Sensitive Assay for the Lactonase Activity of Serum Paraoxonase 1 (PON1) by Harnessing the Fluorescence Turn-On Characteristics of Bioorthogonally Synthesized and Geometrically Controlled Chemical Probes

The lactonase activity of paraoxonase 1 (PON1) has a crucial antiatherogenic function, and also serves as an important biochemical marker in human blood because the aberrant lactonase activity of PON1 is a key indicator for a number of diverse human diseases. However, no sensitive fluorescence assays that detect PON1 lactonase activity are available. We report the synthesis of two fluorescence turn-on chemical probes 16a and 16b (16) able to quantify PON1 lactonase activity. The chemical probes were constructed utilizing a disulfide-containing bicyclononyne, derivatives of rhodamine B and carboxyfluorescein, and reactions including copper-free azide–alkyne cycloaddition. Fluorescence quenching in 16 was characterized by spectroscopic studies and was mainly attributed to the effect of contact quenching. Kinetic analysis of 16b confirmed the outstanding reactivity and specificity of 16b with thiols in the presence of general base catalysts. The 16b-based assay was employed to determine PON1 lactonase activity, with a linear range of 10.8–232.1 U L−1 and detection limit (LOD) of 10.8 U L−1, to quantify serum PON1 activity in human sera, and to determine the Ki of 20.9 μM for the 2-hydroxyquinoline inhibition of PON1 lactonase. We are employing 16b to develop high-throughput assays for PON1 lactonase activity.


Introduction
Many human diseases, including cardiovascular and neurodegenerative diseases, cancer, obesity, type 2 diabetes mellitus, aging, Crohn's disease, microbial infections, and non-alcoholic fatty liver disease, are characterized by chronic or acute oxidative stress and exacerbated inflammation in patients [1][2][3][4]. The detrimental disease states can be partly averted by the normal functions of three proteins-PON1, PON2 and PON3-in the paraoxonase (PON) family, which play anti-oxidative and anti-inflammatory roles in safeguarding Here, we report the successful synthesis of the novel fluorescence turn-on chemical probes 16a and 16b (16), capable of accurately quantifying PON1 lactonase activity. We have demonstrated that the 16b-based fluorescence assay is the first of its kind able to directly quantify PON1 lactonase activity in samples (Scheme 1), provide a broad linear detection range of PON1 lactonase activity with low limit of detection (LOD), and accurately determine PON1 lactonase activity in serum from humans. The chemical probe 16b will be an invaluable reagent to meet the pressing need for the sensitive and rapid quantification of PON1 lactonase activity in serum in clinical settings. Scheme 1. Use of the chemical probe 16b to quantitatively analyze catalysis and inhibition of PON1 lactonase activity. The PON1-catalyzed hydrolysis of thiobutyl butyrolactone (TBBL, a thiolactone) would provide a labile hemithioacetal, which spontaneously cleaves to give 1-butanethiol (nBuSH) as one of the products. The thiolate of nBuSH is highlighted in blue. Subsequently, the thiolate nucleophilically attacked (the pink arrow) the disulfide bond in 16b to obliterate the fluorescence quenching effects (the green arrow) and to liberate the 6-carboxyfluorescein (6-FAM) fluorescence. Increments in the 6-FAM fluorescence (the light green arrow) are, therefore, proportional to increases in PON1 lactonase activity (the light blue arrow). a: (TBBL), Tris, Ca 2+ , pH 8.0. R 1 and R 2 , H or the thiolate of nBuSH.

Synthesis of the Fluorescent Chemical Probes 16b
Analogous to a recently reported fluorescence chemical probe for the quantification of butyrylcholinesterase (BChE) activity [38], the fluorescent chemical probes 16b was also synthesized from exo-6 (Scheme S1, Supplementary Materials, SM), a crucial disulfidecontaining bicyclononyne capable of performing the important bioorthogonal and biocompatible copper-free strained-promoted azide-alkyne cycloaddition (SPAAC) reactions Scheme 1. Use of the chemical probe 16b to quantitatively analyze catalysis and inhibition of PON1 lactonase activity. The PON1-catalyzed hydrolysis of thiobutyl butyrolactone (TBBL, a thiolactone) would provide a labile hemithioacetal, which spontaneously cleaves to give 1-butanethiol (nBuSH) as one of the products. The thiolate of nBuSH is highlighted in blue. Subsequently, the thiolate nucleophilically attacked (the pink arrow) the disulfide bond in 16b to obliterate the fluorescence quenching effects (the green arrow) and to liberate the 6-carboxyfluorescein (6-FAM) fluorescence. Increments in the 6-FAM fluorescence (the light green arrow) are, therefore, proportional to increases in PON1 lactonase activity (the light blue arrow). a: (TBBL), Tris, Ca 2+ , pH 8.0. R 1 and R 2 , H or the thiolate of nBuSH.

Synthesis of the Fluorescent Chemical Probes 16b
Analogous to a recently reported fluorescence chemical probe for the quantification of butyrylcholinesterase (BChE) activity [38], the fluorescent chemical probes 16b was also synthesized from exo-6 (Scheme S1, Supplementary Materials, SM), a crucial disulfidecontaining bicyclononyne capable of performing the important bioorthogonal and biocompatible copper-free strained-promoted azide-alkyne cycloaddition (SPAAC) reactions [39] for the effective synthesis of 16b. We previously indicated that the trifunctional nature of exo-6 provides the appropriate molecule platform to modulate the incorporation of specific fluorophores into one of the two termini in exo-6, and to facilitate the tuning of the fluorogenic properties of afforded chemical probes [38]. We employed the rhodamine B-carboxyfluorescein (FAM) pair in this current study because the degree of proximity between the fluorophores in the chemical probes could be harnessed to develop a sensitive FAM-based PON1 lactonase activity assay at alkaline pH.
Both rhodamine B and FAM were modified to introduce azido and NHS ester groups, respectively, before coupling with exo-6 and achieving synthesis of 16. For example, the azido derivative of rhodamine B (9) was synthesized from the amidation reaction of rhodamine B with 8 (Scheme 2). In addition, we developed an efficient method to effectively separate and purify both 5-carboxyfluorescein (5-FAM, 13a) and its geometric isomer 6-carboxyfluorescein (6-FAM, 13b) from the 5(6)-FAM (10) mixture, and to convert each geometric isomer-pure FAM to the corresponding FAM-NHS ester (14) (Scheme 3). As we will reveal later, the acquisition of the geometric isomer-pure FAM and FAM derivatives was essential to better control the fluorogenic property of the chemical probes and develop more sensitive fluorescence assays for the quantification of PON1 lactonase activity.
Molecules 2022, 27, x FOR PEER REVIEW 4 of 22 [39] for the effective synthesis of 16b. We previously indicated that the trifunctional nature of exo-6 provides the appropriate molecule platform to modulate the incorporation of specific fluorophores into one of the two termini in exo-6, and to facilitate the tuning of the fluorogenic properties of afforded chemical probes [38]. We employed the rhodamine B-carboxyfluorescein (FAM) pair in this current study because the degree of proximity between the fluorophores in the chemical probes could be harnessed to develop a sensitive FAM-based PON1 lactonase activity assay at alkaline pH. Both rhodamine B and FAM were modified to introduce azido and NHS ester groups, respectively, before coupling with exo-6 and achieving synthesis of 16. For example, the azido derivative of rhodamine B (9) was synthesized from the amidation reaction of rhodamine B with 8 (Scheme 2). In addition, we developed an efficient method to effectively separate and purify both 5-carboxyfluorescein (5-FAM, 13a) and its geometric isomer 6carboxyfluorescein (6-FAM, 13b) from the 5(6)-FAM (10) mixture, and to convert each geometric isomer-pure FAM to the corresponding FAM-NHS ester (14) (Scheme 3). As we will reveal later, the acquisition of the geometric isomer-pure FAM and FAM derivatives was essential to better control the fluorogenic property of the chemical probes and develop more sensitive fluorescence assays for the quantification of PON1 lactonase activity.  [39] for the effective synthesis of 16b. We previously indicated that the trifunctional nature of exo-6 provides the appropriate molecule platform to modulate the incorporation of specific fluorophores into one of the two termini in exo-6, and to facilitate the tuning of the fluorogenic properties of afforded chemical probes [38]. We employed the rhodamine B-carboxyfluorescein (FAM) pair in this current study because the degree of proximity between the fluorophores in the chemical probes could be harnessed to develop a sensitive FAM-based PON1 lactonase activity assay at alkaline pH. Both rhodamine B and FAM were modified to introduce azido and NHS ester groups, respectively, before coupling with exo-6 and achieving synthesis of 16. For example, the azido derivative of rhodamine B (9) was synthesized from the amidation reaction of rhodamine B with 8 (Scheme 2). In addition, we developed an efficient method to effectively separate and purify both 5-carboxyfluorescein (5-FAM, 13a) and its geometric isomer 6carboxyfluorescein (6-FAM, 13b) from the 5(6)-FAM (10) mixture, and to convert each geometric isomer-pure FAM to the corresponding FAM-NHS ester (14) (Scheme 3). As we will reveal later, the acquisition of the geometric isomer-pure FAM and FAM derivatives was essential to better control the fluorogenic property of the chemical probes and develop more sensitive fluorescence assays for the quantification of PON1 lactonase activity. Subsequent syntheses of the chemical probes 16a and 16b, however, adopted a drastically different approach from that of the BChE chemical probe synthesis [38]. Here, we abandoned the reported one-pot reaction format and employed a two-reaction strategy in order to increase the yields of the chemical probes 16 in the synthesis (Scheme 4). The one-pot reaction used in synthesis of the BChE chemical probe consistently provided a rather poor yield of less than 50% [38]. We reasoned that the low reaction yield problem could be overcome by coupling exo-6 with 14 through a more optimal amidation reaction to give 15, which was later subjected to a high-yield SPAAC reaction with 9 to give the target chemical probes 16. Indeed, we successfully synthesized the chemical probes 16 with good yields by following the two-reaction scheme. Specifically, starting from exo-6 and 14a/14b, the combined two-reaction yields are 76.2% for 16a and 55.1% for 16b. In addition, the yield difference between 16a and 16b suggests that more severe steric hindrance and molecular crowdedness in 14b and 15b clearly overwhelmed the accessibility of the 6-carboxylate and the 6-carboxyl NHS ester group in the molecules, compromising the yields of 15b (61.6%) and 16b (89.4%), respectively. The apparent disadvantage related to the geometrical constraints of the 6-FAM moiety deteriorating the 16b synthesis yield, however, endowed 16b with unique fluorogenic properties useful in developing sensitive assays for PON1 lactonase activity, which will be revealed in the next spectrophotometric studies of the chemical probes.

Contact Quenching and Fluorogenic Properties of the Fluorescent Chemical Probes 16a and 16b
Contact fluorescence quenching in the chemical probes 16a and 16b was characterized by UV-Vis spectrophotometry, and demonstrated the perturbation of the major FAM (λ max , 495 nm) and rhodamine B (λ max ,~550 nm) absorption bands in 16 due to the proximity of the fluorophore moieties, which led to the formation of nonfluorescent intramolecular heterodimers in 16 (Figure 1a). The resulting UV-Vis spectra showed the complete disappearance of the characteristic rhodamine B absorption band and the dramatic diminution of the typical FAM absorption band in 16. Alternatively, the rhodamine B-FAM pair had the ability to perform Förster resonance energy transfer (FRET) and quench the FAM fluorescence in 16, when maintaining an appropriate distance between the fluorophores. FRET quenching, however, was likely only a minor contributor to the observed diminished FAM fluorescence in 16 ( Figure 1a and Figure S1A, SM), because a predominance of FRET effects would require the visible absorption of the fluorophores in 16 in order to stay intact, and to have intensities similar to those of the free fluorophores in the same concentration [40][41][42]. This is obviously not seen in Figure 1a. Indeed, we had to enact a four-fold increase in the concentration of 16 in order to observe rhodamine B's fluorescence in glacial acetic acid solution ( Figure S1B, SM).
Overall, the results were consistent with past studies demonstrating that a short distance (van der Waals contact) between two fluorophores, or between a fluorophorequencher pair, in contact quenching constructs would facilitate direct electronic interaction of the fluorophore-excited state with the counterpart chromophore, prohibit the return of the excited fluorophore electrons to the ground state, and be characterized by the significantly altered visible absorption properties of the chromophore pair [38,42,43]. It is noted that the same visible absorption change was documented recently for the BChE chemical probe, which, although it was originally described as a FRET probe, is more appropriately classified as a contact quenching fluorescence chemical probe [38].
The fluorogenic properties of 16 were confirmed by reacting 16 with L-cysteine in the Tris buffer ( Figure 1b and Figure S1A, SM). Significantly, the contact quenching of FAM fluorescence in 16 was quickly and completely obliterated within 15-30 min, indicating that 16 promptly reacted with the thiol-containing L-cysteine. We decided to study the potential use of 16b as a reagent in the assay used to quantify PON1 lactonase activity, because 16b was unequivocally a more sensitive and reactive chemical probe that provided fluorogenic properties superior to those of 16a (Figure 1b and Figure S1A, SM). The fluorescence turnon properties of 16b in the presence of various compounds with or without a thiol group  Figure 2a) again confirmed the rapid, thiol-responsive reactivity of 16b, and the potential of 16b to be an effective fluorescence chemical probe for the detection of thiols. Clearly, all thiol-containing reactants, including 1-butanethiol (nBuSH, one of the hydrolyzed products of the PON1 substrate 5-(thiobutyl) butyrolactone (TBBL)), were able to swiftly react with 16b and to increase levels of FAM fluorescence. The discriminative reactivity of 16b with thiols was also consistent with the pseudo-first order rate constants (k 1 ) for the reactions (Figure 2b and Figure S2, SM). The qualitative and quantitative measurements of the 16b reactions demonstrate that, among the studied reactants, 2-aminoethanethiol (2-AET) was the most effective in relieving the contact quenching effect in 16b and liberating the FAM fluorescence ( Figure 2 and Figure S2, SM). We therefore decided to more comprehensively study the fluorogenic mechanism of the reaction between 16b and 2-AET, because the results would be critical in establishing a 16b-based assay as a viable method for measuring PON1 lactonase activity.

Contact Quenching and Fluorogenic Properties of the Fluorescent Chemical Probes 16a and 16b
Contact fluorescence quenching in the chemical probes 16a and 16b was characterized by UV-Vis spectrophotometry, and demonstrated the perturbation of the major FAM (λmax, ~495 nm) and rhodamine B (λmax, ~550 nm) absorption bands in 16 due to the prox- quencher pair, in contact quenching constructs would facilitate direct electronic interaction of the fluorophore-excited state with the counterpart chromophore, prohibit the return of the excited fluorophore electrons to the ground state, and be characterized by the significantly altered visible absorption properties of the chromophore pair [38,42,43]. It is noted that the same visible absorption change was documented recently for the BChE chemical probe, which, although it was originally described as a FRET probe, is more appropriately classified as a contact quenching fluorescence chemical probe [38]. The fluorogenic properties of 16 were confirmed by reacting 16 with L-cysteine in the Tris buffer (Figures 1b and S1A, SM). Significantly, the contact quenching of FAM fluorescence in 16 was quickly and completely obliterated within 15-30 min, indicating that 16 promptly reacted with the thiol-containing L-cysteine. We decided to study the potential use of 16b as a reagent in the assay used to quantify PON1 lactonase activity, because 16b was unequivocally a more sensitive and reactive chemical probe that provided fluorogenic properties superior to those of 16a (Figures 1b and S1A, SM). The fluorescence turn-on properties of 16b in the presence of various compounds with or without a thiol group (Figure 2a) again confirmed the rapid, thiol-responsive reactivity of 16b, and the potential of 16b to be an effective fluorescence chemical probe for the detection of thiols. Clearly, all thiol-containing reactants, including 1-butanethiol (nBuSH, one of the hydrolyzed products of the PON1 substrate 5-(thiobutyl) butyrolactone (TBBL)), were able to swiftly react with 16b and to increase levels of FAM fluorescence. The discriminative reactivity of 16b with thiols was also consistent with the pseudo-first order rate constants (k1) for the reactions (Figures 2b and S2, SM). The qualitative and quantitative measurements of the 16b reactions demonstrate that, among the studied reactants, 2-aminoethanethiol (2-AET) was the most effective in relieving the contact quenching effect in 16b and liberating the FAM fluorescence (Figures 2 and S2, SM). We therefore decided to more comprehensively study the fluorogenic mechanism of the reaction between 16b and 2- . The error bars were too short to be visible in some reactions. The symbol * indicates that 16b had no measurable reactivity, meaning that the values of averaged k1 and standard deviation could not be determined.

Mechanistic Studies of the Fluorogenic Reaction of 16b with 2-AET
The kinetic studies of the reaction between a constant 16b concentration (0.4 μM) and varied 2-AET provide insight into the mechanism of the fluorogenic reaction. We have determined the k1 values of the reactions of 16b in the presence of different 2-AET concentrations, and plotted the values of k1 vs. 2-AET to give a straight line (Figure 3a). Therefore, the 2-AET-16b reaction is, overall, second-order, first-order to 16b, and first-order to 2-AET, with the second-order rate constant k2 of the reaction equal to 0.83 M −1 s −1 , as determined from the slope of the line. Again, the chemical probe 16b is similar to the recently 1-butanethiol (nBuSH), respectively. The reactions were carried out in the dark at rt for 1.5 h before the photograph was taken under UV illumination (365 nm). (b) Determination of the pseudo-first order rate constant k 1 from each of the reactions of 16b with various reactants. Please see Section 3.3 for the experimental details. Each reactant was analyzed at least three times in order to acquire the averaged k 1 values and standard deviation (the error bar). The error bars were too short to be visible in some reactions. The symbol * indicates that 16b had no measurable reactivity, meaning that the values of averaged k 1 and standard deviation could not be determined.

Mechanistic Studies of the Fluorogenic Reaction of 16b with 2-AET
The kinetic studies of the reaction between a constant 16b concentration (0.4 µM) and varied 2-AET provide insight into the mechanism of the fluorogenic reaction. We have determined the k 1 values of the reactions of 16b in the presence of different 2-AET concentrations, and plotted the values of k 1 vs. 2-AET to give a straight line ( Figure 3a). Therefore, the 2-AET-16b reaction is, overall, second-order, first-order to 16b, and firstorder to 2-AET, with the second-order rate constant k 2 of the reaction equal to 0.83 M −1 s −1 , as determined from the slope of the line. Again, the chemical probe 16b is similar to the recently reported BChE probe [38]; it supports a S N 2 nucleophilic substitution reaction mechanism, in which the rapid deprotonation of 2-AET is followed by the rate-limiting nucleophilic attack of the corresponding thiolate on the disulfide bond in 16b to cleave the disulfide bond and to release the 6-FAM fluorescence.  A survey of pH and metal ion effects on the 16b-thiol reaction shed further light on the reaction mechanism. We first obtained k1 for the reactions between 0.4 μM of 16b and 5 mM of 2-AET in the presence of differing pH and/or buffers. The pH titration results were diagrammed (Figure 3b) and unambiguously show incremental changes in k1 values, while the pH of the buffers was changed from acidic to alkaline. The values of k1 approximately plateaued at pH 8.5, which they clearly demonstrated characteristics of general base catalysis to accelerate the 2-AET-16b reaction. In addition, the titration study allowed us to determine the pKa1 of 2-AET to be 8.16 ± 0.07, which is a value consistent with the previously reported 8.19 [44].
Moreover, the potential of metal interference in fluorogenic reactions of 16b was determined by kinetic studies of reactions in the presence of 0.4 μM of 16b, 5 mM of L-cysteine, and 1 mM of the studied metal ions ( Figure S3A, SM). Only Cu(II) and Fe(III) significantly inhibited the L-cysteine-16b reaction. We found that, in the L-cysteine-16b reaction containing Mn(II), the cation formed complexes with 16b and precipitated out of the solution, which rendered kinetic analysis of the reaction impractical. Overall, the second-order thiol-16b reaction is characterized by a general base catalysis with a SN2 nucleophilic substitution reaction mechanism; moreover, the metal ions prevalent in blood do not interfere in the thiol-16b reaction.
Finally, we evaluated the relative importance and role of molecular crowdedness surrounding the disulfide bond in 16b in bestowing the ability of 16b to differentially react with various thiol-containing reactants. For example, the k1 value of the reaction with glutathione (GSH, 0.076 min −1 ) was less than one-fifth of that of the reaction with 2-AET (k1 of 0.387 min −1 ) (Figure 2b), which suggests that 16b only moderately reacted with GSH. The previously studied BChE chemical probe also displayed peculiar reactivity toward A survey of pH and metal ion effects on the 16b-thiol reaction shed further light on the reaction mechanism. We first obtained k 1 for the reactions between 0.4 µM of 16b and 5 mM of 2-AET in the presence of differing pH and/or buffers. The pH titration results were diagrammed ( Figure 3b) and unambiguously show incremental changes in k 1 values, while the pH of the buffers was changed from acidic to alkaline. The values of k 1 approximately plateaued at pH 8.5, which they clearly demonstrated characteristics of general base catalysis to accelerate the 2-AET-16b reaction. In addition, the titration study allowed us to determine the pK a1 of 2-AET to be 8.16 ± 0.07, which is a value consistent with the previously reported 8.19 [44].
Moreover, the potential of metal interference in fluorogenic reactions of 16b was determined by kinetic studies of reactions in the presence of 0.4 µM of 16b, 5 mM of L-cysteine, and 1 mM of the studied metal ions ( Figure S3A, SM). Only Cu(II) and Fe(III) significantly inhibited the L-cysteine-16b reaction. We found that, in the L-cysteine-16b reaction containing Mn(II), the cation formed complexes with 16b and precipitated out of the solution, which rendered kinetic analysis of the reaction impractical. Overall, the second-order thiol-16b reaction is characterized by a general base catalysis with a S N 2 nucleophilic substitution reaction mechanism; moreover, the metal ions prevalent in blood do not interfere in the thiol-16b reaction. Finally, we evaluated the relative importance and role of molecular crowdedness surrounding the disulfide bond in 16b in bestowing the ability of 16b to differentially react with various thiol-containing reactants. For example, the k 1 value of the reaction with glutathione (GSH, 0.076 min −1 ) was less than one-fifth of that of the reaction with 2-AET (k 1 of 0.387 min −1 ) (Figure 2b), which suggests that 16b only moderately reacted with GSH. The previously studied BChE chemical probe also displayed peculiar reactivity toward GSH, and had a significantly diminished fluorogenic property in the presence of GSH [38]. The BChE chemical probe apparently had a unique intrinsic ability to differentiate and rapidly react with most thiol-containing compounds, while remaining rather unreactive to GSH. Subsequent studies have demonstrated that the BChE probe was able to impose steric hindrance on GSH and to prevent the corresponding thiolate of GSH from nucleophilically attacking the disulfide bond in the chemical probe. We thus sought to determine whether the same molecular constraints were also the origin of the distinctive reactivity of 16b with GSH.
We again exploited two GSH derivatives, a monoethyl ester derivative 17 and a dimethyl ester derivative 18, both of which were previously synthesized to study the BChE probe ( Figure S3B, SM) [38]. Here, 17 and 18 were used as substitutes for GSH in the kinetic analysis of the reactions with 16b. The chemical probe 16b reacted with the dimethyl ester of GSH (18) similarly to how the BChE probe reacted with 18. The k 1 value (0.064 min −1 ) of the 18-16b reaction was comparable to that of the GSH-16b reaction (0.076 min −1 ; Figure  S3C, SM). However, the k 1 (0.038 min −1 ) value of the reaction of 16b with the monoethyl ester of GSH, i.e., 17, was half of that for the GSH-16b reaction, and was drastically different from that of the reaction of the BChE probe with 17, which showed completely abolished reactivity and delivered a k 1 essentially equal to zero. Evidently, replacing a methyl group with an ethyl group in GSH enhanced the 16b-imposed steric obstruction, but was unable to annihilate the reaction with 17, as was observed in the reaction between the BChE probe and 17. Thus, GSH's glycine carboxylate group, which was either ethylated in 17 or methylated in 18, experienced less constraint caused by 16b than by the BChE probe. The reduced molecular crowdedness in 16b might contribute to its improved reactivity with GSH. Consequently, the steric hindrance of 16b experienced by GSH is still the most plausible explanation for the k 1 value of 0.076 min −1 in the GSH-16b reaction being smaller than that in the 2-AET-16b reaction, but not significantly small enough to dramatically reduce the reaction rate. The same molecular constraints of 16b might also hinder the reaction with nBuSH, to give a k 1 of 0.072 min −1 (Figure 2b), even though the structure of nBuSH is unrelated to that of GSH.
Since GSH is found naturally and abundantly in human blood, with a concentration of 2.09 ± 1.15 µM in the plasma of healthy humans [45], and had reactivity with 16b comparable to that of nBuSH in the 16b-nBuSH reaction, it could potentially interfere in a 16b-based assay used to measure the activity of PON1 lactonase in blood samples. Consequently, as described in Section 3.4, we have developed a strategy to avoid interference from serum GSH, and have successfully exploited 16b to accurately quantify PON1 lactonase activity in serum.

Development of the 16b-Based Fluorescence Turn-On Assay for Measuring PON1 Lactonase Activity
The fluorescence turn-on assay based on 16b (Scheme 1) demonstrated an ability to specifically and accurately measure PON1 lactonase activity when it was qualitatively revealed that 16b reacted only with the catalytic reaction product nBuSH in the presence of TBBL and PON1 (Figure 4a). The quantitative analysis of the PON1-TBBL-16b catalysis system also justified the use of 16b for the accurate detection of PON1 lactonase activity ( Figure S4, SM). The time-dependent fluorescence change curve of a PON1-TBBL-16b reaction exhibited the classical Michaelis-Menten model of enzyme kinetics, and facilitated the subsequent determination of PON1 lactonase activity in terms of initial velocities, v i (Figure 4b). Thus, the chemical probe 16b has the critical characteristics necessary for the sensitive and specific quantification of the PON1 lactonase activity in human blood samples.

Sensitive and Accurate Determination of PON1 Lactonase Activity in Human Serum by the Fluorescence Turn-On Assay in the Presence of 16b
The results demonstrating the ability of the 16b-based assay to accurately quantify PON1 lactonase activity in samples (Figures 4 and S4, SM) encouraged us to employ the nBuSH-responsive, 6-FAM fluorescence-releasing 16b to determine PON1 lactonase activity in serum. We began with the kinetic analysis of the time-dependent fluorescence turn-on of 16b in the presence of rePON1 standards with different activities, in order to acquire the corresponding values of vi. The linear regression analysis of PON1 activity vs. vi showed that the 16b-based assay provided a linear calibration curve with a slope m of 0.033 and a good linear detection range of 10.8-232.1 U L −1 (Figure 5a). The LOD of PON1 lactonase activity determined by the 16b-based assay was calculated to be 10.

Sensitive and Accurate Determination of PON1 Lactonase Activity in Human Serum by the Fluorescence Turn-On Assay in the Presence of 16b
The results demonstrating the ability of the 16b-based assay to accurately quantify PON1 lactonase activity in samples ( Figure 4 and Figure S4, SM) encouraged us to employ the nBuSH-responsive, 6-FAM fluorescence-releasing 16b to determine PON1 lactonase activity in serum. We began with the kinetic analysis of the time-dependent fluorescence turn-on of 16b in the presence of rePON1 standards with different activities, in order to acquire the corresponding values of v i . The linear regression analysis of PON1 activity vs. v i showed that the 16b-based assay provided a linear calibration curve with a slope m of 0.033 and a good linear detection range of 10.8-232.1 U L −1 (Figure 5a). The LOD of PON1 lactonase activity determined by the 16b-based assay was calculated to be 10. The calibration curve in Figure 5a was employed to determine PON1 lactonase activity in serum samples from three healthy males. The 16b-based assay successfully quantified PON1 lactonase activity in the serum samples ( Figure 5b) and provided the PON1 lactonase activity in the range of 17,800-19,500 U L −1 . The PON1 lactonase activity in human serum was previously reported to be 3800 ± 1900 U L −1 , according to a colorimetric Ellman's assay [33], which is only approximately one-fifth of the value provided by the current study. The disparity between the values of PON1 lactonase activity in human serum could be due to the lower sensitivity of Ellman's assay in comparison with the current fluorescence turn-on assay based on 16b. This may suggest that the 16b-based fluorescence assay is more capable of sensitively and accurately determining PON1 lactonase activity in serum samples. The calibration curve in Figure 5a was employed to determine PON1 lactonase activity in serum samples from three healthy males. The 16b-based assay successfully quantified PON1 lactonase activity in the serum samples ( Figure 5b) and provided the PON1 lactonase activity in the range of 17,800-19,500 U L −1 . The PON1 lactonase activity in human serum was previously reported to be 3800 ± 1900 U L −1 , according to a colorimetric Ellman's assay [33], which is only approximately one-fifth of the value provided by the current study. The disparity between the values of PON1 lactonase activity in human serum could be due to the lower sensitivity of Ellman's assay in comparison with the current fluorescence turn-on assay based on 16b. This may suggest that the 16b-based fluorescence assay is more capable of sensitively and accurately determining PON1 lactonase activity in serum samples.

Kinetic Analysis of PON1 Inhibition to Corroborate the Requirement of PON1 Lactonase Activity for Turning on the 6-FAM Fluorescence in the 16b-Based Assay
We further explicated the importance of PON1 lactonase activity in the fluorescence turn-on analysis based on 16b by measuring the inhibition of PON1 catalysis in the presence of an inhibitor. Here, we have studied the inhibitory effect of 2-hydroxyquinoline (HQ) on PON1 catalysis using the 16b-based assay. HQ is a well-known PON1 inhibitor, which inhibits PON1 catalysis vis a competitive mechanism [30]. We successfully observed that the inhibitory assay based on 16b was able to detect a gradual decrease in vi with increasing HQ concentrations (Figures 6a and S5, SM). Kinetic analysis of the HQ inhibition of PON1 catalysis allowed us to acquire a Dixon plot (Figure 6b), which was employed to estimate the Ki value of 20.9 μM, a result over ten-fold higher than the Ki value of 0.9-1.42 μM determined by colorimetric methods [21]. The higher value of Ki obtained from the current study might reflect the approximate nature of the Dixon plot analysis. Nevertheless, the results in Figures 6 and S5 unmistakably demonstrate that the 16b-based assay indeed requires PON1 lactonase activity to liberate the 6-FAM fluorescence, and is well positioned to sensitively measure the variable PON1 lactonase activities in serum crucial to the diagnosis and treatment of human diseases.

Kinetic Analysis of PON1 Inhibition to Corroborate the Requirement of PON1 Lactonase Activity for Turning on the 6-FAM Fluorescence in the 16b-Based Assay
We further explicated the importance of PON1 lactonase activity in the fluorescence turn-on analysis based on 16b by measuring the inhibition of PON1 catalysis in the presence of an inhibitor. Here, we have studied the inhibitory effect of 2-hydroxyquinoline (HQ) on PON1 catalysis using the 16b-based assay. HQ is a well-known PON1 inhibitor, which inhibits PON1 catalysis vis a competitive mechanism [30]. We successfully observed that the inhibitory assay based on 16b was able to detect a gradual decrease in v i with increasing HQ concentrations (Figure 6a and Figure S5, SM). Kinetic analysis of the HQ inhibition of PON1 catalysis allowed us to acquire a Dixon plot (Figure 6b), which was employed to estimate the K i value of 20.9 µM, a result over ten-fold higher than the K i value of 0.9-1.42 µM determined by colorimetric methods [21]. The higher value of K i obtained from the current study might reflect the approximate nature of the Dixon plot analysis. Nevertheless, the results in Figure 6 and Figure S5 unmistakably demonstrate that the 16b-based assay indeed requires PON1 lactonase activity to liberate the 6-FAM fluorescence, and is well positioned to sensitively measure the variable PON1 lactonase activities in serum crucial to the diagnosis and treatment of human diseases.

Materials and Methods
All reagent-grade chemicals were purchased from commercial sources (Sigma-Aldrich, Acros, Alfa Aesar, and Mallinckrodt Baker) except where noted, and were further purified as necessary. 1 H and 13 C NMR spectra were recorded using either a Varian 200 or 400 MHz spectrometer (Varian, Inc., Palo Alto, CA, USA) at Kaohsiung Medical University, Taiwan (KMU). NMR samples were prepared in (CD3)2SO or CDCl3, and the chemical shifts of 1 H and 13 C signals were reported in parts per million based on the internal standard of each deuterated solvent. ESI high-resolution mass spectra were acquired using Solarix Fourier-transfer mass spectrometry (Bruker Taiwan Co., Ltd., Taiwan) at the Department of Chemistry, National Sun Yat-Sen University. The plasmid pET32a-PON1-G2E6 [46] containing the rePON1 gene was a generous gift of D. S. Tawfik (Weizmann Institute of Science, Israel). Synthesis of exo-6 has been reported previously [38] (Scheme S1, SM). TBBL, the substrate required for the measurement of PON1 lactonase activity, was synthesized by following a reported method [30].

Synthesis of Azido-Rhodamine B (9)
Compound 8 was one of the reactants used to synthesize 9 and was obtained from 7 [47]. The synthesis of 9 began with dissolution of rhodamine B (149.2 mg, 0.337 mmol, 1 equiv) in a solution containing acetonitrile (ACN; 4 mL) and chloroform (1 mL). BOP (benzotriazol-1-yloxytris(dimethylamino)phosphonium hexafluorophosphate) reagent (372.6 mg, 0.842 mmol, 2.5 equiv), 8 (101.9 mg, 1.011 mmol, 3 equiv) and N,N-diisopropylethylamine (DIPEA; 600 μL, 3.444 mol, 10.2 equiv) were added in sequence to the rhodamine  Figure S5. The HQ inhibitor constant K i was determined from the minus x-axis value at which the two extrapolated lines intersected, which, in this study, was 20.9 µM. Please see Section 3.4 for the experimental details. Each reaction was analyzed at least three times in order to acquire the averaged values and standard deviation (the error bar) of v i .

5-Carboxyfluorescein (5-FAM, 13a)
The reaction was performed by dissolving 12a (228.3 mg, 0.441 mmol, 1 equiv) in 5 mL of trifluoroacetic acid (TFA) and refluxing in the dark for 4 h. The reaction mixture was allowed to cool to rt and TFA was subsequently removed under reduced pressure. The resulting solid phase was worked up three times by the following procedure: resuspending in diethyl ether (Et 2 O; 3-5 mL), transferring to a test tube, centrifuging, and decanting the supernatant. The final solid was dissolved in MeOH (3 mL) and evaporated under reduced pressure to obtain 13a (159.2 mg, 0.422 mmol, 95 %). 1

6-Carboxyfluorescein (6-FAM, 13b)
Synthesis of 13b was similarly carried out by dissolving 12b (184 mg, 0.356 mmol, 1 equiv) in 5 mL of TFA and refluxing in the dark for 4 h. The reaction mixture was again allowed to cool to rt, followed by removing TFA under reduced pressure. The resulting solid phase was resuspended in an Et 2 O/ hexane solution (2:3; 3-5 mL), transferred to a test tube, and centrifuged; lastly, the supernatant was decanted. The workup procedure was performed two more times before dissolving the remaining solid in MeOH (3 mL). The resulting MeOH solution was evaporated under reduced pressure to obtain 13b (108 mg, 0.286 mmol, 80 %). 1

6-FAM NHS Ester (14b)
The synthesis of 14b was achieved by the addition of 13b (100 mg, 0.265 mmol, 1 equiv), NHS (76 mg, 0.66 mmol, 2.5 equiv) and EDC (76 mg, 0.396 mmol, 1.5 equiv) in sequence to 2 mL of DMF. The acquired solution was similarly reacted and worked up by following the procedures for the 14a synthesis reported above to yield 14b (97.4 mg, 0.205 mmol, 77.4 %). 1  For the synthesis of 15a, exo-6 (31.9 mg, 0.097 mmol, 1 equiv), 14a (55.2 mg, 0.117 mmol, 1.2 equiv), and DIPEA (75.7 mg, 102 µL, 0.583 mmol, 6 equiv) were sequentially dissolved in 2 mL of DMF to initiate the reaction, which was stirred and allowed to proceed in the dark at rt in a nitrogen atmosphere overnight. The resulting reaction mixture was evaporated under reduced pressure, redissolved in a limited volume of an acetic acid (AcOH)/MeOH/DCM (1:25:75) solution, and loaded onto a silica column pre-equilibrated with the same solution. Compounds were eluted from the column in the dark; fractions containing the product were pooled and evaporated under reduced pressure to give 15a (53.4 mg, 0.078 mmol, 80.2%). 1  Similarly, exo-6 (40.6 mg, 0.124 mmol, 1 equiv), 14b (70.2 mg, 0.148 mmol, 1.2 equiv), and DIPEA (96.1 mg, 129.5 µL, 0.742 mmol, 6 equiv) were sequentially dissolved in 2 mL of DMF; the reaction proceeded and the products treated following the same procedures as were used for 15a synthesis. The reaction described in this section afforded 15b (52.2 mg, 0.761 mmol, 61.6%). 1  Dioxane (2 mL) was used to dissolve 15a (47.7 mg, 0.070 mmol, 1 equiv) and 9 (37 mg, 0.070 mmol, 1 equiv) and to give the reaction mixture, which was refluxed in the dark for 5 h. The resulting reaction mixture was evaporated under reduced pressure, redissolved in a limited volume of an AcOH/hexane/EA (1:10:100) solution, and loaded onto a silica column pre-equilibrated with the same solution to wash out impurities in the dark. Additional compounds were eluted out of the column by an AcOH/EA (1:100) solution; fractions containing the product were pooled and evaporated under reduced pressure to give 16a (80.5 mg, 0.066 mmol, 95%). 1  The method of 16a synthesis described above was analogously applied for the synthesis 16b. Here a dioxane (2 mL) solution containing 15b (40 mg, 0.058 mmol, 1 equiv) and 9 (30.6 mg, 0.058 mmol, 1 equiv) was reacted and worked out to yield 16b (63.1 mg, 0.052 mmol, 89.4%). 1

Spectroscopic Measurements
The UV-Vis absorption spectra and kinetic studies were recorded using an Amersham Biosciences Ultrospec 2100 pro spectrophotometer (KMU), and fluorescence spectra and kinetic studies were obtained with a PerkinElmer LS-55 fluorescence spectrometer (KMU) with an excitation wavelength of 325 nm for 6-FAM detection in a 1 cm standard quartz cuvette.

Specificity and Mechanism of the Reactions of the Fluorescent Chemical Probe 16b with Thiols
The preference of 16b to react with certain compounds was characterized by the obtained pseudo-first order rate constant (k 1 ) values. A typical experiment for k 1 determination is briefly described below. A quartz cuvette containing 1388 µL of a Tris buffer (50 mM of Tris, 1 mM of Ca 2+ , pH 8.0) was mounted onto the temperature controller of the PerkinElmer LS-55 fluorescence spectrometer first. The measurement was initiated after the sequential transfer of 150 µL of a 50 mM solution of one of the reactants (DL-dithiothreitol (DTT), L-glutamate, glycine, L-cysteine, GSH, L-serine, L-lysine, L-methionine, 2-mercaptoethanol, 2-AET, or nBuSH) in the Tris buffer and 12 µL of 16b (50 µM in DMF) to the Tris solution in the cuvette to give reactant = 5 mM and 16b = 0.4 µM, respectively, in the reaction. The progress of the reaction at 25 • C was detected by the liberation of the 6-FAM fluorescence (λ max = 525 nm) until signal saturation occurred, typically after 15-90 min. The data of normalized fluorescence intensity at 525 nm were acquired by subtracting the background fluorescence intensity at 525 nm of 16b from the original measurement of fluorescence intensity at 525 nm. The normalized data were fitted to a single-exponential equation for first-order kinetics F(t) = F 0 + F max (1 − e −k1t ) (F(t), normalized 6-FAM fluorescence at a specific time point t) to give k 1 (GraphPad, La Jolla, CA, USA). Similar kinetic analysis was exploited to determine k 2, in which the concentration of 2-AET in the reaction was 1 mM, 2.5 mM, 5 mM, 10 mM, 12.5 mM or 15 mM. A second-order rate constant (k 2 ) value was acquired from the slope of a linear regression curve based on the plot of k 1 vs. 2-AET. The pH effect on the reaction of 16b with 2-AET was determined by the same first-order kinetics in the presence of different pH buffers including MES (pH 5.5-6.5), PIPES (pH 6.5-7.4), EPPS (pH 7.4-8.6) and CHES (pH 8.6-9.5). The pH titration data enabled calculation of the pK a1 value for 2-AET according to the equation k 1 = k 1,max [1 + 10 (pKa1−pH) ] (GraphPad, USA). The similar k 1 determination of the L-cysteine-16b reactions in the presence of 1 mM of metal ions (K + , Li + , Na + , Cd 2+ , Co(II), Cu(II), Mg 2+ , Mn(II), Ni(II), Zn 2+ , Fe(III)) and the background of 1 mM Ca 2+ unveiled the effects of metal ions on the reaction. The potential of 16b to react discriminately against GSH was likewise evaluated by the first-order kinetics of the reactions between 16b and 5 mM of the monoethyl (17) and the dimethyl (18) ester derivatives of GSH [38]. Kinetic experiments were performed in quadruplicate in order to determine k 1 for a specific reaction; the provided k 1 was the mean ± SD of the four experiments. The plasmid pET32a-PON1-G2E6 was transformed into BL21(DE3) cells in order to overexpress rePON1, which was shown to have a catalytic activity similar to human PON1 [46], and was subsequently purified according to an established method [46]. The purified rePON1 protein concentration was determined by the Bradford assay (Bio-Rad), and rePON1 lactonase activity was calibrated using Ellman's colorimetric method [48]. PON1 lactonase activity is expressed as U L −1 in which one unit (1 U) is defined as the amount of PON1 able to hydrolyze 1 µmol of TBBL per min. Analysis of PON1 activity for different PON1 samples in the presence of 16b was performed qualitatively in vials and quantitatively in the fluorescence spectrometer. The vial assay included four different solutions in which each vial contained a different combination of 2 µM of 16b (dissolved in dimethyl sulfoxide (DMSO), 0.5%), 5 mM of TBBL (dissolved in ACN, 0.2%), and 7.79 U L −1 of rePON1 (50% glycerol) in the Tris buffer. The solutions were prepared in the dark in four different vials and reacted at rt in the dark for 15 min. The liberated fluorescence was visualized and photographed, while the samples were illuminated with 365 nm light from a hand-held lamp.
PON1 lactonase activity was quantitatively determined by utilizing the 16b-based assay in solutions containing 16b (0.8 µM in DMSO, 1.6%), rePON1 (8.6-232.1 U L −1 , 0.25-1.25% glycerol), and TBBL (10 mM in ACN, 1%) in the Tris buffer in a quartz cuvette previously mounted onto the temperature controller of the PerkinElmer LS-55 fluorescence spectrometer. PON1 catalysis was carried out at 25 • C for 30 min; reaction progression was followed by monitoring the fluorescence intensity at 525 nm as a function of reaction time. The normalized fluorescence intensity data at 525 nm were acquired by subtracting a background fluorescence of 16b at the same wavelength from the original fluorescence intensity reading. Initial velocities (v i ) of PON1 catalysis were calculated from the fluorescence changes in windows of the steady state reactions. Human PON1 lactonase activity was measured after a trained phlebotomist at KMU Hospital obtained 20 mL of whole blood from three healthy male volunteers; the blood samples were centrifuged, the serum was separated from the plasma and diluted 100-fold, and the diluted serum fraction was used as a substitute for rePON1 in separate 16b-based reactions. It is noted that, before the addition of TBBL (10 mM in ACN, 1%), GSH in diluted serum was consumed in the presence of 16b (0.8 µM in DMSO, 1.6%) in the Tris buffer at 25 • C for 30 min. Finally, effects of the inhibitor HQ on PON1 catalysis were studied viz 16b-based analysis, in which each reaction contained 16b (0.8 µM in DMSO, 1.6%), rePON1 (250.9 U L −1 , 1.25% glycerol), TBBL (5 or 10 mM, 1% or 1.5% ACN, respectively), and HQ (50, 100, or 200 µM in ACN, 1% or 1.5%) in the Tris buffer. Each specific PON1 catalysis was performed three times with the reported v i as a mean ± SD of the experiments.

Conclusions
In summary, the current study has successfully provided critical information essential to: (1) the effective organic synthesis of the chemical probes 16 enabled by the modulated bioorthogonal chemistry; (2) the demonstration of the importance of contact fluorescence quenching and disulfide cleavage on the fluorogenic properties of 16; (3) the development of the first fluorescence turn-on assay for sensitive and accurate quantification of PON1 lactonase activity in samples of human serum. The assay exploited the fluorogenic property of the novel chemical probe 16b in the presence of PON1 and TBBL to release the 6-FAM fluorescence, and delivered a linear detection range of 10.8-232.1 U L −1 and an LOD of 10.8 U L −1 in the measurement of PON1 lactonase activity. The specificity and sensitivity of the 16b-based assay thus provides an excellent method to accurately and dynamically quantify PON1 lactonase activity in samples with broad applications in clinical settings and basic research.
The first major contribution of the study is highlighted by the synthetic method used for the chemical probes 16. Here, we developed a facile approach to separate 5-FAM and 6-FAM, which proved to be critical to improving the sensitivity of chemical probes in the study (Scheme 3). In addition, we have shown that the two-reaction method for the synthesis of 16 (Scheme 4) was more effective than the previously reported one-pot format [38], and provided higher yields when covalently linking two different chromophores to exo-6 in a programmed manner. The higher yield provided by the two-reaction approach in the synthesis of 16 is attributed to the outstanding reactivity of the SPAAC reaction to rapidly generate a triazole product, which, however, might be bulky enough to prevent subsequent amidation reactions with electrophiles. The synthesis of a chemical probe with high yields may require amidation reactions of electrophiles with exo-6 to be performed first, followed by the SPAAC reaction. We are collaborating with colleagues to determine the X-ray structures of 16. The molecular structures of 16 will provide crucial information about not only the extent of molecular crowdedness in 16, but also the importance of contact quenching effects on the fluorogenic properties of 16, which will be discussed in the next paragraph.
The second significant aspect of this study is that it provides evidence associating contact effects with the quenching of the FAM fluorescence in the chemical probes 16 (Figure 1a). The UV-Vis spectra support the assertion that the rhodamine B and FAM moieties in 16 were in van der Waals contact, in order to enable efficient FAM fluorescence quenching due to the distance between the fluorophores likely being on the sub-nanometer length scale. Moreover, the distance between rhodamine B and 6-FAM in 16b is probably shorter than that between rhodamine B and 5-FAM in 16a. The subtle distance difference between rhodamine B and FAM in 16 was enough to enhance contact quenching in 16b more prominently than in 16a, and to establish 16b as the more sensitive chemical probe ( Figure 1b and Figure S1A, SM). Again, elucidating the X-ray structures of 16a and 16b will help us to better understand the dominant role of contact quenching in 16, and to decipher the peculiar but unique reactivity of 16b with nBuSH.
We were concerned that the rather low reactivity of 16b with nBuSH would somehow affect the sensitivity of the fluorescence turn-on assay for the quantification of PON1 lactonase activity (Figure 2 and Figure S2, SM). The differential reactivity of 16b with various thiolates, including nBuSH, could not be solely explained by steric hindrance, electrostatic interactions, or a hydrophobic effect, and was likely a combination of multiple factors simultaneously. Nevertheless, 16b preferentially reacted with small, charged, and hydrophilic thiolates. Therefore, based on the understanding of the fluorogenic reaction mechanism for 16b ( Figure 3 and Figure S3, SM), we are synthesizing better lactonase substrates to substitute for TBBL in PON1 catalysis, in order to further improve the sensitivity and accuracy of the PON1 lactonase activity assay.
Finally, this study has clearly demonstrated the development and use of the first fluorescence turn-on assay to accurately quantify PON1 lactonase activity in samples including sera (Figure 4, Figure 5 and Figure S4, SM). The fluorogenic property of the assay exclusively depends on the PON1 lactonase activity releasing nBuSH, which was subsequently able to nucleophilically attack the disulfide bond in 16b ( Figure 6 and Figure S5, SM). As stated above, we are undertaking on-going studies aimed at developing fluorescence assays that will measure PON1 lactonase activity more sensitively and accurately. The critical information provided by current and future studies will pave the way for attaining highthroughput diagnostic tools capable of efficiently and effectively detecting aberrant PON1 lactonase activity in serum and other biological samples, so as to facilitate the diagnosis and treatment of human diseases.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of MOST and KMU.