting 1 ml of culture medium, and immediately frozen at −80°C for later use in titer assays.

Chlamydia immunofluorescence, structured illumination, and titer assays.

For immunofluorescence assays, Pathfinder Chlamydia lipopolysaccharide (LPS) stain (Bio-Rad, Hercules, CA) was used to stain fixed coverslips to visualize chlamydial inclusions and counterstained with 4′,6-diamidino-2-phenylindole (DAPI) to visualize host cell nuclei. Each coverslip was photographed under ×400 magnification using a Z1 Axiovert Observer epifluorescence microscope and the accompanying Zen Blue software (Zeiss, Oberkochen, Germany). Percent infectivity and inclusion size were determined at 40 h hpi, as described previously (27). For each treatment, percent infectivity was averaged from 10 fields per coverslip. For inclusion size, 10 inclusions were measured in 5 fields per coverslip using the spline contour tool in the Zen Blue software package. To determine the production of infectious progeny via titer assay, samples were thawed, vortex-mixed, sonicated, and then serially diluted for infection of HeLa cell monolayers. After 40 h, coverslips were fixed with methanol and stained with Pathfinder stain to visualize chlamydial inclusions. Inclusions were counted on 2 coverslips per sample using ×200 magnification and a Zeiss Z1 Axiovert Observer inverted light and epifluorescence microscope. Data are expressed as the average IFU per milliliter ± standard error of the mean (SEM). Statistical analyses for percent infectivity, inclusion size, and titer were performed using Student's t test, and P values for each treatment were calculated. Final titer numbers and SEMs were plotted as a bar chart. For examination of RB morphology and size, Chlamydia-infected cells were fixed in 100% methanol at 18 hpi, blocked with 3% bovine serum albumin, and immunolabeled with monoclonal antibody specific for the Chlamydia major outer membrane protein. Cells were subsequently labeled with anti-goat IgG-Alexa Fluor 594, mounted, and imaged with a Zeiss Elyra PS.1 microscope in structured illumination mode.

Estimation of mDAP in lysates of C. trachomatis treated with excess l-glutamate.

To determine the effect of excess l-glutamate in shifting the preference of DapF_Ct away from epimerization (i.e., reduced production of mDAP), lysates of C. trachomatis grown in the presence of excess l-glutamate were used as a nutrient for the mDAP-requiring E. coli ΔdapD mutant. Estimation of mDAP in lysates was based on growth of the ΔdapD mutant, compared to a standard growth curve of the same mutant grown with known concentrations of mDAP. To prepare lysates, l-glutamate-treated C. trachomatis cells were harvested after 40 h and the entire harvested volume (1 ml) was stored at −80°C prior to use. To release cellular contents, including d-glutamate isomers or DAP isomers, harvested C. trachomatis cells were lysed by repeated (six times) sonication for 30 s, with 30-s intervals between pulses, followed by filtration of the lysate through a 0.25-μm syringe filter. Lysates were preserved at 4°C prior to quantification of mDAP. For estimation of mDAP levels, the E. coli ΔdapD mutant was pregrown in LB supplemented with 100 μg/ml mDAP. Bacteria were washed three times with LB to remove residual mDAP, pelleted, and resuspended in 1 ml LB. To measure mDAP present in the l-glutamate-treated and untreated C. trachomatis lysates, 100 μl of each lysate was added to 900 μl of the E. coli ΔdapD mutant resuspended in LB, in duplicate, followed by incubation at 37°C on a shaker incubator for 18 h. OD₆₀₀ values for all samples were plotted as a bar chart demonstrating the estimated amounts of mDAP on the secondary y axis. To rule out the presence of any growth-inhibiting compounds in either the l-glutamate-treated lysates of C. trachomatis or DMEM, 10 μg/ml mDAP was added in experimental tubes and growth was measured using a spectrophotometer. OD₆₀₀ values for all samples were plotted as a bar chart.

C. trachomatis recovery from l-glutamate treatment.

To test the reversibility of growth inhibition caused by l-glutamate treatment, HeLa cells grown for 24 h were infected with C. trachomatis at an MOI of 0.5 in triplicate in three plates for IFU titer. The first plate containing C. trachomatis-infected HeLa cells was incubated for 40 h with complete DMEM containing no l-glutamate or containing 75 mM l-glutamate, followed by harvesting of the C. trachomatis for the titer assay. The second plate was washed once with complete DMEM after 40 h and then refed with complete DMEM containing 75 mM l-glutamate and incubated for an additional 40 h. In the third plate, after 40 h of 75 mM l-glutamate treatment, the wells were similarly washed, and the medium was replaced with complete DMEM lacking l-glutamate. After an additional 40 h, cells were harvested for chlamydial titer assays.

Modulation of glutamate racemase activity of DapF_Ct by varying the amounts of mDAP and l-Ara in the dapF_Ct-complemented E. coli ΔdapF murI mutant.

ATM1465 (E. coli ΔdapF murI transformed with pBAD18::dapF_Ct) was pregrown in LB supplemented with 100 μg/ml d-glutamate and 100 μg/ml thymine. After the bacteria were washed three times with LB to remove residual d-glutamate, pelleted cells were resuspended in 1 ml LB. Two sets of three tubes each containing LB were prepared. No mDAP was added to the first tube, whereas 1 mM mDAP and 10 mM mDAP were added to the second and third tubes of each set, respectively; 0.2% l-arabinose and 1% l-arabinose as inducer were added to all tubes of set 1 and set 2, respectively, for low- and high-level expression of dapF_Ct. No l-arabinose and 1% glucose were used as negative controls to validate the induction by l-arabinose and the repression by glucose. Finally, each tube was inoculated with washed ATM1465 at an initial OD₆₀₀ of 0.02. All tubes were incubated on a shaker incubator at 37°C, and growth was measured every hour for 6 h. The observed OD values were plotted for observation of the growth pattern. All experiments were performed three times.

Cell viability assay.

Cytotoxicity arising from addition of either l-glutamate or mDAP to HeLa cells was assessed with the MTT assay. This assay measures the amount of active NAD(P)H inside HeLa cells as an indication of cytotoxicity. Available NAD(P)H in viable cells reduces the yellow-colored MTT to a purple formazan precipitate, which is further dissolved by HCl followed by OD₅₇₀ measurement. HeLa cells were plated in DMEM plus 10% FBS at a density of 5 × 10⁴ cells/0.2 ml per well in 96-well microtiter plates. For mDAP cytotoxicity measurements, HeLa cells were preincubated for 24 h with 20 mM, 40 mM, or 100 mM mDAP, followed by refeeding with the same medium and incubation at 37°C with 5.0% CO₂ for an additional 40 h. For l-glutamate-related cytotoxicity, 50 mM or 75 mM l-glutamate was added after 24 h of growth of HeLa cells, followed by incubation of plates at 37°C with 5.0% CO₂ for 40 h. After incubation, the medium was replaced with phenol red-free DMEM and the MTT assay was performed as described by the manufacturer (Cayman, Ann Arbor, MI, USA). Viability was calculated in terms of OD₅₇₀ values, and data were plotted as a bar chart. Student’s t test was applied to validate the experimental data.