Enolase Is Implicated in the Emergence of Gonococcal Tolerance to Ceftriaxone

Antibiotic tolerance is associated with antibiotic treatment failure, and molecular mechanisms underlying tolerance are poorly understood. We recently succeeded in inducing tolerance to ceftriaxone (CRO) in an N. gonorrhoeae reference isolate. In a prior in vitro study, six biological replicates of WHO P strains were exposed to CRO (10× the MIC) followed by overnight growth, and tolerance was assessed using a modified Tolerance Disc (T.D.) test. In the current study, we characterized the mutation profile of these CRO-tolerant phenotypes. The whole genome was sequenced from isolates from different replicates and time points. We identified mutations in four genes that may contribute to ceftriaxone tolerance in N. gonorrhoeae, including a mutation in the enolase (eno) gene that arose independently in three lineages.

Antibiotic tolerance or persistence is defined as the ability of a bacterial subpopulation to survive high antibiotic concentrations to which the bacteria are fully susceptible without an increase in minimum inhibitory concentration (MIC) [8,9]. Antibiotic tolerance can contribute to treatment failure and the emergence of antibiotic resistance [10][11][12][13]. Tolerancerelated treatment failure can occur when a subpopulation of bacteria in a dormant state survives the lethal effect of the antibiotic [8,9]. Tolerance to antibiotics has been shown to play a role in the clinical persistence of infections caused by Escherichia coli, Pseudomonas aeruginosa and Mycobacterium tuberculosis [13][14][15]. Tolerance has also been shown to emerge prior to, and facilitate the emergence of, antimicrobial resistance (AMR) in E. coli [16]. Thus, tolerance mutations may pave the way for a rapid subsequent evolution of AMR.
We recently succeeded in inducing tolerance to ceftriaxone (CRO) via intermittent cyclic CRO exposure in N. gonorrhoeae reference isolates (WHO P) [17]. In the current study, using whole-genome sequencing (WGS), we characterized the mutation profile of these CRO-tolerant phenotypes.

Phenotypes
We recently induced tolerance to ceftriaxone (CRO) in N. gonorrhoeae reference isolates (WHO P) [17]. A highly susceptible CRO reference strain, WHO P with a CRO MIC of 0.004 µg/mL, was used [18]. Briefly, a direct colony suspension method was used, wherein Antibiotics 2023, 12, 534 2 of 7 WHO P reference colonies from overnight cultures were suspended in Gonococcal (GC) broth (containing 15 g/L bacto protease peptone, 1 g/L soluble starch, 4 g/L K 2 HPO 4 (174.18 g/mol), 1 g/L KH 2 PO 4 (136.08 g/mol), 5 g/L NaCl (58.44 g/mol), supplemented with 1% BD BBLTM IsoVitaleX,) adjusted to a turbidity of 0.5-1.0 McFarland (McF) and exposed to a high concentration (conc.) of CRO (10× the MIC (0.04 µg/mL)) for a fixed duration (3 h) in a cyclic manner. After each cycle, overnight cultures were suspended in (GC.) broth (50 µL in 5 mL), containing 0.04 µg/mL of final conc. of CRO antibiotic (Merck Life Science, Darmstadt, Germany). After this, antibiotics were removed by washing the cultures twice with 10 min centrifugation (1400× g) and resuspending the pellets (1 mL) in fresh GC. broth (5 mL) and incubated overnight (21 h) at 36 • C in a 6.0% CO 2 incubator. The experiment was carried out in 15 mL falcon tubes placed on a roller-mixer (RM 5, CAT, Staufen, Germany). Following each exposure cycle, the CRO MIC was determined using a CRO gradient E-test ranging from 0.016 to 256 µg/mL (BioMérieux, France). The samples were stored at −80 • C in skim milk containing 30% glycerol. The cyclic CRO interval exposure experiment was carried out for six biological replicates for 7 consecutive days.
Six and two populations were tested with and without the antibiotic, respectively.
To detect the presence of tolerant phenotypes, the CRO-exposed population was analyzed using a modified T.D. test protocol (Figure 1; [19]). Briefly, the samples, after the CRO exposure cycle, were inoculated on BD TM Columbia Agar with 5% Sheep Blood and incubated at 36 • C overnight (18-24 h) in a 6.0% CO 2 incubator. This was followed by a T.D. test wherein 0.008 µg of CRO antibiotic was added to a 6 mm Whatmann ® antibiotic assay disc. After~18 h of incubation, the CRO discs were replaced by nutrient discs, i.e., discs containing G.C. broth, and incubated overnight. Additionally, 10 µL of G.C. medium was added to the same nutrient disc and incubated for an additional night as N. gonorrhoeaetolerant colonies emerged 48 h after adding the nutrient disc. The tolerant colonies were inoculated on fresh blood agar plates and stored in 30% glycerol skim milk at −80 • C.
We recently induced tolerance to ceftriaxone (CRO) in N. gonorrhoeae reference isolates (WHO P) [17]. A highly susceptible CRO reference strain, WHO P with a CRO MIC of 0.004 µg/mL, was used [18]. Briefly, a direct colony suspension method was used, wherein WHO P reference colonies from overnight cultures were suspended in Gonococcal (GC) broth (containing 15 g/L bacto protease peptone, 1 g/L soluble starch, 4 g/L K2HPO4 (174.18 g/mol), 1 g/L KH2PO4 (136.08 g/mol), 5 g/L NaCl (58.44 g/mol), supplemented with 1% BD BBLTM IsoVitaleX,) adjusted to a turbidity of 0.5-1.0 McFarland (McF) and exposed to a high concentration (conc.) of CRO (10× the MIC (0.04 µg/mL)) for a fixed duration (3 h) in a cyclic manner. After each cycle, overnight cultures were suspended in (GC.) broth (50 µL in 5 mL), containing 0.04 µg/mL of final conc. of CRO antibiotic (Merck Life Science, Darmstadt, Germany). After this, antibiotics were removed by washing the cultures twice with 10 min centrifugation (1400× g) and resuspending the pellets (1 mL) in fresh GC. broth (5 mL) and incubated overnight (21 h) at 36 °C in a 6.0% CO2 incubator. The experiment was carried out in 15 mL falcon tubes placed on a rollermixer (RM 5, CAT, Staufen, Germany). Following each exposure cycle, the CRO MIC was determined using a CRO gradient E-test ranging from 0.016 to 256 µg/mL (BioMérieux, France). The samples were stored at −80 °C in skim milk containing 30% glycerol. The cyclic CRO interval exposure experiment was carried out for six biological replicates for 7 consecutive days. Six and two populations were tested with and without the antibiotic, respectively.
To detect the presence of tolerant phenotypes, the CRO-exposed population was analyzed using a modified T.D. test protocol ( Figure 1; [19]). Briefly, the samples, after the CRO exposure cycle, were inoculated on BD TM Columbia Agar with 5% Sheep Blood and incubated at 36 °C overnight (18-24 h) in a 6.0% CO2 incubator. This was followed by a T.D. test wherein 0.008 µg of CRO antibiotic was added to a 6mm Whatmann ® antibiotic assay disc. After ~18 h of incubation, the CRO discs were replaced by nutrient discs, i.e., discs containing G.C. broth, and incubated overnight. Additionally, 10 µL of G.C. medium was added to the same nutrient disc and incubated for an additional night as N. gonorrhoeae-tolerant colonies emerged 48 h after adding the nutrient disc. The tolerant colonies were inoculated on fresh blood agar plates and stored in 30% glycerol skim milk at −80 °C.

Genetic Characterization of Eno and tatC Genes Associated with Ceftriaxone Tolerance in WHO-P and Global Neisseria spp. Collection
The putative SNPs identified in the relevant genes associated with ceftriaxone tolerance were further examined in the global collection comprising the genomes of N. gonorrhoeae (n = 17,871), including WHO-P and commensal Neisseria spp. (n = 1136), whose provenance and metadata are described elsewhere [21].

Discussion
Previously, we reported that tolerance to CRO can be induced in CRO-susceptible (<0.004 µg/mL) N. gonorrhoeae WHO P reference isolate [17]. In the current analysis, we describe, for the first time, the mutations associated with emergence of CRO tolerance in N. gonorrhoeae.
Enolase, a glycolytic enzyme, which catalyzes the conversion of 2-phospho-D-glycerate to phosphoenolpyruvate, is involved in carbon metabolism [22]. It is also a component  Figure 2 and Table 2).
None of the above mutations were observed in the global collection of 17,871 N. gonorrhoeae isolates.
Of note, no tolerant colonies were observed in lineage 1 and all tolerant WHO P colonies were found to have a CRO MIC ≤ 0.008 µg/mL, comparable to the MIC of the control samples. No increase in CRO MIC was observed for either the tolerant or the control isolates.

Discussion
Previously, we reported that tolerance to CRO can be induced in CRO-susceptible (<0.004 µg/mL) N. gonorrhoeae WHO P reference isolate [17]. In the current analysis, we describe, for the first time, the mutations associated with emergence of CRO tolerance in N. gonorrhoeae.
Enolase, a glycolytic enzyme, which catalyzes the conversion of 2-phospho-Dglycerate to phosphoenolpyruvate, is involved in carbon metabolism [22]. It is also a component of the RNA degradosome, which is involved in RNA processing and gene regulation [23,24]. In Pseudomonas aeruginosa, enolase influences tolerance to oxidative stress by affecting the production of ahpB and ahpC in an OxyR-independent manner [25]. In another study, oxidative stress response genes gor and ahpC were found to play a role in antibiotic tolerance of Streptococcus mutans biofilms [26]. In CRO-tolerant N. gonorrhoeae, a mutation in the enolase gene arose independently in three lineages at time point 7, implicating enolase in CRO-tolerance.
Mutations in tatC and edd genes were identified at timepoint 7 in lineage 1. Mutations in tatC and other genes, gltI, hlpA, ruvC, ddlB and ydfI, were found to result in tolerance to tosufloxacin in E. coli [27]. TatC, the primary substrate receptor, is part of the twinarginine translocation (Tat) system needed to transport folded proteins across biological membranes [28]. In Zymomonas mobilis the efficient export of NADP-containing glucosefructose oxidoreductase to the periplasm depends on an intact twin-arginine motif [29] and, thus, although the mutations in tatC were identified only at one time point, these mutations may be implicated in tolerance in N. gonorrhoeae.
We were unable to perform the complementation experiments necessary to prove that the mutations we identified played a role in generating tolerance. Other limitations of our study include the small sample sizes and the fact that we only used one antimicrobial to assess the pathway to tolerance. Future studies could remedy these shortcomings as well as assess if tolerance plays a meaningful role in gonococcal treatment failure and the emergence of AMR, as has been shown for other bacterial species [13].
Our global phylogenetic analysis revealed that none of the isolates had the above mutations, which suggests that these mutations might have a high fitness cost and, therefore, not be seen in the natural population. This also raises the possibility that these putativetolerance-associated mutations are transient. A number of studies have found that transient mutations can act as stepping stones to antibiotic resistance in N. gonorrhoeae and other bacteria [30,31]. Several, in vitro antibiotic exposure experiments have shown that the development of AMR is always preceded by the emergence of tolerance, in which several partial resistance mutations can occur [11,14,16,32,33]. Our findings raise the possibility that transient mutations may, likewise, emerge in response to exposure to certain antimicrobials. These may, in turn, facilitate the emergence of AMR, as has been shown in E. coli [16]. Interestingly, isolates from lineages 3, 4 and 5 with a mutation in the eno gene did not have mutations in tatC and edd genes and vice versa, suggesting that there may be multiple pathways involved in CRO tolerance in N. gonorrhoeae.