pubs.acs.org/ac Article An Immunochemical Approach to Quantify and Assess the Potential Value of the Pseudomonas Quinolone Signal as a Biomarker of Infection Enrique J. Montagut, M. Teresa Martin-Gomez, and M. Pilar Marco* Cite This: Anal. Chem. 2021, 93, 4859−4866 Read Online ACCESS Metrics & More Article Recommendations * sı Supporting Information ABSTRACT: Quorum sensing (QS) is a bacterial cell density-based communication system using low molecular weight signals called autoinducers (AIs). Identification and quantification of these molecules could provide valuable information related to the stage of colonization or infection as well as the stage of the disease. With this scenario, we report here for the first time the development of antibodies against the PQS (pseudomonas quinolone signal), the main signaling molecule from the pqs QS system of Pseudomonas aeruginosa, and the development of a microplate-based enzyme-linked immunosorbent assay (ELISA) able of quantifying this molecule in complex biological media in the low nanometer range (LOD, 0.36 ± 0.14 nM in culture broth media). Moreover, the PQS ELISA here reported has been found to be robust and reliable, providing accurate results in culture media. The technique allowed us to follow up the PQS profile of the release of bacterial clinical isolates obtained from patients of different disease status. A clear correlation was found between the PQS immunoreactivity equivalents and the chronic or acute infection conditions, which supports the reported differences on virulence and behavior of these bacterial strains due to their adaptation capability to the host environment. The results obtained point to the potential of the PQS as a biomarker of infection and to the value of the antibodies and the technology developed for improving diagnosis and management of P. aeruginosa infections based on the precise identification of the pathogen, appropriate stratification of the patients according to their disease status, and knowledge of the disease progression. H ealthcare-associated and community-acquired infections (HAI and CAI, respectively) have become a major burden for public health worldwide,1 and Pseudomonas antibiotics for the treatment of infections caused by resistant bacteria, such as carbapenem-resistant P. aeruginosa.5 The lack of reliable tools for rapid diagnostic of bacterial infections has aeruginosa is particularly found among the most commonly contributed to the overprescription and misuse of broad isolated microorganisms in clinical settings. According to the spectrum antibiotics, strengthening the generation of multi- U.S. Centers for Disease Control and Prevention (CDC), drug-resistant bacteria.6 Current diagnostic methods based on about 51,000 P. aeruginosa HAI cases occur each year in U.S. a culture plate require several days to obtain conclusive results, hospitals. This pathogen has developed a vast adaptive which often is an unaffordable time for effectively treating the response and significant resistance to innate immune effectors infection.7 Many diagnostic alternatives have emerged as and to antibiotics, particularly during chronic infections. Such potential complements of the standard techniques, providing is the case in patients suffering from cystic fibrosis (CF), a more complete and accurate results. For instance, matrix- disease caused by a genetic disorder that predominantly affects assisted laser desorption ionization−time of flight−mass the mucociliary clearance of the respiratory tract,2 creating an spectrometry (MALDI-TOF-MS) analysis has been success- ideal scenario for opportunistic pathogens such as P. fully used to characterize the protein footprint from bacterial aeruginosa. This pathogen colonizes the lungs of these patients isolates of CF.8 Also, polymerase chain reaction (PCR) at the age of 2−3 years old and ends up persisting in 60−80% methods have appeared as interesting alternatives to identify of CF patients as a chronic infection, compromising their life pathogens.9 However, these techniques require complex selection of the targets, tedious and time-consuming and frequently ending in respiratory failure and premature death.3 On the other hand, antibiotic therapy rarely leads to elimination of a pathogen from chronically infected CF Received: November 10, 2020 respiratory airways. In 2017, multidrug-resistant P. aeruginosa Accepted: March 4, 2021 caused an estimated 32,600 infections among hospitalized Published: March 11, 2021 patients and 2700 estimated deaths in the United States.4 Recently, the World Health Organization (WHO) has published a list declaring the urgent need of finding new © 2021 American Chemical Society https://dx.doi.org/10.1021/acs.analchem.0c04731 4859 Anal. Chem. 2021, 93, 4859−4866 Analytical Chemistry pubs.acs.org/ac Article extractions, highly trained personnel, and/or expensive equip- measured in clinical samples, partially due to the increased ment,10 which prevent their wide implementation on clinical sample complexity and low detectability attained by some of settings. Point of care (PoC) devices may be a promising these techniques. With this scenario, we have focus our option in future routine clinical diagnostics, providing rapid attention on the development of specific antibodies against the reliable responses and being low cost and user-friendly to use. main signaling molecule of the pqs QS system, with the final PoC performance would allow us to change the management aim of developing immunochemical diagnostic technologies of the patient and would provide faster diagnosis than suitable to assess the potential value of these specific QS conventional methods, ideally in almost every place and molecules as biomarkers of infection. In this paper, we report, situation. Most of the time, these devices rely on the use of a for the first time, the development of a highly sensitive biorecognition element that specifically binds the correspond- microtiter-based enzyme-linked immunosorbent assay ing biomarker target,11 such is the case of antibodies. (ELISA) to quantify the PQS and its use to analyze the Host−pathogen interaction involves the release of a large concentration levels of this molecule in culture media where amount of molecules, which have potential as biomarkers for clinical isolates obtained from infected patients have been diagnosis, patient stratification, disease monitoring, or for grown. Preliminary data point to the potential value of this QS guiding antibiotic treatment as being the subject of molecule as the biomarker to diagnose P. aeruginosa infection investigation.12,13 Because of their key role in the development and to stratify patients in respect to the stage of infection. of pathogenesis, the signaling molecules of the quorum sensing (QS) system could be a potential biomarker of infection. QS is a bacterial communication mechanism based on the release of ■ EXPERIMENTAL SECTION General Methods and Instruments. See the Supporting low molecular weight signals that regulate the coordinated Information. genetic expression of a wide array of physiological activities.14 Synthesis of PQS Hapten for Immunization. The These signals, also called autoinducers (AIs), promote their hapten 3-(2-heptyl-3-hydroxy-4-oxo-1,4-dihydroquinolin-6-yl)- own biosynthesis and their concentration increase as a function propanoic acid (13) was synthesized following a similar of population density. At a particular AI concentration approach to that described by Reen et al.30 for the synthesis of threshold, the genetic expression of decisive factors such as the PQS through a five-step synthetic pathway from 3-(4- the secretion of virulence factors and biofilm formation is aminophenyl)propanoic acid and methyl 3-oxodecanoate (3), triggered. P. aeruginosa QS is a well-studied communication obtained by a nucleophilic reaction of 2,2-dimethyl-1,3- network composed of four systems (las, rhl, pqs, and iqs) dioxane-4,6-dione (Meldrum’s acid) with octanoyl chloride performing in a complex hierarchical manner. Particularly, the and subsequent methanolysis. All the intermediates and the quinolone-based pqs system, responsible for the regulation of final product were purified and characterized by nuclear virulence factors including pyocyanin, elastase, lectin, and magnetic resonance (NMR) spectroscopy and ultraperform- rhamnolipids,15 has particularly attracted the attention of the ance liquid chromatography (UPLC)-MS/MS.31 scientific community due to its QS-dependent and independ- Synthesis of the Bioconjugates PQS-KLH and PQS- ent functions.16 Its main signaling molecule, 2-heptyl-3- BSA. A solution of the PQS hapten (13) (3.32 mg, 10 μmol) hydroxy-4(1H)-quinolone or the PQS (pseudomonas quino- in anhydrous DMF (400 μL) was cooled to 4 °C. Afterward, lone signal), specific for P. aeruginosa, is responsible for tri-n-butylamine (2.62 μL,11 μmol) and isobutyl chloroformi- important functions besides its signaling activity. For instance, ate (1.56 μL, 12 μmol) were added to the hapten solution and the PQS mediates cytotoxicity and has been indicated to the mixture was left stirring for 15 min at 4 °C and 30 min at balance the growth and death in P. aeruginosa populations.17 room temperature. Then, 200 μL of the reaction mixture was The PQS also mediates iron acquisition due to its strong slowly added over the protein solution (bovine serum albumin chelating character and its activity to prompt the expression of (BSA) or keyhole limpet hemocyanin (KLH), 2.5 mg/mL, 2 genes related to the biosynthesis of the siderophores mL in 10 mM phosphate buffered saline (PBS)). Then, the pyoverdine and pyochelin. 18,19 Furthermore, the PQS mixture was left stirring for 2 h at RT and overnight at 4 °C promotes the formation of outer membrane vesicles (OMVs) without agitation. The bioconjugates were purified by dialysis due to its remarked lipophilic character. The molecule induces against 0.5 mM PBS (5 × 5 L) and Milli-Q water (1 × 5 L) the curvature of the bacterial membrane and resides as an and stored freeze-dried at −80 °C. A small fraction (20 μL) of integral component, maintaining its biological activity and PQS-BSA was kept for MALDI-TOF-MS analysis, rendering a achieving the communication between bacteria through the hapten density of 17 haptens per molecule of BSA (see Table aqueous media.20 Eventually, the PQS was also found to S1). Hapten densities of PQS-BSA and PQS-KLH were also modulate the host immune response by dysregulating several assessed by fluorescence, measuring the intensity of the defense mechanisms and cytokine expression.21 emission at 445 nm, of the native protein and the The PQS has been found in several body fluids of infected bioconjugates, and interpolating the value on a calibration patients,22,23 such as, for example, in sputum, approximately at curve build with the PQS hapten, rendering the same value of the micrometer range by thin-layer chromatography (TLC) previous MALDI-TOF-MS analysis (see Figure S1). analysis,24 although other more selective techniques found it at Polyclonal Antisera (PAb). See the Supporting Informa- concentrations in the low nanometer range.25 While mass tion. spectrometry has been the most commonly used technology to ELISA. As385/HHQ-BSA ELISA. Microtiter plates were quantify these molecules,26,27 electrochemical28 or bioreporter coated with the HHQ-BSA conjugate in coating buffer (0.04 methods29 have also been reported. These measurements μg/mL, 100 μL/well) overnight at 4 °C and covered with an require complex sample treatment methods such as extraction adhesive plate sealer. The day after, plates were washed with using organic solvents and tedious purification or enrichment PBST (4 × 300 μL/well) and solutions of PQS standards (2 steps to accomplish the necessary detectability. On the other μM to 0.13 nM in PBST-EDTA, 50 μL/well) followed by the hand, only in very few cases, these alkylquinolones have been As385 (dil 1/48000 in PBST-EDTA, 50 μL/well) addition and 4860 https://dx.doi.org/10.1021/acs.analchem.0c04731 Anal. Chem. 2021, 93, 4859−4866 Analytical Chemistry pubs.acs.org/ac Article the microplates were left without agitation for 30 min at room the expression of a series of genes, biosynthesizing virulence temperature. After another washing step, a solution of diluted factors, and favoring biofilm progression. Rapid detection of goat anti-rabbit IgG-HRP (1/6000 in PBST) was added (100 these signaling molecules could give clinicians an early μL/well) and incubated for 30 min at room temperature. The indication of infection or knowledge about the stage of the plates were washed again, and the substrate solution (100 μL) disease. was added and left for 30 min at room temperature in the dark. Over the last few years, biosensors and PoC devices have The enzymatic reaction was stopped by adding 4 N H2SO4 emerged as promising alternatives for a more rapid and solution (50 μL/well), and the absorbance was read at 450 nm. efficient detection of biomarkers of the disease. Among them, Immunoassay Evaluation. Performance of the assays was antibody-based technologies have demonstrated their capa- evaluated through the modification of different physicochem- bility of providing accurate and reliable diagnostic results ical parameters (pH, ionic strength, the presence of a thanks to their great affinity, which allows us to develop highly surfactant (% Tween 20), solubility with addition of organic sensitive technologies and their specificity. However, accom- solvents, competence time, incubation time, or cation plishing high quality antibodies with the necessary affinity and complexation by EDTA) in the competitive step. specificity is a crucial factor, which is mainly dependent on the Cross-Reactivity Studies. HHQ, PQS, and HQNO immunogen and of the strategy used to produce such calibration curves (10 μM to 1 pM) were built in PBST and antibodies. measured on the As385/HHQ-BSA ELISA following the Synthesis of the Immunizing Hapten. The pseudomo- procedure described above. The standard curves obtained were nas quinolone signal (PQS) is an autoinducer used specifically fitted to the four-parameter equation mentioned above, and by P. aeruginosa to regulate virulence (see Figure 1) and for the IC50 value was used to calculate the cross-reactivity according to the following equation CR (%) = IC50(cross reactant)/IC50(analyte) × 100. Analysis of Clinical Isolates Culture Samples. Samples. Clinical isolates were obtained from lower respiratory tract samples, mainly sputum specimens, of patients diagnosed of acute or chronic respiratory infections. Clinical samples were cultured in MacConkey agar, blood agar, and chocolate agar and incubated for up to 5 days at 37 °C. P. aeruginosa isolates were selected and stored frozen at −20 °C. Prior to the analyses, P. aeruginosa isolates were cultured overnight at 37 °C in blood agar plates. The day after, a portion of the grown bacteria was diluted in Mueller−Hinton (MH) culture broth (3 mL) and shaken at 37 °C. When the optical density at 600 nm (OD600) reached a value of 0.2−0.3, a dilution (1/1000 in 20 mL of MH) was performed and the resulting solution was shaken at 37 °C. Aliquots were taken at the selected times for measurement of OD600, for colony forming unit (CFU) calculation, and measurement of the PQS by the ELISA. The Figure 1. Retrosynthetic scheme for the synthesis of the PQS hapten rest of the isolates were grown using the same experimental (8), analogous to the 2-heptyl-3-hydroxy-4(1H)-quinolone (PQS) procedure, yet the aliquots were extracted just at a selected quorum sensing molecule of the pqs system from P. aeruginosa. The time of 8 h. PQS concentrations measured by the ELISA were hapten was synthesized through a five-step synthetic pathway from 3- expressed as PQS immunoreactivity equivalents (IRequiv.) due (4-aminophenyl)propanoic acid and methyl 3-oxodecanoate. to the potential specific interferences caused by other alkylquinolones of different chain lengths potentially present this reason constitutes an excellent target to develop new in the culture media. diagnostic and therapeutic strategies. Due to its small size, it is Analyses of the Samples Using As385/HHQ-BSA ELISA. not immunogenic by itself, for which a reason-appropriate Culture broth samples were analyzed following the same immunizing hapten design was performed to ensure the procedure as described above except for coating the plates with production of high affinity antibodies that would establish HHQ-BSA at 0.04 μg/mL (in coating buffer, 100 μL/well) and strong non-covalent interactions, such as electrostaticity, diluting As385 64,000 times with PBST-EDTA. The standards hydrogen bonds, hydrophobicity, and van der Waals forces, used to build the standard curves were prepared in MH broth with our target.32,33 The quinolone ring and the characteristic diluted 10 times with PBST-EDTA, and the samples were catechol moiety of the native PQS structure were considered diluted 10 times with PBST-EDTA before measuring them the most important epitopes, with the capability of establishing with the ELISA. To assess the accuracy of the method, blind such kinds of noncovalent interactions. Therefore, an spiked samples using diluted MH culture broth (1/10) were immunizing hapten was designed incorporating a spacer arm prepared and measured using the above-described ELISA for at the C-6 position to promote the exposure of the moiety three different days and three replicates within the same day. containing these functional groups. The end of the spacer arm The results are expressed as the mean of all replicates. ■ was provided with a carboxylic acid, which eventually allowed the conjugation of the hapten to the lysine residues of an RESULTS AND DISCUSSION immunogenic protein through the use of orthogonal chemistry Understanding quorum sensing (QS) is key to apprehend (active ester, anhydride mixed, etc.).34 about pathogenesis. QS depends on specific signaling The synthesis of the PQS hapten was inspired from a molecules or autoinducers, which are responsible for triggering strategy previously described by Pesci and co-workers35 based 4861 https://dx.doi.org/10.1021/acs.analchem.0c04731 Anal. Chem. 2021, 93, 4859−4866 Analytical Chemistry pubs.acs.org/ac Article on two sequential modifications of the C-3 position in the PQS refer to the values in the diluted sample. The concentrations biological precursor, 2-heptyl-4-quinolone or HHQ. Hapten are expressed in nanometers, and the data shown correspond synthesis (see Figure 1) started with esterification of 3-(4- to the average of three different days using at least 2 well/ aminophenyl)propanoic acid to obtain the protected aniline replicates per concentration. derivative (9). The aniline derivative was subsequently reacted The assay was able to work at pH values between 4.5 and by amination reaction with β-ketoester (3) in the presence of a 9.5, although a slightly better performance was observed at pH catalytic amount of p-toluene sulfonic acid to obtain the 7.5 based on its detectability and maximum absorbance. The intermediate (10). Afterward, the quinolone derivative (11) addition of an organic solvent such as DMSO did not produce was prepared through Conrad−Umpach thermic cyclization in a significant enhancement; although the Amax increased, at diphenyl ether at 270 °C. Formylation of the C-3 position concentrations greater than 2−5%, the detectability decreased. through the Duff reaction36 allowed us to obtain the quinolone On the other hand, the sensitivity was dramatically affected by derivative (12) as the precursor of the desired hapten. Tween 20. A substantial signal enhancement was observed Concomitant hydrolysis of the methyl ester group occurred; when Tween 20 was not added to the assay buffer (see Figure however, this fact no longer affected the synthetic procedure. S2), while the IC50 value was significantly lower (better Finally, oxidation of the C-3 position through Dakin reaction detectability) than under the standard conditions using a buffer allowed us to obtain the PQS hapten (13). This reaction relies with 0.05% of this detergent. However, since it is known that on a characteristic hydrogen peroxide nucleophilic addition the presence of a small amount of Tween 20 helps in that promotes the [1,2]-aryl migration and elimination, minimizing nonspecific interactions and solubilizing the subsequently obtaining an hydroxyl group directly bonded to analyte, it was decided to keep a concentration of just the previously formylated position.37 Overall, the pure 0.01%. The detectability was slightly better at conductivity proposed PQS proposed hapten (13) was obtained in five values greater than 15 mS/cm; however, the Amax decreased synthetic steps with an overall yield of 16%. substantially, and for this reason, it was decided to keep the The PQS hapten was used to prepare KLH and BSA protein conductivity at 15 mS/cm for further experiments. Further- bioconjugates following the anhydride mixed method. This more, it was added to the assay buffer (0.1 mM) of EDTA as a method relies on the carboxylic acid activation using a result of the accuracy experiments (explained in the section chloroformate and a sterically hindered base, forming an below). anhydride that afterward reacts with the lysine residues present ELISA Evaluation and Characterization. Under these in the protein.38 Hapten bioconjugates were obtained with a conditions, the assay showed an LOD of 0.17 ± 0.01 nM and a high density ratio (PQS-BSA ≈ 17; PQS-KLH ≈ 116−130), as dynamic range compressed between 0.53 ± 0.04 and 24.37 ± demonstrated by MALDI-TOF-MS and fluorescence analysis 3.18 nM (see Table 1 and Figure 2), which is below the (Table S1 and Figure S1). ELISA for PQS Detection. Antisera (As385, As386, and Table 1. Features of the As385/HHQ-BSA ELISA for the As387) were increased against PQS-KLH in female New Detection of the PQS Zealand white rabbits. The avidity of obtained As for the BSA conjugates (HHQ-BSA and PQS-BSA) and the competition of PBST MH diluted 1/10 the PQS for the binding to the antibodies were assessed by Amin 0.03 ± 0.01 0.02 ± 0.01 combined experiments, comparing the absorbance of the assay Amax 1.45 ± 0.04 1.11 ± 0.03 in the presence and absence of the analyte. After a preliminary slope −0.72 ± 0.01 −0.73 ± 0.11 selection, the most promising combinations were analyzed by IC50 3.87 ± 0.53 6.05 ± 0.16 two-dimensional titration experiments. The first set of dynamic range 0.53 ± 0.04 to 24.37 ± 3.18 0.97 ± 0.25 to 36.16 ± 6.20 experiments allowed us to select the best As/bioconjugate LOD 0.17 ± 0.01 0.36 ± 0.14 combinations for competitive assays based on the largest R2 0.997 ± 0.002 0,995 ± 0,003 reduction of the signal when the PQS was present in different combinations, while the second one allowed us to select the average concentrations found in P. aeruginosa bacterial cultures concentrations of the immunoreagents providing the best (μM) and in the same range of concentrations (low nanometer ratio/noise signal. As a result, four different As/bioconjugate range) compared to the ones found in sputum samples by combinations were evaluated in a competitive format (see Abdalla and co-workers.25,26,39 Furthermore, the detectability Table S2), from which the combination of As385/HHQ-BSA achieved with the As385/HHQ-BSA ELISA is greater than was selected for further studies due to the higher detectability those obtained with other techniques such as those based on in further experiments. bioreporters29,40 or electrochemical methods.28 Regarding LC- The performance of the assay under different physicochem- MS/MS methods, the ELISA reported here allows us to obtain ical conditions such as the effect of time, pH, concentration of higher or similar detectability without including additional a non-ionic surfactant, ionic strength, or the presence of an preconcentration steps. These methods require the use of solid organic solvent was evaluated (see Figure S2). A significant phase extraction and/or organic solvent extraction steps, variation of the assay features by introducing a preincubation ending up in a 20 times concentration factor of the sample step was not observed, and the optimum competition time was for obtaining the reported LOQs.22,39 found to be 30 min. The PQS is the most important and active signaling The concentrations of the BSA conjugate and As dilution molecule in the P. aeruginosa pqs QS system. However, other used in the assay run in PBST-EDTA buffer were 0.04 μg/mL quinolones produced by this pathogen have also an important and 1/48,000, respectively. In the case of the assay run in MH role in communication or interspecies interaction. The other diluted 1/10, the concentrations of the BSA conjugate and As two molecules are HHQ, the precursor of the PQS in the dilution were 0.04 μg/mL and 1/64,000, respectively. The biosynthetic pathway and involved in the signaling process, parameters and features of the MH 1/10 curve correspond and and HQNO, an N-oxide quinolone able to inhibit the electron 4862 https://dx.doi.org/10.1021/acs.analchem.0c04731 Anal. Chem. 2021, 93, 4859−4866 Analytical Chemistry pubs.acs.org/ac Article Figure 2. (A) Calibration curves of the As385/HHQ-BSA ELISA for the detection of the PQS run in buffer (PBST-EDTA) and in 1/10 diluted MH broth, under the conditions established (see Table 1). Each calibration point was measured in triplicates on the same ELISA plate, and the results show the average and standard deviation of analysis made on three different days. (B) Matrix effect of MH broth undiluted and diluted 2, 5, 10, and 20 times with PBST on the As385/HHQ-BSA ELISA. The calibration curves were run using the conditions established for the assay in PBST. Modification of the assay conditions allowed us to achieve similar immunoassay features as when the assay was run in buffer (see Table 1 and panel A). The results shown are the average and standard deviations of analysis made on two different days measured by duplicates each day. transport chain in other bacteria showing remarkable production profile of QS molecules such as the PQS can vary antistaphylococcal properties.41 The only structural difference among strains coming from different infection types or between HHQ, HQNO, and the PQS is one oxidized position infection status3 and might have clinical relevance. However, (C-3 or N-1), the reason why the recognition pattern by the such types of samples may contain a variety of components As385/HHQ-BSA ELISA had to be assessed. Despite the from the media or from the bacteria, which may interfere with similarity between the different molecules, the percentage of the assay. For this reason, on the first instance, the matrix effect cross-reactivity was less than 2% for HQNO (IC50 = 236.2 caused by measurement in Mueller−Hinton (MH) culture nM) and about 13% in the case of HHQ (IC50 = 27.2 nM) broth, frequently used to grow P. aeruginosa strains, was (see Table S5). Moreover, additional specificity experiments evaluated. These media contain nitrogenous compounds, were performed in which calibration curves of the PQS were vitamins, starch, and amino acids, which added to the products built under the presence of constant concentrations of either released by the bacteria itself such as virulence factors, HHQ or HQNO (Figure S3). It allowed us to evaluate the proteins, and other exo-products, which could make the potential cross-reactivity for each point of the calibration curve sample quite complex. To estimate the extension of this effect, at different concentrations of cross reactants. The experiments calibration curves were built in MH culture media, at different showed that, indeed, the percentages of cross-reactivity were dilution factors, and measured with the As385/HHQ-BSA not constant and diminished when the concentrations of the ELISA. The resulting calibration curves and the effect caused PQS increased. Hence, the ELISA developed recognized a by MH in the assay performance can be seen in Figure 2B. The much better PQS despite the great structural similarities of the measurement in MH produced a substantial increase of the other two autoinducer molecules of the pqs system, which maximum signal and a decrease in the detectability of the pointed to the possibility to quantify specifically the PQS from assay. This effect could be diminished by increasing the samples where all three quinolones will be present. Specificity dilutions of the MH media with the assay buffer; however, the studies using other molecules structurally related (i.e., same features as the assay run in buffer could not be reached quinolone-type antibiotics ciprofloxacin and norfloxacin) and unless the MH broth was diluted at least 20 times, which non-structurally related (i.e., the virulence factor pyocyanin or compromised too much the detectability for our purposes. In IQS, 2-(2-hydroxylphenyl)-thiazole-4-carbaldehyde, the main light of this behavior, two-dimensional titration experiments signaling molecule of the iqs QS system), potentially present in were carried out using MH media diluted 10 times with the clinical or bacterial culture samples, were also performed. In all assay buffer to set up new assay conditions in respect to the cases, the As385/HHQ-BSA ELISA did not respond to concentration of immunoreagents. As it can be seen in Figure concentration values up to 10 μM of all mentioned molecules, 2A, the assay run in MH diluted 1/10 provided very good and thus, the CR was assumed to be less than 0.01% for all of features in a similar range to the ones obtained in buffer. The them. IC50 value was 6.05 ± 0.16 nM, while the obtained LOD was As385/HHQ-BSA ELISA intra-plate, inter-plate, and inter- 0.36 ± 0.14 nM (Table 1), which taking into account, the day precision were assessed, as it can be observed in Table S3, dilution factor would turn into 60.5 ± 1.6 and 3.6 ± 1.4 nM, where the percentage of the coefficient of variation (%CV) at respectively. The PQS is normally found in the micrometer three different concentrations is shown (IC20, IC50, and IC80). range39 in P. aeruginosa cultures, and thus, the decrease in The reproducibility is only slightly lower when measuring detectability caused by the MH dilution factor did not have to concentrations close to the limit of quantification during mean a negative impact on the PQS production kinetics different days, but even at those very low concentration values, assessment, allowing the quantification from a considerably the %CV is below near 10% on intra-plate and inter-plate low concentration. studies. Accuracy of the assay was evaluated by preparing blind ELISA Implementation for Clinical Isolate Analysis. spiked samples in MH culture broth and measuring them with With these results, as a pilot study, the ELISA was used to the developed ELISA. Interestingly, the results showed a evaluate the PQS release profile on culture samples where P. critical underestimation of the spiked PQS concentrations aeruginosa was grown. The total amount of exo-products and (Figure 3). We hypothesized that this effect might be caused 4863 https://dx.doi.org/10.1021/acs.analchem.0c04731 Anal. Chem. 2021, 93, 4859−4866 Analytical Chemistry pubs.acs.org/ac Article Figure 3. Results from the accuracy study. The graph shows the linear regression analysis of the concentration of the PQS spiked in MH broth and the concentration measured with the As385/HHQ-BSA ELISA. Assays were run in diluted MH culture media 1/10 using PBST (blue line) and PBST-EDTA (red line). Each calibration point was measured in triplicates on the same ELISA plate, and the results show the average and standard deviation of analysis made on three different days. by the described chelating properties of the PQS. Thus, it is known that the role of PQS chelating ferric iron (Fe3+) in a non-deliverable manner19 causes an iron starvation response and stimulates a concentration-dependent increase production of iron scavenging siderophores such as pyoverdine and pyochelin. PQS-Fe3+ has been found to accumulate in the membrane of the cell, acting as a storage molecule. Considering such behavior, the PQS might be forming complexes with the cations present in MH culture broth. To probe this hypothesis, the calibration curve was run in the presence of different concentrations of ethylenediaminetetra- acetic acid (EDTA), a powerful ion chelator able to displace Figure 4. Bacterial growth, expressed as OD 600 and PQS the divalent cation complexation by the PQS (Figure S4). The immunoreactivity equivalents (IRequiv.) measured in MH broth addition of EDTA caused a substantial increase of the media where P. aeruginosa clinical isolates PAAI6 and PACI6 were maximum absorbance; nevertheless, it produced an increase cultured. Samples were taken at the selected times and measured using the As385/HHQ-BSA ELISA. Each calibration point was in detectability. In light of these results, we assessed the effect measured in triplicates on the same ELISA plate, and the results show of EDTA in the buffer (PBST-EDTA) on the assay accuracy. the average and standard deviation of analysis made on two different As it is shown in Figure 3, EDTA produced a significant days. improvement of the accuracy achieving a much better correlation between spiked concentrations and the value measured with the ELISA. These results indicate that the patients suffering from P. aeruginosa infection with distinct PQS in the form of a complex is not recognized by the severity degrees. For this purpose, clinical isolates from 12 antibodies in the same manner as in the free form. Thus, patients with acute or chronic infections were grown in MH at subsequent studies with the As385/HHQ-BSA ELISA were 37 °C. After 8 h (time chosen as a compromise based on the performed in the presence of EDTA in assay buffer. above-described behavior), sample aliquots were taken and With these results, we addressed the investigation of the measured with the PQS ELISA developed. As it can be PQS production profile of clinical isolate growth on MH broth observed in Figure 5, high PQS IRequiv. values (0.5−3 μM) media. Since previously performed experiments pointed at a were measured on the culture media from isolates belonging to different production profile of HHQ, the PQS biosynthetic patients (1−5) with an acute infection, while the PQS IRequiv. precursor, depending on whether the clinical isolates belonged values of the isolates from patients (8−12) with a chronic to patients with a chronic or an acute infection, and the PQS infection were much lower (nanometer range) or even below profiles of the release of isolates from patients with an acute the LOD under conditions used. (PAAI6) and a chronic (PACI6) P. aeruginosa airway infection, In light of this outcome, it seems clear that the concentration respectively, were initially measured. of QS molecules released to the media can be correlated with Interestingly, similarly to HHQ, PQS quantifiable levels the type of infection or the disease status of the patient. It has were reached after 5 and 12 h of growth for the clinical isolates been often reported that the behavior of strains inducing acute PAAI6 and PACI6, respectively, while after 48 h, the PQS or chronic infections is substantially different in terms of IRequiv. value of PAI6 was higher than that of PAAI6 (see virulence and exo-product release caused by an adaptation to Figure 4), as already observed for HHQ. Moreover, the PQS the hostile host environment. Thus, on chronic lung infections, release correlated to the growth, as evidenced by the OD600 P. aeruginosa adapts to the host environment by evolving and the calculated CFUs. The decrease in the OD600 at 48 h of toward a state of reduced bacterial invasiveness that favors growth observed in PAAI6 might be explained by the bacterial persistence without causing overwhelming host injury. prolonged exposure to high concentrations of self exo-products In that respect, the results found with the here reported PQS with lytic activity, such as pyocyanin and the PQS.42−44 ELISA are in agreement with this situation. The low levels of The next set of experiments was seeking to demonstrate that the PQS released by the isolates from chronic infected patients this was a general behavior for clinical isolates coming from would justify a potential lower production of virulence factors. 4864 https://dx.doi.org/10.1021/acs.analchem.0c04731 Anal. Chem. 2021, 93, 4859−4866 Analytical Chemistry ■ pubs.acs.org/ac Article ASSOCIATED CONTENT * sı Supporting Information The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.analchem.0c04731. Immunogen and coating antigen preparation, compet- itive assay selection, physicochemical parameter opti- mization, cross-reactivity studies, and characterization clinical isolates (PDF) Figure 5. PQS IRequiv. recorded from a collection of clinical isolates ■ AUTHOR INFORMATION Corresponding Author from patients with different clinical profiles. Samples were grown in M. Pilar Marco − Nanobiotechnology for Diagnostics (Nb4D), MH broth for 8 h, and the aliquots taken were diluted five times with Department of Surfactants and Nanobiotechnology, Institute PBST-EDTA prior to the ELISA analyses. Clinical isolates 1−5 were for Advanced Chemistry of Catalonia (IQAC) of the Spanish obtained from patients undergoing acute infection, and isolates 8−12 were obtained from patients undergoing chronic infection. Isolate Council for Scientific Research (CSIC), 08034 Barcelona, number 7 corresponds to the reference strain PAO1. The reference Spain; CIBER de Bioingeniería, Biomateriales y number of clinical isolates can be found in Table S7. Each calibration Nanomedicina (CIBER-BBN), 08034 Barcelona, Spain; point was measured in triplicates on the same ELISA plate, and the Phone: +34 93 4006184; Email: pilar.marco@cid.csic.es results show the average and standard deviation of analysis made on two different days. Authors Enrique J. Montagut − Nanobiotechnology for Diagnostics (Nb4D), Department of Surfactants and Nanobiotechnology, Institute for Advanced Chemistry of Catalonia (IQAC) of the From a diagnostic perspective, in addition to identifying the Spanish Council for Scientific Research (CSIC), 08034 microorganisms causing the disease, being able to distinguish Barcelona, Spain; CIBER de Bioingeniería, Biomateriales y between a chronic and an acute infection is very relevant to Nanomedicina (CIBER-BBN), 08034 Barcelona, Spain; apply appropriate therapeutic strategies and improve disease orcid.org/0000-0003-4418-0018 management. M. Teresa Martin-Gomez − Microbiology Department, Vall ■ d’Hebron University Hospital (VHUH), 08035 Barcelona, Spain; Genetics and Microbiology Department, Universitat CONCLUSIONS Autònoma de Barcelona (UAB), 08193 Barcelona, Spain In this work, for the first time, antibodies against the most Complete contact information is available at: relevant and biologically active quinolone from the pqs QS https://pubs.acs.org/10.1021/acs.analchem.0c04731 system of P. aeruginosa have been reported. The produced antibodies were used to develop the indirect competitive Notes As385/HHQ-BSA ELISA for the quantification of the PQS, The authors declare no competing financial interest. achieving a detectability that is within the range of concentrations found in biological and clinical matrices. The results obtained when measuring the PQS IRequiv. released by ■ ACKNOWLEDGMENTS This work has been funded by the Ministry of Economy and clinical isolate growth on MH culture broth demonstrated the Competitiveness (SAF2015-67476-R and RTI2018-096278-B- ability of the As385/HHQ-BSA ELISA to stratify patients C21) and Fundación Marató de TV3 (TV32018-201825-30- depending on the state of the disease. If this immunochemical 31). The Nb4D group is a consolidated research group (Grup strategy is applied for a higher number of QS molecules and/or de Recerca) of the Generalitat de Catalunya and has the infection biomarkers, then it might provide interesting support from the Departament d’Universitats, Recerca i information for diagnostic or prognostic purposes as well as Societat de la Informació de la Generalitat de Catalunya about the pathogenic strain, the status of the patient, and/or (expedient: 2017 SGR 1441). CIBER-BBN is an initiative the progression of the disease. In such a case, the technique funded by the Spanish National Plan for Scientific and developed in this work could be envisaged as a future Technical Research and Innovation 2013−2016; Iniciativa complement of the standard clinical analysis and help clinicians Ingenio 2010, Consolider Program, and CIBER Actions are on taking decisions about the treatment and the management financed by the Instituto de Salud Carlos III with assistance of infected patients. On the other hand, the obtained limits of from the European Regional Development Fund. E.J.M. wishes to thank the FPI fellowship (BES-2016-076496) from the detection and dynamic ranges led us to consider applying the Spanish Ministry of Economy and Competitiveness. The method for the direct quantification of the PQS in clinical Custom Antibody Service (CAbS) is acknowledged for the respiratory samples from infected patients and its evaluation as assistance and support on production of PQS antibodies. ■ the biomarker of infections. Eventually, the versatility and specificity of antibodies would also allow us to implement the REFERENCES developed technique in a wide variety of sensors and/or PoC (1) Horcajada, J. P.; Montero, M.; Oliver, A.; Sorli, L.; Luque, S.; devices, supporting the rapid and straightforward diagnostic of Gomez-Zorrilla, S.; Benito, N.; Grau, S. Clin. Microbiol. Rev. 2019, 32, infections caused by P. aeruginosa and subsequently improving No. e00031-19. the management of the patients. (2) Lipuma, J. J. Clin. Microbiol. Rev. 2010, 23, 299−323. 4865 https://dx.doi.org/10.1021/acs.analchem.0c04731 Anal. Chem. 2021, 93, 4859−4866 Analytical Chemistry pubs.acs.org/ac Article (3) Valentini, M.; Gonzalez, D.; Mavridou, D. A.; Filloux, A. Curr. (31) Montagut, E. J.; Vilaplana, L.; Martin-Gomez, M. T.; Marco, M. Opin. Microbiol. 2018, 41, 15−20. P. ACS Infect. Dis. 2020, 6, 3237−3246. (4) Centers for Disease Control and Prevention, N. C. f. E. a. Z. I. D. (32) Ballesteros, B.; Barceló, D.; Sanchez-Baeza, F.; Camps, F.; N. Division of Healthcare Quality Promotion (DHQP) Biggest Marco, M.-P. Anal. Chem. 1998, 70, 4004−4014. Threats and Data: 2019 AR Threats Report; 2019, Last update June 18, (33) Pinacho, D. G.; Sánchez-Baeza, F.; Marco, M.-P. Anal. Chem. 2020. 2012, 84, 4527−4534. (5) E. Tacconelli, E. C.; Savoldi, A.; D. Kattula, F. Burkert, Global (34) Hermanson, G. T. Bioconjugate Techniques, (2nd Edition), priority list of antibiotic-resistant bacteria to guide research, discovery Elsevier; 2008. and development of new antibiotics. WHO 2017. (35) Pesci, E. C.; Milbank, J. B. J.; Pearson, J. P.; McKnight, S.; (6) Pang, Z.; Raudonis, R.; Glick, B. R.; Lin, T.-J.; Cheng, Z. Kende, A. S.; Greenberg, E. P.; Iglewski, B. H. Proc. Natl. Acad. Sci. U. Biotechnol. Adv. 2019, 37, 177−192. S. A. 1999, 96, 11229−11234. (7) Burns, J. L.; Rolain, J. M. J. Cystic Fibrosis 2014, 13, 1−9. (36) Masurier, N.; Moreau, E.; Lartigue, C.; Gaumet, V.; Chezal, J.- (8) Bittar, F.; Rolain, J.-M. Clin. Microbiol. Infect. 2010, 16, 809−820. M.; Heitz, A.; Teulade, J.-C.; Chavignon, O. J. Org. Chem. 2008, 73, (9) Jung, J. Y.; Yoon, H. K.; An, S.; Lee, J. W.; Ahn, E. R.; Kim, Y.-J.; 5989−5992. Park, H.-C.; Lee, K.; Hwang, J. H.; Lim, S.-K. Sci. Rep. 2018, 8, 10852. (37) Yaremenko, I. A.; Vil’, V. A.; Demchuk, D. V.; Terent’ev, A. O. (10) Chen, J.-W.; Lau, Y. Y.; Krishnan, T.; Chan, K.-G.; Chang, C.- Beilstein J. Org. Chem. 2016, 12, 1647−1748. Y. Front. Microbiol. 2018, 9, 1104. (38) Chaudhary, A.; Girgis, M.; Prashad, M.; Hu, B.; Har, D.; Repič, (11) Reali, S.; Najib, E. Y.; Treuerné Balázs, K. E.; Chern Hui Tan, O.; Blacklock, T. J. Tetrahedron Lett. 2003, 44, 5543−5546. A.; Váradi, L.; Hibbs, D. E.; Groundwater, P. W. RSC Adv. 2019, 9, (39) Ortori, C. A.; Dubern, J.-F.; Chhabra, S. R.; Cámara, M.; 21486−21497. Hardie, K.; Williams, P.; Barrett, D. A. Anal. Bioanal. Chem. 2011, (12) Mohan, A.; Harikrishna, J. Indian J. Med. Res. 2015, 141, 271− 399, 839−850. (40) Fletcher, M. P.; Diggle, S. P.; Crusz, S. A.; Chhabra, S. R.; 273. Cámara, M.; Williams, P. Environ. Microbiol. 2007, 9, 2683−2693. (13) Martinez-Sagasti, F.; Velasco-López, E.; Domingo-Marín, S.; (41) Hotterbeekx, A.; Kumar-Singh, S.; Goossens, H.; Malhotra- Gil-Perdomo, J. M. Rev. Esp. Quimioter. 2018, 31, 43−46. Kumar, S. Front. Cell. Infect. Microbiol. 2017, 7, 106. (14) Miller, M. B.; Bassler, B. L. Annu. Rev. Microbiol. 2001, 55, (42) D’Argenio, D. A.; Calfee, M. W.; Rainey, P. B.; Pesci, E. C. J. 165−199. Bacteriol. 2002, 184, 6481−6489. (15) Lee, J.; Zhang, L. Protein Cell 2015, 6, 26−41. (43) Das, T.; Manefield, M. PLoS One 2012, 7, No. e46718. (16) Lin, J.; Cheng, J.; Wang, Y.; Shen, X. Front. Cell. Infect. (44) Hall, S.; McDermott, C.; Anoopkumar-Dukie, S.; McFarland, Microbiol. 2018, 8, 230. A. J.; Forbes, A.; Perkins, A. V.; Davey, A. K.; Chess-Williams, R.; (17) Häussler, S.; Becker, T. PLoS Pathog. 2008, 4, No. e1000166. Kiefel, M. J.; Arora, D.; Grant, G. D. Toxins 2016, 8, 236. (18) Diggle, S. P.; Matthijs, S.; Wright, V. J.; Fletcher, M. P.; Chhabra, S. R.; Lamont, I. L.; Kong, X.; Hider, R. C.; Cornelis, P.; Cámara, M.; Williams, P. Chem. Biol. 2007, 14, 87−96. (19) Popat, R.; Harrison, F.; da Silva, A. C.; Easton, S. A. S.; McNally, L.; Williams, P.; Diggle, S. P. Proc. R. Soc. B 2017, 284, 20170200. (20) Mashburn-Warren, L.; Howe, J.; Garidel, P.; Richter, W.; Steiniger, F.; Roessle, M.; Brandenburg, K.; Whiteley, M. Mol. Microbiol. 2008, 69, 491−502. (21) Kim, K.; Kim, S.-H.; Lépine, F.; Cho, Y.-H.; Lee, G. R. Microb. Pathog. 2010, 49, 174−180. (22) Barr, H. L.; Halliday, N.; Barrett, D. A.; Williams, P.; Forrester, D. L.; Peckham, D.; Williams, K.; Smyth, A. R.; Honeybourne, D.; Whitehouse, J. L.; Nash, E. F.; Dewar, J.; Clayton, A.; Knox, A. J.; Cámara, M.; Fogarty, A. W. J. Cystic Fibrosis 2017, 16, 230−238. (23) Barr, H. L.; Halliday, N.; Cámara, M.; Barrett, D. A.; Williams, P.; Forrester, D. L.; Simms, R.; Smyth, A. R.; Honeybourne, D.; Whitehouse, J. L.; Nash, E. F.; Dewar, J.; Clayton, A.; Knox, A. J.; Fogarty, A. W. Eur. Respir. J. 2015, 46, 1046−1054. (24) Collier, D. N.; Anderson, L.; McKnight, S. L.; Noah, T. L.; Knowles, M.; Boucher, R.; Schwab, U.; Gilligan, P.; Pesci, E. C. FEMS Microbiol. Lett. 2002, 215, 41−46. (25) Abdalla, M. Y.; Hoke, T.; Seravalli, J.; Switzer, B. L.; Bavitz, M.; Fliege, J. D.; Murphy, P. J.; Britigan, B. E. Infect. Immun. 2017, 85, No. e00176-17. (26) Kushwaha, M.; Jain, S. K.; Sharma, N.; Abrol, V.; Jaglan, S.; Vishwakarma, R. A. ACS Chem. Biol. 2018, 13, 657−665. (27) Lanni, E. J.; Masyuko, R. N.; Driscoll, C. M.; Aerts, J. T.; Shrout, J. D.; Bohn, P. W.; Sweedler, J. V. Anal. Chem. 2014, 86, 9139−9145. (28) Buzid, A.; Shang, F.; Reen, F. J.; Muimhneacháin, E. Ó .; Clarke, S. L.; Zhou, L.; Luong, J. H. T.; O’Gara, F.; McGlacken, G. P.; Glennon, J. D. Sci. Rep. 2016, 6, 30001. (29) Müller, C.; Fetzner, S. Appl. Microbiol. Biotechnol. 2013, 97, 751−760. (30) Reen, F. J.; Clarke, S. L.; Legendre, C.; McSweeney, C. M.; Eccles, K. S.; Lawrence, S. E.; O’Gara, F.; McGlacken, G. P. Org. Biomol. Chem. 2012, 10, 8903−8910. 4866 https://dx.doi.org/10.1021/acs.analchem.0c04731 Anal. Chem. 2021, 93, 4859−4866