Thiolactone modulators of quorum sensing revealed through library design and screening

2.1. Library Design and Synthesis

We designed a 21-member library of thiolactone AHL analogues, with all of the compounds containing a conserved thiolactone head group and a variable acyl tail (Figure 2C). A brief description of our design process and compound synthesis is provided here. Thiolactones 12 and 15, previously shown to be agonists in LuxR and LasR (see above) and analogs of OHHL and OdDHL, respectively, were included in the library to serve as control compounds.^{28, 29} To complement these two controls, we also included the thiolactone analog (18) of TraR's native ligand, OOHL, along with five other thiolactone analogs (19–23) of native AHLs used by a range of bacteria, including P. aeruginosa (BHL, 19), P. chlororaphis (C6-HL, 20), Pantoea ananatis (C6-HL, 20; C7-HL, 21; C8-HL, 22) Burkholderia cepacia (C8-HL, 22), and Acidithiobacillus ferrooxidans (C12-HL, 23).³⁴ The remaining thiolactone library members were chimeric ligands based on acyl groups that we have previously identified in AHL-based LuxR-type receptor agonists and antagonists (Figure 1B).^{20–22, 35} Thiolactones 30/31 and 32/33 were modeled after AHLs 8 and 9, which are strong antagonists of both LuxR and TraR. Likewise, thiolactones 28/29 were based on AHL 7, which is a moderate antagonist of LuxR. Phenylacetanoyl HL 5 was previously shown to be largely inactive in many LuxR-type receptors,²¹ and we therefore included thiolactone analogs 24/25 of AHL 5 to test whether this inactivity profile would be maintained in thiolactones. The 3-nitro phenylacetanoyl thiolactones 26/27 were based on AHL 6, which is an extremely strong LuxR agonist²⁰ and a moderate LasR antagonist.^{21, 22} To assess the importance of stereochemistry on ligand activity, thiolactones 14, 15, and 24–35 were synthesized in both racemic (DL) and enantiopure (L) form. The L-thiolactone enantiomer was chosen based on several previous studies that have shown that the active enantiomer of native AHL signals is the L-form.^{21, 27} Such an analysis of the stereochemical requirements for thiolactone modulation for LuxR-type proteins is yet to be reported.

The thiolactone derivatives that lacked 3-oxo functionality were synthesized by routine EDC couplings between L-homocysteine thiolactone and various carboxylic acids. (Figure 2A). The remainder of the library was synthesized by reacting Meldrum's acid with the requisite alkyl acid chloride to afford the Meldrum's acid derivative, which was then coupled to L-homocysteine thiolactone (Figure 2B). Racemic thiolactones were made in similar manner from DL-homocysteine thiolactone (See Experimental Section).

2.2. Library Assay Design

Small molecules are usually screened for LuxR-type agonism or antagonism using a bacterial strain containing a reporter gene for a given LuxR-type protein.⁹ These strains typically lack a functional LuxI-type synthase, yet retain the functional LuxR-type receptor. Exogenous native AHL therefore must be added to activate the LuxR system. These strains provide a straightforward way to examine the agonistic and antagonistic activities of non-native ligands (by adding only the compound of interest or the compound in competition with the native AHL ligand (at its EC₅₀ value), respectively).

We utilized four bacterial reporter strains in this study to examine the LuxR-type modulatory activities of the thiolactone library in LasR, LuxR, and TraR. Two strains were selected for the LasR screens: Escherichia coli DH5α (pJN105L + pSC11)³⁶ and P. aeruginosa PA01 MW1 (pUM15).³⁷E. coli DH5α (pJN105L + pSC11) is a heterologous reporter strain containing one plasmid for the LasR gene and a second plasmid containing the promoter region for LasI fused to β-galactosidase (β-gal). LasR activity is read-out using a standard colorimetric assay with ortho-nitrophenyl-β-galactosidase (ONPG) as the substrate for β-gal. The PA01 MW1 (pUM15) strain is a LasR reporter in P. aeruginosa that lacks a functional LasI and contains a plasmid with a LasR responsive promoter for Yellow Fluorescent Protein (YFP), which facilitates straightforward evaluation of LasR activity using fluorescence. Examining the thiolactone library in both of these strains allowed us to study the effects of these compounds on LasR in an isolated system (E. coli) and then in the presence of P. aeruginosa's more complex QS network (including RhlR and QscR, in PAO1). (We note that E. coli and P. aeruginosa have different compound uptake/efflux profiles, and this feature should be taken into account when comparing small molecule screening data between the two strains (see below)).

V. fischeri ESI 114 (Δ-LuxI)³⁸ and A. tumefaciens WCF (pCF372)³⁹ were used to examine the activity of the thiolactone library in LuxR and TraR, respectively. The V. fischeri mutant strain lacks a functioning LuxI synthase, but retains its native lux operon, allowing a quantitative luminescent readout based on LuxR activity. Similarly, A. tumefaciens WCF (pCF372)³⁹ lacks a functioning TraI, yet contains a plasmid with a TraR responsive promoter for the β-gal gene, thereby allowing for direct quantitation of TraR activity.

We used bacteriological assay protocols for small molecule screening that were analogous to those reported in our earlier studies (See Experimental Section). All synthetic compounds were screened at 10 μM in both agonism and antagonism assays in the bacterial reporter strains. Notably, these compound concentrations parallel those used by our laboratory in past studies,^{22, 40} allowing for comparisons to be made between the assay data reported here and this past work. No effects on bacterial growth were observed over the time course of the reporter gene assays (4–16 h).

2.3. Antagonism and Agonism Assays in LasR

The primary antagonism and agonism data for the thiolactone library in the LasR E. coli and P. aeruginosa reporter strains are shown in Table 1. Several intriguing trends in ligand activity are immediately apparent upon analysis of these data. First, all of the thiolactones exhibited slight to moderate agonistic activities in the antagonism assay in the P. aeruginosa LasR reporter strain. Natural AHL thiolactone analogues 12, 14/15, 19, and 20 and phenylacetanoyl thiolactones 24/25, 29, and 32–34 were capable of agonizing LasR by greater than 20% in the antagonism assay, yet only the natural AHL analogs (12, 14/15, 19, and 20) showed appreciable agonistic activity in the corresponding agonism assays in the same strain. Compounds capable of agonizing LasR in both assays can be clearly defined as LasR agonists,²¹ but those that fail to agonize in the agonism assay could operate by a different mechanism (see below). Second, while native AHL thiolactone mimics 19 and 20 exhibited an expected correlation between agonism and antagonism trends in the P. aeruginosa LasR reporter strain, this pattern is not mimicked in the E. coli LasR reporter strain. Conversely, 3-NO₂ thiolactones 26/27 were highly active agonists in the E. coli LasR reporter strain, yet minimally active in the P. aeruginosa reporter strain. This activity profile is the opposite of its parent AHL 6, which is one of the stronger LasR inhibitors known.²¹ We note, however, that 3-NO₂ AHL 6 is also an extremely potent agonist of LuxR.²⁰ Third, the remaining potent LasR agonists identified in the E. coli reporter strain (22 and 34) were also significant LasR antagonists in this strain. On the whole, the only LasR agonists in the thiolactone library capable of agonizing LasR in the agonist and antagonist assays in both reporter strains were thiolactone analogs of natural AHL ligands: OdDHL analog 14/15, OHHL analog 12, and N-dodecanoyl thiolactone 23. These results, while unexpected, corroborate with those previously reported by Passador and co-workers for 15.⁴¹

In contrast to the P. aeruginosa LasR reporter strain in which no antagonists were found, several highly potent LasR antagonists were found using the E. coli reporter strain. The non-3-oxo aliphatic thiolactones 20–22 and phenylacetanoyl thiolactones 24/25 and 28–34 were all found to be ~50–80% inhibitors of LasR this strain. Many of these compounds show complex trends in activity. For example, both the racemic and enantiopure phenylacetanoyl thiolactones 28–31 are good to excellent inhibitors in the E. coli LasR reporter strain, yet are mild agonists in the P. aeruginosa reporter strain antagonism assays. This is particularly intriguing since neither the 4-methoxy nor the 4-Br phenylacetanoyl parent AHLs (7 and 8) were very active in LasR; rather, they are both active antagonists in LuxR.²⁰ Thiolactones 24/25 and 32/33, which contain either an unsubstituted phenyl or a biphenyl tail, exhibit even more exaggerated differences in activity; they are simultaneously good antagonists in the E. coli strain and good agonists in the antagonism assay in the P. aeruginosa strain. The parent AHLs (control phenyl AHL 5 and biphenyl AHL 9) display minimal activity in LasR as well.²¹ These data trends for thiolactone analogs versus their parent AHLs clearly suggest that minor perturbations (i.e., O to S) to the AHL head group can have dramatic effects on overall compound activity and receptor selectivity.

We note again here that the P. aeruginosa reporter strain retains the two other LuxR-type proteins that govern QS in this organism (QscR and RhlR), while the E. coli strain lacks these receptors. The lack of these receptors in E. coli could be one reason for the complicated thiolactone activity trends for LasR outlined above when comparing the two strains. Differences in membrane permeability between the two strains could be another contributing factor (see above). Both RhlR and QscR can regulate LasR to some degree, with QscR directly repressing LasR using its identical cognate ligand, OdDHL.¹⁹ In previous work, we have shown that many of our non-native AHLs can simultaneously inhibit LasR and QscR. In this context, thiolactones 24, 25, and 28–33 could also be inhibiting QscR in addition to LasR. A QscR inhibitor would yield LasR agonistic activity in the P. aeruginosa strain, yet would exhibit LasR antagonistic activity in the QscR null E. coli strain. This is the trend observed above for these compounds. Additional experiments with QscR are necessary to evaluate this hypothesis.

Since it was difficult to determine a definitive acyl group SAR for the thiolactone LasR agonists and antagonists, we scrutinized the thiolactone stereochemistry to determine what role it played, if any, in dictating compound activity. If stereochemistry played a large role in receptor binding, we hypothesized that the enantiopure L-thiolactones would be approximately twice as active as the racemic thiolactones (screened at the same concentration). Our reasoning proved incorrect, however, as the enantiopure L-thiolactones and their racemates largely exhibited approximately the same activity (Table 1). (We note that no appreciable racemization was observed in our synthesis procedure for the L-thiolactones). This result suggests that however these ligands modulate LasR, the interaction is less specific than that of native AHL ligands. Thiolactones 35 and 34, however, are a special case. Although they are LasR antagonists in the E. coli strain and LasR agonists in the P. aeruginosa strain, 34 (the racemate) has higher activity in both cases than 35 (the enantiopure L-form). For these compounds, the D-enantiomer may be the more active form of the compound.

2.4. Possible Agonism Model in LasR for Thiolactones

The interesting trends in LasR agonism for the thiolactones in the presence of OdDHL within the individual strains caused us to consider the mechanism of LasR activation in more detail (Figure 3). Many of the thiolactone library members have similar architectures as OdDHL (most notably the aliphatic derivatives), allowing us to make the reasonable assumption that they can potentially bind in the OdDHL binding pocket in LasR. A simple schematic of LasR's mechanism of action is shown in Figure 3A. OdDHL binds to LasR, which then homodimerizes and binds to the promoter region of one of many genes under LasR control including LasI, RhlI, RhlR, and RsaL, acting as a transcriptional activator.¹⁶ Similarly, when a non-native LasR agonist replaces OdDHL in reporter gene assays, LasR:non-native ligand homodimers are believed to form (Figure 3B). In competitive antagonism assays, it is generally assumed that the non-native ligand is able to outcompete OdDHL for the binding pocket, either by creating inactive homodimers or by preventing dimerization from occurring. Based on the work of Bassler and co workers,³² it is also possible that the inactive homodimers are able to bind DNA, but the homodimer:DNA complex is then unable to successfully interact with RNA polymerase. Two alternate scenarios are possible, however, if one considers LasR heterodimer formation. Such heterodimers would consist of one LasR bound to OdDHL and one LasR bound to a non-native ligand (Figures 3C and D).

In one scenario (Figure 3C), non-native ligands may act as what we term cooperative agonists if they are able to form viable heterodimers with LasR, yet are either unable to form homodimers or form inactive homodimers. This could lead to an observed high agonistic activity in a competitive antagonism assay against OdDHL Therefore, if the homodimers of the non-native ligand are either inactive or unable to form, negligible agonistic activity would be observed for the non-native ligand alone in an agonism assay. For example, thiolactones 24/25, 29, and 32–34 exhibit this trend in the P. aeruginosa strain. Conversely, in a second scenario (Figure 3D), if the synthetic ligand is an agonist when OdDHL is absent and is an antagonist when OdDHL is present, both the non-native and OdDHL homodimers could likely agonize LasR, while the heterodimers are inactive. An antagonistic effect would therefore be observed when the natural ligand is present (Figure 3D). We have named compounds that could bind in this latter manner “bimodal binders” because of their dual antagonistic and agonistic properties. We uncovered thiolactones displaying this activity profile in both LuxR and TraR (see below). Additional biochemical experiments are of course required to validate these models for LasR and related LuxR-type proteins, and are currently underway in our laboratory. Nonetheless, we find these models helpful for understanding the screening data obtained above for the thiolactones in LasR, and below in LuxR and TraR.

2.5. Antagonism and Agonism Assays in LuxR and TraR

The antagonism and agonism primary screening data for the thiolactone library in LuxR and TraR are shown in Table 2. Several exquisitely active compounds were identified, indicating the ability of acylated thiolactones to strongly modulate LuxR-type receptors beyond LasR. Sixteen thiolactones (including racemic compounds) were found to inhibit LuxR by over 50%. Notably, thiolactones 15, 18, 20, 27/28, 30/31, and 32/33, all of which are over 90% inhibitors, represent some of the most potent LuxR antagonists reported to date. The strong antagonistic activity of 15, the thiolactone mimic of OdDHL, is analogous to OdDHL itself in LuxR.²² Thiolactone 15 was also incapable of activating LuxR, corroborating the work of Schaefer et al.²⁹ Surprisingly, 3-NO₂ phenylacetanoyl thiolactones 26/27 were only weak agonists of LuxR, even though their parent 3-NO₂ AHL (6) is a “super activator” of LuxR.²⁰ This change in activity may be a result of the possible bimodal binder activity profile for 26/27 in LuxR (see above), as it is also behaves as a weak LuxR antagonist. However, several thiolactone displayed activities that tracked well with their parent AHLs. Thiolactones 28/29 were found to have analogous LuxR antagonistic activities as their methoxy AHL analog (7), which is a moderate inhibitor of LuxR. Similarly, thiolactones 30–34, modeled after our lactone-based LuxR inhibitors, were all strong LuxR antagonists, while 24/25 (based on inert AHL 5) were only weakly inhibitory.

Turning to TraR, we only found one thiolactone, the OOHL mimic 18, which displayed appreciable agonistic activity in TraR (50%; Table 2). Thiolactone 18 is notable as very few non-native TraR agonists have been reported to date.²¹ Several thiolactones, however, appeared to be either capable of what we now propose to be cooperative agonism (see above), or were extremely strong TraR antagonists. Compounds 19, 20, 26, and 27 behaved as cooperative agonists, with 19 and 20 agonizing TraR in the presence of OOHL to over 70% more than OOHL alone. The N-hexanoyl thiolactone 20 is the non-3-oxo analog of OHHL, suggesting that close similarities in ligand structure may permit active heterodimer formation. In turn, thiolactones 21 and 22 were very strong TraR inhibitors, with 22 surpassing the activity of its parent N-octanoyl HL (3) and inhibiting TraR by 99%.²² Compound 21 has a one carbon shorter acyl chain than 22 and is 20% less active, correlating well with acyl chain length activity trends seen for similar AHL analogs in TraR.²²

Analogous to the LasR primary screening data, thiolactone stereochemistry did not appear critical for LuxR agonism or antagonism. However, it did influence TraR agonism and antagonism trends in some cases. For example, the racemic thiolactone 30 and L-thiolactone 31 showed a reversal in activity; 30 is a 25% TraR antagonist, while 31 is a 60% TraR agonist in the antagonism assay. These data suggest that the D-stereoisomer may have higher TraR inhibitory activity, and the L-stereoisomer then has higher agonistic activity. In another example, L-thiolactone 29 exhibited very strong agonistic activity in the TraR antagonism assay (2× the activity of OOHL), while the racemic thiolactone 28 is only a 24% TraR agonist. For both 28 and 29, agonistic activity is only observed when OOHL is present. These data suggest that 28 and 29 could function as cooperative agonists of TraR (see above), with the L enantiomer imparting the majority of the activity. Indole thiolactones 34 and 35 showed a similar agonistic activity profile as 28 and 29, although not quite as exaggerated.

Overall, the thiolactone analog of OdDHL (15) was highly active in all strains tested—either as a LasR agonist (94% in E. coli and 88% in P. aeruginosa) or an antagonist in LuxR (92%) and TraR (65%). These trends match those observed for OdDHL in each strain, and suggest that, at least for 15, sulfur replacement in the lactone ring does not significantly affect its interactions with a LuxR-type receptor.

2.6. Dose Response Analysis for Selected L-Thiolactones

Dose response analyses for selected active, L-thiolactones (12, 14, 15, 20–23, 25, 27, 29, 31, 33, and 35) were conducted to determine their IC₅₀ and EC₅₀ values in the bacterial reporter strains. These data are listed in Table 3, and revealed several agonists and antagonists with sub-micromolar activities (15, 18, 20–22, 31, and 33). The majority of these compounds were LuxR antagonists. The most active compound identified overall was N-octanoyl L-thiolactone (22), which could antagonize LuxR by 50% at a 10-fold lower concentration than OHHL. Interestingly, thiolactone 22 was also the strongest LasR antagonist identified in the E. coli reporter strain, capable of antagonizing LasR by 50% at a 10-fold higher concentration than OdDHL. Following this trend, the second and third most active LuxR antagonists were also the next most active LasR antagonists in the E. coli strain (i.e., the N-heptanoyl (21) and biphenyl (33) L-thiolactones).

The dose response data for four thiolactone agonists are worthy of attention (12, 15, 18, and 23; Table 3). Interestingly, all four were native AHL analogs. The thiolactone analog of OOHL (18) is especially notable, as to our knowledge it is the most active non-native TraR agonist reported to date (albeit 100-fold less active then OOHL). In turn, OdDHL thiolactone analog 15 had EC₅₀ values approximately equal to OdDHL in both the E. coli and P. aeruginosa LasR reporter strains (~10 nM and 1 μM, respectively). OHHL thiolactone analog 12 was also a strong agonist in both LasR reporter strains, but with EC₅₀ values approximately 10-fold higher than OdDHL in each strain. Interestingly, unlike the thiolactone analogs of OOHL (18) in TraR and OdDHL (15) in LasR, OHHL thiolactone analog (12) was a mild LuxR antagonist instead. Finally, thiolactone 23, the non-3-oxo thiolactone analog of OdDHL, was found to be ~100-fold less active then OdDHL in the E. coli LasR reporter strain, yet was only ~10-fold less active than OdDHL in the P. aeruginosa reporter strain. This difference in activity is significant because many of the thiolactones examined in the primary assays displayed the opposite trend in the two strains, that is, heightened activity in the E. coli versus the P. aeruginosa LasR reporter strain.

Further comparison of the thiolactone dose response data to that for our previously reported AHLs (1–9) revealed that the oxygen → sulfur replacement can both augment and diminish relative compound activities in AHL analogs (Table 3).^{20, 21} For example, thiolactones 21, 22 and 33 largely exhibit similar agonistic or antagonistic potencies as their parent AHLs (2, 3, and 9) in LasR (E. coli), LuxR and TraR. The N-dodecanoyl thiolactone (23), however, displayed reduced agonistic activity relative to its AHL analog (4), suggesting that the sulfur replacement can impact ligand activity in some cases. Likewise, the phenylacetanoyl thiolactone 25 was a moderate LasR inhibitor in E. coli, while it parent AHL (5) was largely inactive in most LuxR-type receptors. Notably, the 3-NO₂ thiolactone 27 and indole thiolactone 35 were single digit micromolar agonists of LasR in E. coli, while their corresponding AHL analogs (6 and 10) were sub-micromolar LasR antagonists.²¹ Clearly, additional thiolactones will need to be studied to gain more insight into the subtle SARs behind these activity switches. Structural studies of these compounds complexed to various LuxR-type proteins could be particularly illuminating, as indicated by the recent X-ray crystallographic work of Hughson and co-workers on related thiolactones bound to CviR.³³

2.7. Thiolactone Hydrolysis Study

We sought to examine the hydrolytic stability of the thiolactone analogs described above, as generating probe compounds with heightened hydrolytic activity relative to AHLs was one motivation for this study. Like lactones, thiolactones are also hydrolyzable. However, direct comparison of the hydrolysis rates of AHLs versus their thiolactone analogs has, to our knowledge, yet to be reported. We provide such an analysis here.

We chose to analyze 15 as it was the most active thiolactone modulator identified in this study. A biological assay was developed to determine the functional half-life of 15 and its corresponding AHL analog, OdDHL. Our assay did not directly measure lactone hydrolysis, but rather the ability of the remaining non-hydrolyzed ligand to activate LasR. Previous experiments have shown that the hydrolysis half-life for OdDHL is approximately two days, while racemization of the chiral center was found to be less than 5% over the course of a week.²⁷ These data suggest that most ligand degradation is due to hydrolysis and not epimerization. Previous hydrolysis experiments have been monitored by NMR spectroscopy and performed in deuterated buffers containing 50% DMSO.^{27, 42} The use of high levels of DMSO reduces the biological relevance of the assay. In order to assess ligand degradation under more biologically relevant conditions, we monitored ligand activity levels using a LasR reporter strain (PA01 MW1 (pUM15); introduced above).³⁶

Our hydrolysis assay protocol utilized the P. aeruginosa LasR reporter to detect ligand activity after incubation in growth media for specified times using fluorescence (YFP; See Experimental Section). Fluorescence values corresponded to the degree of ligand degradation prior to contact with the reporter strain, and were normalized to that of a freshly prepared ligand sample (i.e., the positive control). Since we predict that the ligand degradation is due to hydrolysis, we assumed a pseudo first-order rate and plotted the natural log of the ligand activity as a percent of the positive control versus time. The slope of the graph therefore could be used to determine the half-life of the ligand according to the formula t_1/2=ln(2)/-slope (Figure 4).

Using this reporter assay procedure, we determined the half-life for OdDHL to be 48.2 h. This value corresponds closely with the hydrolysis half-life for OdDHL previously reported by Spring and co-workers.²⁷ A similar analysis of thiolactone 15 revealed its half-life to be 82.3 h, almost twice that of OdDHL. Yates and co-workers have shown that once ring-opened, the lactone does not reclose in appreciable quantities until pH values of less than 2.⁴² Since sulfur compounds are stronger nucleophiles than their oxygen analogs,⁴³ we hypothesize that 15 is either able to remain closed for longer than OdDHL or it is able to ring close at higher pHs, shifting the half-life to longer times. We note that the LasR, LuxR, and TraR reporter gene assays above were conducted for no longer than 16 h, so the majority of both OdDHL and 15 is likely in the closed form and ligand stability (or free thiol reactivity) did not have a significant impact on the observed activities. Nevertheless, these hydrolysis data suggest that thiolactone AHL analogs, such as 15, could have value for biological experiments conducted over longer time periods.

4.1. Acylated Thiolactone Synthesis

Thiolactones lacking 3-oxo functionalities in their acyl groups were synthesized in solution using previously reported EDC coupling protocols and purified using acid-base extraction procedures.^{46, 47}

Thiolactones containing 3-oxo acyl groups were prepared from acylated Meldrum's acid derivatives. The Meldrum's acid derivatives were synthesized by dissolving Meldrum's acid (0.5 g, 3.5 mmol) in CH₂Cl₂ (50 mL) under nitrogen at 0 °C. DMAP (0.86 g, 7 mmol) was added and allowed to dissolve completely. Thereafter, the alkyl acid chloride (3.5 mmol) was added to the reaction mixture slowly over 1 h at 0 °C. The reaction mixture was allowed to stir for an additional hour at 0 °C, and then overnight (~14 h) at rt. The reaction mixture was washed 2× with 2M HCl and 1× with brine, and the organic layer was dried over MgSO₄, filtered, and concentrated to yield the crude acylated Meldrum's acid derivative

The 3-oxo thiolactones were synthesized by stirring the homocysteine thiolactone (25 mg, 0.16 mmol) with the Meldrum's acid derivative (0.16 mmol) in acetonitrile (15 mL) at rt under nitrogen for 2 h, followed by refluxing for 5 h. The acetonitrile was then removed, and the reaction residue was redissolved in ethyl acetate and washed 1× each with saturated sodium bicarbonate, 1M NaHSO₄, and brine. The organic layer was dried over MgSO₄, filtered, and concentrated to yield the 3-oxo thiolactone products. Further purification by flash silica gel column chromatography (25% EtOAc/hexanes) was necessary for 14 and 18.

Purities and isolated yields for these compounds were 91–99% and 35–60%, respectively. See Supplementary Material for full compound characterization details.

4.2. Bacterial Reporter Gene Assays

Reporter gene assays for LasR (in E. coli), LuxR, and TraR were performed according to our previously reported methods.²¹ The LasR assay in the P. aeruginosa PA01 MW1 (pUM15) strain was modified from a literature procedure.³⁷ See Supplementary Material for full details of bacteriological protocols.

4.3. Thiolactone Hydrolysis Study

Thiolactone ligand was incubated in Luria-Bertani medium at 37 °C for predetermined times, and then added to a 96-well plate. The P. aeruginosa PA01 MW1 (pUM15) reporter strain was cultured overnight, pelleted, washed, and resuspended in a minimal amount of media. These cells were then added to the 96-well plate containing ligand, incubated for 8 h, and the wells were analyzed for optical density and YFP fluorescence. Fluorescence was normalized to cell density per well, and each time point was referenced to the cell-density-normalized readout of ligand prepared at time 0 (i.e., the positive control).