Disrupting quorum sensing as a strategy to inhibit bacterial virulence in human, animal, and plant pathogens

Abstract The development of sustainable alternatives to conventional antimicrobials is needed to address bacterial virulence while avoiding selecting resistant strains in a variety of fields, including human, animal, and plant health. Quorum sensing (QS), a bacterial communication system involved in noxious bacterial phenotypes such as virulence, motility, and biofilm formation, is of utmost interest. In this study, we harnessed the potential of the lactonase SsoPox to disrupt QS of human, fish, and plant pathogens. Lactonase treatment significantly alters phenotypes including biofilm formation, motility, and infection capacity. In plant pathogens, SsoPox decreased the production of plant cell wall degrading enzymes in Pectobacterium carotovorum and reduced the maceration of onions infected by Burkholderia glumae. In human pathogens, lactonase treatment significantly reduced biofilm formation in Acinetobacter baumannii, Burkholderia cepacia, and Pseudomonas aeruginosa, with the cytotoxicity of the latter being reduced by SsoPox treatment. In fish pathogens, lactonase treatment inhibited biofilm formation and bioluminescence in Vibrio harveyi and affected QS regulation in Aeromonas salmonicida. QS inhibition can thus be used to largely impact the virulence of bacterial pathogens and would constitute a global and sustainable approach for public, crop, and livestock health in line with the expectations of the One Health initiative.


Introduction
The rise of antimicr obial r esistance (AMR) has led the international community to reinforce hygiene measures to limit the spread of infections.Priorities include reducing the use of antimicr obials, whic h is a major issue, as their excessive use in livestock, human health, and a gricultur e induces serious AMR concerns worldwide.In this context, the World Health Organization (WHO) has initiated a "Global Action Plan" to tackle AMR (GAP-AMR) (Harbarth et al. 2015, World Health Organization 2015 ).Developing sustainable and global approaches to reducing bacterial infections with neither toxic side effects nor associated resistance is extr emel y c hallenging.Moving fr om conv entional c hemical a ppr oac hes to nov el bio-based alternativ es is thus an attractiv e pr ospect.Bacteriopha ges, antimicr obial peptides, biological contr ol, and enzymes ar e a ppealing candidates for a v ariety of applications (Nigam et al. 2014, Khardori et al. 2020 ).Enzymes in particular may be of interest, as their catalytic performances may allow them to be used in small quantities while having a br oad effect.Ne v ertheless, to limit the emer gence of AMR, r ethinking how to inhibit bacterial virulence remains a major concern.Over the last two decades, enzymes able to disrupt quorum sensing (QS) have gained considerable interest with regards to limiting bacterial infections while not challenging bacterial sur-vi val (Rém y et al. 2016 ).QS is a bacterial comm unication pr ocess r el ying on the pr oduction and detection of autoinducible signaling molecules (AIs) that bacteria use to coordinate behaviors in a cell-density-dependent manner, including infection-related beha viors .AIs ar e c hemicall y div erse, including acyl-homoserine lactones (AHLs), furanosyl diester AI-2, peptides, and hormones (e.g. e pine phrine) that can be recognized by one or several bacteria depending on the signal (Rémy et al. 2018 ).Among AIs, special attention has been paid to AHLs, as they ar e widel y used by Gr am-negativ e bacteria, including human, animal, and plant pathogens.Enzymes such as lactonases, acylases, and oxidoreductases have been studied for their ability to interfere with AHLbased QS, a strategy referred to as quorum quenching (QQ).AHLs are signaling molecules presenting a conserved lactone ring and an acyl chain moiety that may vary in length (from C 4 to C 18 ), substitutions (e.g.hydroxyl and carbonyl), and degree of saturation.Basicall y, AHL-based QS r elies on a synthase, usuall y noted "I," that produces one or several AHLs that diffuse freely through cell membranes, and a sensor regulator, noted "R" which, upon AHL binding, becomes active and induces the expression of numerous genes that often include the synthase I, thus generating autoinduction (P a penfort and Bassler 2016 ).As a broad panel of AHLs is involved in QS regulation, QQ enzymes (QQE) require Here, we aimed to evaluate Sso Pox lactonase, a hyperstable enzyme of gr eat inter est for biotechnological applications, on a wide panel of AHL-using bacteria.Our objective was to probe the potential of lactonase technology to deal with bacterial threats in different fields of application.As the development of an alternative to antibiotics and conventional antimicrobials is challenging, this study aims to demonstrate that the same lactonase, and its related variants, can be used as a global alternative in a range of important applications.To this end, four variants of the lactonase Sso Pox, obtained by semirational engineering strategies, wer e consider ed on the basis of their distinct specificities to w ar d AHLs (Billot et al. 2022, Daude et al. 2023 ).As all bacteria do not necessarily use the same AHLs for their respective QS, a wide panel of pathogenic bacteria involved in plant infections ( Bur kholderia glumae , Dic keya c hrysanthemi , Dic keya dadantii , and P ectobacterium carotovorum ), human infections ( Acinetobacter baumannii , Burkholderia cepacia , Pseudomonas aeruginosa ), and animal infections ( Aeromonas salmonicida and Vibrio harveyi ) was selected.The ability of these strains to produce AHLs was c hec ked by the identification of an AHL synthase in their r espectiv e genomes (Table 1 ).
The chosen enzymatic variants sho w ed different abilities to interfere with pathogenic bacteria.Nevertheless, each strain was sensitive to at least one variant, resulting in alteration of QS-related behaviors in these str ains.These r esults suggest that enzymemediated QQ, combining one or mor e v ariants , ma y be considered as a multipurpose approach to strengthening the antimicrobial arsenal, as promoted by the GAP-AMR.Lactonases could thus constitute a sustainable a ppr oac h to addr ess health thr eats in the animal-human-environmental interface in line with the One Health initiative (McEwen andCollignon 2018 , One Health 2024 ).

Strains and bacterial cultures
The following strains were used in this study: NaCl, 10 g l −1 tryptone, 5 g l −1 yeast extract, and pH 7), except for D .dadantii , D .chrysanthemi , and both P. carotovorum , which were pr ecultiv ated in Tryptic Soy Broth (TSB).A culture step was then performed as indicated below, before the different phenotypic assa ys .Chromobacterium violaceum CV026 was cultivated in LB at 37 • C and used as a reporter strain for the detection of AHLs.
Regarding phytopathogenic bacteria, the inoculation of B. glumae from a 16-hour pr ecultur e was carried out at 1:100 dilution in fresh LB.Subsequently, 3 ml of this culture was poured into 12-well plates and was then incubated for 24 hours at 33 • C, with stirring at 350 rpm.The cultures of D .dadantii , D .chrysanthemi , and both P. carotovorum from the DSMZ and Vegepolys collections, wer e performed fr om a 6-hour pr ecultur e diluted at 1:100 in TSB.A quantity of 1.5 ml of the culture was tr ansferr ed to 24-well plates and then incubated for 16 hours at 30 • C, with agitation at 400 rpm.
Finall y, for the v eterinary str ains, cultur es of A. salmonicida wer e performed from a 16-hour pr ecultur e and diluted at 1:1000 in TSB.Cultur es wer e gr own for 24 hours at r oom temper atur e without stirring in a 24-well plate.Vibrio harveyi was pr ecultiv ated in LB for 16 hours and inoculated 1:100 in autoinducer bioassay (AB) medium (Gr eenber g et al. 1979 ) [0.3 M NaCl, 0.05 M MgSO 4 , and 0.2% vitamin-free casamino acids (Difco), adjusted to pH 7.5 with KOH and supplemented with 200 μl of sterile 1 M potassium phosphate (pH 7.0), 200 μl of 0.1 M l -arginine, and 250 μl of glycerol 80% for a final volume of 20 ml].The cultur es wer e tr ansferr ed to 24-well plates and incubated for 24 hours at 30 • C with stirring (350 rpm).
Sso Pox variants were added during the cultures, and their respectiv e concentr ations wer e adjusted according to their specific kinetic parameters .T he v ariant and the concentr ation used for eac h str ain ar e indicated in Table 1 .

Production of SsoPox enzymes and determination of their concentration
Enzyme production was performed as described previously (Hiblot et al. 2013 ).Briefly, Esc heric hia coli BL21 (DE3)-pGr o7/Gr oEL strain containing the plasmid of the Sso Pox variant (W263I, V82I-A266G-A275Y, V82I-A275G, and V82I-K271L) was used.Variants W263I, V82I-A266G-A275Y, V82I-A275G, and V82I-K271 L are hereafter r eferr ed to as A, B, C, and D, r espectiv el y.Pr ecultur es wer e incubated at 37 • C for 16 hours in LB supplemented with antibiotics (ampicillin at 100 μg ml −1 and c hlor amphenicol at 34 μg ml −1 ).Bacteria were then inoculated at 1:100 in ZYP-5052 autoinducer medium supplemented with both antibiotics described above and incubated at 37 • C with 180 rpm agitation.When cultures reached an OD 600 nm between 0.8 and 1, bacteria were induced by adding 0.2% l -arabinose for chaperon production and CoCl 2 at 0.2 mM final concentration, to stabilize the active site of Sso Pox lactonase.The temper atur e was r educed to 23 • C and cultur es wer e gr own for 20 hours with stirring at 180 rpm.Cells were harvested by centrifugation at 4 000 × g for 20 minutes at 10 • C. Pellets were solubilized in lysis buffer (HEPES 50 mM pH 8, NaCl 150 mM, DNAseI 10 μg ml −1, lysozyme 0.25 mg ml − 1 , and PMSF 0.1 mM).Cells in the l ysis buffer wer e stor ed at −80 • C for at least 16 hours .T ha wed cells were sonicated three times for 30 seconds, with an amplitude of 45% (QSonica sonicator Q700 sonicator ®, QSonica, USA).Finall y, cells wer e heated to 80 • C for 30 minutes and then centrifuged to pellet the debris .T he supernatants were collected and saturated with 75% of ammonium sulfate at 4 • C for 16 hours to precipitate Sso P ox.T he enzyme w as pelleted do wn b y centrifugation at 10 000 × g for 15 minutes at 10 • C. The pellets were resuspended in 8 ml of HEPES 50 mM pH 8.0 and NaCl 150 mM buffer and then filtered at 0.8 μm.Ammonium sulfate was r emov ed via desalting (HiPrep 26/10 desalting, ÄKTA pure, GE Healthcare, USA ÄKTA pure).The sample was concentrated using 30 kDa centricon, and then injected into a size-exclusion c hr omatogr a phy column (GF Hiload 16/600 Superdex 75 pg) to purify the enzyme.Enzyme purity was then c hec ked by SDS-PAGE and pr otein concentr ation was measured using a Bradford assay (Bradford 1976 ).

Detection of AHL
Six-well plates containing 3 ml of LB agar supplemented or not supplemented with enzyme, were prepared to identify AHL production.A quantity of 10 μl of bacterial culture was streaked in line on the agar next to another line of 10 μl of the r eporter str ain C. violaceum CV026 from a 16-hour pr ecultur e in LB.This strain produces violacein when exogenous AHLs are detected (McClean et al. 1997 ).Sso Pox action on QS was e v aluated by the abolition of violacein production upon degradation of AHL produced by the tested strains.Plates were incubated at room temperature for 48 hours to observe violacein production by CV026.

Assessment of biofilm formation
Biofilm formation was measured for each strain using the crystal violet staining method (Sigma ®) (Hoffman et al. 2005 ).Inoculation for each strain was performed as detailed above in the section "Strains and bacterial cultures."After incubation for 16 hours ( P .carotovorum and Dickeya sp.), 24 hours ( A. salmonicida , Burkholderia sp., P. aeruginosa , and V. harveyi ) and 48 hours ( A. baumannii ) at the a ppr opriate temper atur es, planktonic cells wer e car efull y r emov ed fr om the 12-or 24-w ell plates.Wells w ere w ashed with phosphate buffered saline (PBS) and completely dried at 37 • C.Then, 3 ml (for 12-well plates) and 2 ml (for 24-well plates) of 0.05% crystal violet was added to stain the attached biofilm, and each plate was stained for 10 minutes.Crystal violet was gently r emov ed, and the wells were washed with PBS.The fixed crystal violet was finally dissolved in 3 or 2 ml of absolute ethanol, depending on the plate, and 200 μl was tr ansferr ed to a 96-well plate.Biofilm formation was quantified by measuring the absorbance at OD 595 nm in a microplate reader.The following classification was applied for the determination of biofilm formation: Absorbances below 0.3 indicate the absence of an attached biofilm, absorbances between 0.3 and 1 r e v eal a weak biofilm formation, and absorbances above 3 are evidence of a strong biofilm (Xu et al. 2016 ).

Measurement of plant cell wall degrading enzymes
Plant cell wall degrading enzymes (PCWDE) assays were performed by measuring the activity of pectate l yase, pol ygalacturonase , and cellulase .
The pectate lyase plates wer e pr epar ed according to a pr e viousl y r eported pr ocedur e (Starr et al. 1977 ).Briefly, the plates wer e composed of 1% (w/v) pol ygalactur onic acid, 1% (w/v) yeast extract, 10% (v/v) 0.1% bromothymol blue, and 1.5% agar in distilled w ater.The solution w as heated, without boiling, until all the components were dissolved, and the pH was adjusted to 7. If necessary, the pH was readjusted after autoclave sterilization.Supernatants fr om ov ernight cultur es tr eated or not tr eated with Sso Pox wer e harvested, and 10 μl of the supernatants was added to the plates and incubated for 4 hours at 37 • C. Pol ygalactur onic acid degr adation halos were revealed by pouring 4 N HCl on the plates to pr ecipitate nondegr aded pol ygalactur onic acid.Pectate l yase activity in the supernatants was assessed by the presence of clear halos in the dish.
Pol ygalactur onase activity measur ement was performed as follows (Cui et al. 2020 ).Briefly, plates were poured with a mixture of 0.5% pol ygalactur onic acid, 0.2% sucrose, 0.2% (NH 4 ) 2 SO 4 , and 1.5% agar.A quantity of 10 μl of cell-free supernatants was spotted on the plates.After incubation for 4 hours at 37 • C, the plates wer e r e v ealed with 4 N HCl.The formation of halos indicated the degr adation of pol ygalactur onic acid by pol ygalactur onase in the supernatants.
Finally, the detection of cellulase activity was measured, as previously described (Cui et al. 2020 ).Plates were prepared with 0.1% carboxymethyl cellulose, 0.8% a gar ose and 0.38% sodium phosphate, and pH was adjusted to 7. A quantity of 10 μl of supernatants was spotted on the plate and after 4 hours of incubation at 37 • C the plates were revealed with 5 minutes incubation at room temper atur e with 0.1% Congo red, follo w ed b y tw o w ashes with 1 M NaCl.Carboxymethyl cellulose degradation was detected by the formation of a white halo, indicating the presence of cellulase activity in the supernatants.

Swarming motility assay
Sw arming motility w as monitored for B. glumae .Briefly, 3 μl from a 24-hour culture was spotted on 0.7% LB agar supplemented or not supplemented with Sso Pox B. 24-well plates were incubated at 30 • C for 48 hours, maintaining a high humidity atmosphere to avoid a gar ose drying.Motility ar eas wer e measur ed using Ima geJ.Experiments were performed in four replicates.

Virulence assay on onion slices
Red onions ( Allium cepa ), purc hased fr om a local market, were used in this assay to assess the virulence of B .glumae.T he onions were disinfected for 15 minutes in a 70% ethanol bath, and thor oughl y rinsed with sterile water.A sterile knife was used to cut the onions into six pieces .T he slices were separated and perforated with tips prior to infection .Bacteria were centrifuged at 10 000 × g for 10 minutes and supernatants were removed.
The onions were then infected with 10 μl of culture adjusted to 0.25 × 10 8 CFU ml −1 with PBS.To e v aluate the effect of Sso Pox B on B .glumae virulence , when indicated, onion slices wer e pr etr eated with 0.09 mg ml −1 of lactonase and left to dry.The onion slices were then infected with bacterial culture.PBS was inoculated as negativ e contr ol.The onion slices were placed on Petri dishes on a filter paper soaked with 3 ml of sterile water and placed in a box containing paper to w el moistened with 30 ml of distilled water to maintain high humidity.The onion slices were incubated in the dark at 30 • C for 5 days (Karki et al. 2012 ).Three onions were used in this assay, and 15 slices were included in each condition.
To quantitativ el y assess virulence, the macer ated tissues wer e r emov ed using a spatula and the percentage of macerated tissues was quantified by weighing the slices before and after removing maceration (Carstens et al. 2018 ).

Cytotoxicity of P. aeruginosa toward J774.1 macr opha ges
J774.1 m urine macr opha ges wer e used to e v aluate the cytotoxicity of P. aeruginosa.These macr opha ges wer e cultiv ated in Roswell Park Memorial Institute (RPMI) 1640 broth (Gibco™ , Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% foetal bovine serum (FBS) and 1% stre ptom ycin and penicillin (Gibco ™ , Thermo Fisher Scientific) and incubated at 37 • C in a humidified atmosphere containing 5% CO 2 .When macrophages r eac hed 75% confluence, cells were harvested in a 50-ml tube and centrifuged for 5 minutes at 750 × g and 4 • C. Cells were resuspended in RPMI supplemented with 10% FBS to r emov e the antibiotics, and then centrifuged a gain.Finall y, cells wer e placed in a 96-well plate at 10 5 cells per well in 100 μl RPMI with 10% FBS.Cells were incubated for 20 hours at 37 • C in a humidified atmosphere containing 5% CO 2 .
Pseudomonas aeruginosa , cultivated as described abo ve , was centrifuged and the pellet was diluted in RPMI broth supplemented with 10% FBS.A quantity of 10 μl of bacteria was added to J774.1 macr opha ges at a multiplicity of infection (MOI) of 10.Supernatant toxicity was also e v aluated by adding 10 μl to the J774.1 macr opha ges.Centrifugation (10 minutes, 500 × g , 4 • C) was performed to bring bacteria into contact with the macr opha ges, before incubating the 96-well plates for 45 minutes at 37 • C with 5% CO 2 .Finall y, supernatants wer e collected after a 5-minute centrifugation step at 200 × g and 4 • C.
Cytotoxicity was assessed by a CyQUANT™ lactate dehydrogenase (LDH) cytotoxicity assay kit (Thermo Fisher Scientific) according to the manufacturer's instructions, by measuring LDH released by macr opha ges dama ged by bacteria.The negativ e control was measured by adding PBS, corresponding to 0% cytotoxicity, and the positiv e contr ol by adding the commercial lysis buffer, corresponding to 100% toxicity, to noninfected macrophage cells.

RN A extr action and QS gene expression measurement
Aeromonas salmonicida was grown as pr e viousl y described, with or without the addition of Sso Pox enzyme.A volume of 200 μl of culture was used for RNA extraction with the RNA PureLink ® mini kit (Thermo Fisher).After extr action, a DNase tr eatment was performed with the TURBO DNA-free TM kit (Thermo Fisher) to elim-

Selection of Sso Pox candidates for a broad QQ potential study
The importance of QQE specificity to efficiently tackle QSregulated bacterial virulence has been thoroughly demonstrated (Koch et al. 2014, Rémy et al. 2020 ), and r e vie w ed (Murugay ah and Gerth 2019 , Billot et al. 2020 , Sikdar andElias 2020 ).These publications underlined that the right QQE may disrupt a precise QS signal with maximum efficacy.Protein eng ineering strateg ies were implemented, particularly on the highly robust Sso Pox, leading to dr astic c hanges in activities to w ar d AHLs (Bzdrenga et al. 2017, Billot et al. 2020 ).Hence, to demonstrate the potential of the QQ strategy to w ar d a wide r ange of pathogens r el ying on differ ent QS systems , four Sso P ox variants , generated by our group (Hiblot et al. 2013, Billot et al. 2022, Daude et al. 2023 ), with distinct specificities on AHL substr ates, wer e car efull y selected and tested fr om a collection of pr e viousl y m utated enzymes.Briefly, the collection was r ationall y designed through site saturation mutagenesis targeting r esidues involv ed in substr ate binding and activ e site flexibility, leading to catalytic activity modulation to w ar d AHLs (Hiblot et al. 2013, Billot et al. 2022 ).Her e, four v ariants wer e selected fr om labor atory scr eenings, r eferr ed to as Sso Pox A-D (A: W263I, B: V82I-A266G-A275Y, C: V82I-A275G, and D: V82I-K271L), and were thoroughl y c har acterized (Table 3 ), r e v ealing (i) impr ov ed degr adation activities to w ar d se v er al AHLs and (ii) distinct and complementary specificities to w ar ds different types of AHLs .Sso P ox A was selected for its high catalytic activity to w ar ds long-chain AHLs such as 3-oxo-C 12 -HSL, with a k cat /K M > 10 5 M −1 s −1 (Hiblot et al. 2013 ).This enzyme has already been shown to efficiently tackle the QSrelated virulence of various pathogens, notably several strains of P. aeruginosa (Hr aiec h et al. 2014, Guendouze et al. 2017, Rémy et al. 2020 ), and C. violaceum (Mion et al. 2021 ) .This enzyme thus constitutes a good candidate to target pathogens using long-chain AHLs.Sso Pox B is a m ultiple v ariant with impr ov ed specificity to w ar ds AHLs carrying a 6-carbon acyl c hain, suc h as C 6 -HSL and 3-oxo-C 6 -HSL (10 4 M −1 s −1 and 10 3 M −1 s −1 , r espectiv el y).This v ariant is the best candidate so far to target these AHLs.Similarly, variant Sso Pox D was selected for its broader specificity towards midr ange AHLs, suc h a C 6 , C 8 , and C 10 acyl chain AHLs .T his enzyme could be qualified as br oad-r ange, since it is the only one to r eac h a k cat /K M value > 10 3 M −1 s −1 on six different AHLs, and up to 10 4 M −1 s −1 to w ar ds C 8 -HSL and 3-o xo-C 10 -HSL.Finally, Sso Po x C was the variant selected to disrupt QS systems r el ying on shortc hain AHLs, suc h as C 4 -HSL.While its catal ytic efficienc y to w ar d C 4 -HSL r emains r elativ el y low (10 3 M −1 s −1 ), it is the best activity reported so far for an Sso Pox variant, the wild-type enzyme being almost inactive to w ar ds this substrate (11.62 M −1 s −1 ) (Hiblot et al. 2013 ).Furthermore, its efficiency at disrupting the QS-regulated phenotypes of a C 4 -HSL model bacterium has been pr e viousl y established (Daude et al. 2023 ).
Taken as a whole, these four variants harbour complementary activity spectra, making it possible to tackle heterogeneous QS systems using different kind of AHLs more accurately and specificall y.Hence, this narr ow selection provides a sound basis for targeting the QS-regulated virulence phenotypes of a large spectrum of micr oor ganisms causing deleterious issues in human health, fish farming, and food crops.

Sso Pox inhibits virulence factors in phytopathogens
Many bacterial strains are of critical interest in agriculture and ornamental plant production, due to their pathogenicity.In particular, Dic keya and P ectobacterium ar e soft r ot-inducing bacteria, present in the top 10 phytopathogenic bacteria (Mansfield et al. 2012 ), damaging a wide variety of crops (Charkowski 2018 ).Other bacterial species are also known to be noxious to specific cultures, for instance B. glumae to rice (Ham et al. 2011 ) .Here, the ability of Sso Pox lactonase to disrupt QS of str ains fr om thr ee r ele v ant genera of phytopathogenic bacteria was evaluated.
Pr e vious studies r eported v arious AHL le v els and types for Dickeya sp. and P. carotovotum spp., such as 3-o xo-C 6 -HSL, 3-o xo-C 8 -HSL, C 6 -HSL, and C 10 -HSL for the former, and 3-o xo-C 6 -HSL, 3-o xo-C 8 -HSL, and C 8 -HSL for the latter (Crépin et al. 2012 ).In this respect, Sso Pox C and Sso Pox D were selected to perform QQ experiments (Table 3 ).The production of AHLs was checked for each selected phytopathogenic bacterium using the reporter strain C. violaceum CV026 (McClean et al. 1997 ).The production of AHLs was detected for each strain, leading to different levels of violacein production (Fig. 1 ).In the presence of Sso P ox variants , no violacein was produced, indicating that the enzyme effectiv el y degr aded the AHLs produced by the phytopathogenic bacteria (Fig. 1 ).
The production of PCWDEs is a common bacterial virulence factor (Toth et al. 2003 ) that can be under AHL-mediated QS control (Barnard et al. 2007, Liu et al. 2008 ).Ther efor e, to e v aluate QQ efficiency, semiquantitative assays were used to detect the activity of three PCWDEs , polygalacturonase , pectate lyase , and cellulase , in the supernatants from cultures of D .dadantii , D .chrysanthemi , P. carotovorum DMSZ30168, and P. carotovorum DV1708-1686 (Fig. 1 ).
The three PCWDE activities were detected in culture supernatants of both Dic keya str ains, r egardless of tr eatment with Sso P ox lactonase , indicating that QQ is not sufficient to abolish the expression of PCWDEs in these strains (Fig. 1 A and B).In addition to AHL-mediated QS, Dic keya sp.ar e known to use another signal known as virulence factor modulating (Vfm), which has been shown to be involved in PCWDE production in some strains (Nasser et al. 2013 , Helman andChernin 2015 ).Ho w e v er, in other Dickey a species , such as D. chrysanthemi , expression of the pel gene w as sho wn to be under the control of C 10 -HSL (Hosseinzadeh et al. 2017 ).T he interpla y between the two QS systems, AHLs and Vfm, has not been fully described and may be strain-dependent.Here, w e sho w ed that AHL degr adation did not lead to a decr ease in PCWDE production, suggesting that another signal dominates the regulation of this phenotype.
In Pectobacterium sp., the three enzymatic activities pectate l yase, pol ygalactur onase, and cellulase were decreased in the culture supernatants after lactonase treatment (Fig. 1 C and D).Specificall y, pectate l yase and pol ygalactur onase activities dr amaticall y decr eased, while cellulase activity was onl y slightl y r educed.This r esult is consistent with pr e vious studies, whic h demonstrated that AHL-mediated QS regulates PCWDE in Pectobacterium , r el ying on gene invalidation and recombinant lactonase expression experiments (Dong et al. 2001, Smadja et al. 2004, Palmer et al. 2011, Kusada et al. 2017 ).Ho w e v er, this is the first demonstration that PCWDE production by Pectobacterium sp. can be quenched by the exogenous addition of a purified lactonase.
To extend these assa ys , biofilm formation, another commonly studied virulence factor, was assessed for all Dickeya and Pectobacterium strains.In previous studies, biofilm and pellicle production b y Dicke ya and Pectobacterium w as sho wn to vary betw een strains and culture conditions (Mee-Ngan et al. 2005, Joshi et al. 2015 ).In the conditions tested, no biofilm production was obtained and ther efor e no difference between treatments was observed.
Another phytopathogenic bacterium tested was B. glumae , a Gr am-negativ e rice pathogen causing bacterial panicle blight and gr ain r ot (She w et al. 2019, Choi et al. 2021 ).This bacterium is responsible for up to 75% of rice production losses in se v er al pr oducer countries, including Ja pan (She w et al. 2019, Cui et al. 2021 ).Due to global warming, B. glumae infections could become more common and regular over the coming decades (Shew et al. 2019 ), especially due to the range of its optimal growth temper atur e (fr om 30 • C to 35 • C) (Mizobuchi et al. 2020 ).In addition to rice pathogenicity, B. glumae is involved in wilting symptoms in sunflo w ers , tomatoes , sesame , perilla, eggplants , and hot peppers (Jeong et al. 2003 ), and of rot in onions (Karki et al. 2012 ).B. glumae has been described to pr oduce C 8 -HSL thr ough the TofIR QS system (Kim et al. 2004 ).The production of AHLs was e v aluated using the r eporter str ain CV026.Violacein pr oduction was observed in untreated samples .Con versely, no violacein production was observed upon supplementation with Sso Pox B, indicating AHL degradation by the lactonase (Fig. 2 A).Biofilm formation was assessed, but no biofilm formation was obtained in the tested conditions and no impact of Sso Pox treatment was observed (Fig. 2 B).Swarming motility, a surface movement initiated by a group of bacteria and driven by rotating flagella (Kearns 2010 ), was e v aluated.After 24 hours of incubation, untreated B. glumae were weakly motile and the zone surrounding the inoculation spot displayed a rounded shape.In contrast, bacteria grown on LB agar containing Sso Pox B sho w ed increased motility, together with a change in colony pattern, the shape of which became irregular ( Fig. S1 , Supporting Information ).Pr e vious studies sho w ed that motility regulation and efficiency is multifactorial and depends on QS, temper atur e, and the pr oduction of rhamnolipids, among other factors (Nickzad et al. 2015 ).Here, we report that AHL degradation at 30 • C modulates swarming motility.As motility was associated with virulence and pathogenesis of B. glumae in rice (Kim et al. 2018 ), the impact of Sso Pox on the ability to infect plants was e v aluated using a macer ation assay on r ed onions .T his infection model is an alternative to the rice infection model, and enables to assess the virulence of B. glumae with a high correlation between virulence in rice and onions (Karki et al. 2012 ).Onion slices were inoculated with B. glumae culture and measure-  ments of the macerated area were performed.Onion slices infected with B. glumae were highly macerated (i.e.64%), demonstrating the ability of B. glumae to infect onions (Fig. 2 C).In contr ast, the pr etr eatment of onion slices with Sso Pox B significantly inhibited the infection by B. glumae ( P-value < .0001), decreasing the av er a ge macer ation to 26%, corr esponding to a 2.5-fold decr ease.In summary, these r esults demonstr ate that Sso Pox B degr aded AHL pr oduced by B. glumae , alter ed motility, and decr eased maceration in an onion infection model.Pr e vious studies using B. glumae strains expressing the aii A lactonase gene also reported a decrease in AHL production, as well as reduced virulence in rice (Cho et al. 2007, Park et al. 2010 ) .Taken as a whole, these results suggest that QS inhibition decreases the degree of infections caused by B. glumae and that motility increase in vitro is not necessaril y corr elated with an incr ease in macer ation symptoms and onion infections .Further investigations in planta are needed to confirm the potential of QQ as a biocontrol solution against this phytopathogen.

Sso Pox activity against human pathogens
The ESKAPEE group designates pathogens known to escape drug ther a pies because of their m ulti r esistance to antibiotics (Cie ślik et al. 2021 ).The acr on ym ESKAPEE r efers to Enterococcus f aecium , Staphylococcus aureus , Klebsiella pneumoniae , A. baumannii , P. aeruginosa , Enterobacter species, and E. coli .These pathogens have been classified by the WHO in a priority list of antibiotic-resistant bacteria for which the development of new therapeutic solution is urgent (World Health Organization 2017 ).Among these species, se v er al bacteria use a QS system to regulate their virulence (Gallo w ay et al. 2012 ), and the disruption of this system of communication is studied as an alternative to antibiotic molecules (Nigam et al. 2014 ).
First, we tested the impact of QQ on A. baumannii .This Gramnegative bacterium causes a broad array of infections, including pneumonia, urinary tract infections, bloodstream infections, and skin infections, and has been found in man y healthcar e envir onments (Bhar gav a et al. 2010, Antunes et al. 2011 ).Se v er al lactones wer e r eported in A. baumannii , mainl y long-c hain AHLs v arying from C 10 to C 16 .3-OH-C 12 -HSL have been further described as the main AHL in A. baumannii (Saipriya et al. 2020 ).AHLs have also been shown to influence biofilm formation which is an important virulence factor involved in antibiotic resistance and survival pr operties (Bhar gav a et al. 2010, Mayer et al. 2020 ).The Sso Pox A variant was used to assess its QQ activity against A. baumannii .As the kinetic parameters could not be performed on 3-OH-C 12 -HSL, which is not commercially a vailable , the Sso P ox v ariant was c hosen with respect to its activity against long-chain AHLs, including 3-oxo-C 12 -HSL (Table 3 ).Culture of A. baumannii failed to induce violacein production in the reporter strain CV026 (Fig. 3 A).Nevertheless, an AHL synthase abaI was identified in the genome of the A. baumannii strain used in this study.In addition, C. violaceum CV026 is particularly adapted to the detection of AHLs from C 4 to C 8 , and w as sho wn not to produce violacein in the presence of long-c hain AHLs, e v en at high concentr ations (McClean et al. 1997 ).Ther efor e, biofilm formation was dir ectl y e v aluated as a QS phenotype.After 48 hours of culture, a high biofilm formation was found in the untreated condition, whereas supplementation with Sso Pox A lactonase induced a 4-fold decrease in the quantity of biofilm (Fig. 3 A).Similar results were obtained with A. baumannii ATCC 17978 and the lactonase Aii20J, significantl y decr easing the formation of biofilm (Mayer et al. 2020 ), as well as in clinical isolates of A. baumannii (López et al. 2017 ), underlining the potential use of QQ as an alternative treatment to fight A. baumannii infections.
Burkholderia cepacia , another human pathogen, was then considered.B. cepacia is an important pathogen involved in cystic fibrosis (Geisenberger et al. 2000 ), known to produce AHLs.C 8 -HSL has been described as the main AHL, while the production of C 6 -HSL has also been reported (Riedel et al. 2001, Vial et al. 2007 ).Here, 24hour cultures of B. cepacia treated or not treated with Sso Pox B were used for AHL detection, but no violacein production was observed in any of the samples, although the AHL synthase cepI was identified in the genome of the strain used for this assay (Fig. 3 B).Therefore, biofilm formation was studied as a direct readout for QS and QQ using crystal violet staining, as pr e viousl y described.In our conditions, the amount of biofilm was reduced upon treatment with Sso Pox B (Fig. 3 B).Biofilm formation is a phenotype known to be under QS regulation in members of B. cepacia complex , such as B. cepacia H111 (Huber et al . 2001 ).Her e, the r esults suggest that QQ may also influence biofilm formation in B. cepacia ATCC 25416 str ain, although further inv estigations would be r equir ed to confirm QQ efficiency on other specific phenotypes.
Finally, another human pathogen, P. aeruginosa , was tested.Pseudomonas aeruginosa is known to be responsible for hospitalacquir ed infections, particularl y in imm unocompr omized patients and patients with cystic fibrosis (Eberl andTümmler 2004 , P aller oni 2010 ).The QS of P. aeruginosa PA14 inv olves tw o major AHL-based QS systems, the LasIR and RhlIR, whic h pr oduce C 4 -HSL and 3-oxo-C 12 -HSL (Li et al. 2022 ).In this experimental setup, no AHL was detected by the r eporter str ain C. violaceum CV026 (Fig. 4 A), but the same r eporter str ain detected AHLs in ethyl acetate extractions of P. aeruginosa supernatants (Rémy et al. 2020 ).The addition of Sso Pox A, the variant selected for its degrading acti vity against 3-o xo-C 12 -HSL (Table 3 ), induced a 3-fold decrease in biofilm formation (Fig. 4 B), confirming pr e vious r esults obtained with P. aeruginosa PA14 and PAO1 (Hr aiec h et al. 2014, Guendouze et al. 2017, Rémy et al. 2020 ).The QQ activity of Sso Pox A was further e v aluated in an in vitro cytotoxicity model using macr opha ges.The bacterial cytotoxicity at a MOI of 10 was assessed to w ar d J774.1 macr opha ge cell lines.Untr eated cultur es of P. aeruginosa wer e str ongl y cytotoxic towards macr opha ges and 80% cytotoxic-ity was measured, while upon enzymatic treatment the cytotoxicity a gainst macr opha ges decr eased down to 20%, demonstrating the efficiency of QQ at reducing P. aeruginosa cytotoxicity and virulence factor production in this model.The cytotoxicity of cellfree supernatants was also evaluated.Cytotoxicity was detected in untreated supernatants, where 40% of cytotoxicity was measured, while cytotoxicity decreased below 10% in supernatants treated with Sso Pox A (Fig. 4  In addition, recent studies have demonstrated that the use of a QQ lactonase can significantly modulate P. aeruginosa PA14 susceptibility or resistance to antimicrobials (Rémy et al. 2020 , Sikdar andElias 2022 ).The synergistic effect of combining the QQE and antimicr obials (e.g.bacteriopha ges and antibiotics) to counteract P. aeruginosa PA14 virulence was also e v aluated demonstr ating the promising potential of this enzyme against human pathogenic strains (Mion et al. 2019 ).

Sso Pox inhibits virulence factors in aquaculture fish pathogens
In aquaculture, a significant proportion of periodical disease outbreaks is due to bacterial pathogens (Dadar et al. 2017 ).Among these diseases, vibriosis and aeromonasis are of crucial interest in aquaculture (J ay aprakashvel and Subramani 2019 ).To treat these pathogens, vaccination, antibiotics, and bacteriopha ges ar e the most studied methods (Defoirdt et al. 2011, Richards 2014, Dadar et al. 2017 ), and QQ is a de v eloping method, whic h is being considered to fight bacterial pathogens (J ay aprakashvel and Subramani 2019 ).
The fish pathogen A. salmonicida , was first tested.Aeromonas sp.ar e Gr am-negativ e bacteria living ubiquitousl y in aquatic envir onments and are opportunistic pathogens of animals, mainly fish, and humans (Tala gr and-Reboul et al. 2017, Vincent et al. 2019 ).Aeromonas salmonicida is responsible for furunculosis and septicaemia in fish (Tewari et al. 2014, Kim et al. 2015, Menanteau-Ledouble et al. 2016 ), and is the cause of se v er e economic losses in the aquaculture industry (Vincent et al. 2017 ).Aeromonas salmonicida uses two QS systems, the AI-2 and the AHL system, based on AsaIR (Swift et al. 1997, Jangid et al. 2007 ).
Pr e vious studies hav e r eported that the A. salmonicida str ain ATCC 33658 mainly produces C 4 -HSL and C 6 -HSL, as confirmed by thin layer c hr omatogr a phy (Swift et al. 1997, Tala gr and-Reboul et al. 2017 ), as well as C 8 -HSL, C 10 -HSL, C 12 -HSL, and C 14 -HSL, as detected by gas c hr omatogr a phy (Cataldi et al. 2007, Tala gr and-Reboul et al. 2017 ).
In this study, the impact of Sso Pox A was assessed focusing on AHL production, biofilm formation, and asaRI gene expression (Fig. 5 ).First, the production of AHLs by A. salmonicida was confirmed by the reporter strain C. violaceum CV026, while the addition of Sso Pox A resulted in the absence of violacein production, confirming its ability to degrade the corresponding AHLs (Fig. 5 A).
To study the impact of QS and QQ on A. salmonicida phenotypes, biofilm formation was then measured.Although Aeromonas str ains ar e known to pr oduce biofilm (Tala gr and-Reboul et al. 2017 ), no biofilm was detected in our conditions (Fig. 5 B).Therefor e, the expr ession of the genes involved in A. salmonicida QS system was c hec ked to e v aluate the impact of the lactonase on the QS of the strain.The expression of asaI was significantly decreased in the presence of Sso Pox A, as compared to the control without the enzyme.Ho w e v er, the expr ession of the asaR gene was not significantly impacted by the presence of the enzyme in comparison to the control without the enzyme (Fig. 5 C).These results show that Sso Pox is able to affect QS in A. salmonicida ATCC 33658.Indeed, AHL degradation by Sso Pox prevents AsaR activation by AHL binding, ther efor e limiting the induction of asaI expression.Further impact on A. salmonicida virulence remains to be studied but in two other A. salmonicida strains, the AsaIR system was correlated to virulence factor modulation.In A. salmonicida subsp.ac hromogenes , a asaI m utant was not able to produce AHLs, and its virulence was decreased in Arctic charr ( Salvelinus alpinus L.) (Schwenteit et al. 2011 ).In A. salmonicida AE03, also a asaI mutant, the production of C 4 -and C 6 -HSL was abolished, while C 8 -, C 10 -, and C 12 -HSL wer e still detected and the maxim um biofilm biomass and thickness were decreased (Liu et al. 2018 ).
A second marine gamma pr oteobacterium, V. harveyi , was then tested.Vibrio harveyi natur all y liv es in marine habitats, especially in warm waters and r epr esents, together with other Vibrio species, a serious pathogen of wild and farmed fish and inv ertebr ates (Zhang et al. 2020 ).Among its virulence arsenal, the productions of extr acellular toxins, exopr oteases, and sider ophor es (Defoirdt et al. 2008 ) are known to be regulated by QS.In addition to its ecological and economic significance, V. harveyi uses a QS system that differs fr om an y other Gr am-negativ e bacteria.This QS system r elies on three AIs, including the AHL 3-OH-C 4 -HSL (HAI-1), that are sensed extr acellularl y by their cognate sensor kinase, and the QS signal is transmitted to the LuxR regulator via a phosphorelay cascade (Cao and Meighen 1989, Chen et al. 2002, Ng et al. 2011 ).Finally, V. harveyi is also known for its easy QS readout, bioluminescence , which is in volved in the milky sea effect (Lapota et al. 1988, Miller et al. 2005 ).As no degradation signal was obtained with any Sso Pox variant using 3-OH-C 4 -HSL, no kinetic parameters could be determined.Ho w e v er, two v ariants identified for their ability to degrade AHLs with short acyl chains were selected, namely Sso Pox C and D (Table 3 ).Similarl y, the r eporter str ain C. violaceum CV026 did not detect 3-OH-C 4 -HSL, so that no direct evidence for degradation of this AHL by the selected variants could be observed (Fig. 6 A; Fig. S2 , Supporting Information ).Ne v ertheless, both v ariants successfully inhibited bioluminescence (Fig. 6 C), indicating that QS could be disrupted by Sso Pox.In addition, QS disruption was c hec ked using biofilm formation as a second QS biomarker .Interestingly, while biofilm formation was not at all impacted by the  but their use has thus far been limited to laboratories (Xu et al. 2022 ).To study the potential of QQ and particularly that of lactonases as a r eliable str ategy to limit antibiotic use against vibriosis, in vivo studies need to be performed (Tinh et al. 2007, Pande et al. 2013, Noor et al. 2019 ).
Most studies are based on the use of bacterial mutants unable to produce and/or perceive QS signals .T herefore , the outcome might differ from QQ approaches that induce a metabolic sink by constantl y degr ading AHL pr oduced by bacteria.Indeed, the addition of AHL-degrading bacteria ( Bacillus sp.) protected giant river pr awn larv ae ( Macrobrac hium rosenbergii ) a gainst V. campbellii viru-lence (Pande et al. 2013 ), and the direct injection of a lactonase in Car assius aur atus significantl y pr otected goldfish a gainst infection b y A. h ydrophila (P eng et al. 2021 ).

Conclusion
Taken as a whole, these results show that lactonase-mediated QS interference is of utmost interest when it comes to targeting the virulence of bacterial pathogens.Although the different bacteria did not r espond equall y to lactonase tr eatments, all wer e ne v ertheless affected in at least one QS-related phenotype.Depending on the communication molecules used and the bacterial regulatory QS circuit, the very same lactonase can be engineered by specific point mutations to efficiently target a broad range of bacteria and tackle their virulence.In vivo studies are needed to reinforce the pr e vious work and assess the effectiveness of Sso Pox treatment.To harness this broad-spectrum potential for a variety of applications, enzymes need to be competitive in comparison with conventional approaches, and enzyme engineering seems an attr activ e way to make this tec hnology cost effectiv e.
Mor eov er, cautious attention must be paid to regulatory requirements, as optimized enzymes issued fr om geneticall y modified micr oor ganisms m ust fit within curr ent r egulatory fr ame works, such as international regulations on health pr oducts, v eterinary medicinal products, and plant protection products.Although several aspects need to be considered before reaching industrial applications, this study suggests that lactonases could constitute a sustainable a ppr oac h to addr ess health thr eats in the animalhuman-en vironmental interface .T hese results , combined with the high stability of Sso P ox, will pa ve the way for concrete applications and broaden the arsenal available to fight bacterial infections.

Ac kno wledgements
W e thank V egepolys V alley for providing field isolates of phytopathogenic strains.

Figur e 1 .
Figur e 1. T he effect of Sso Pox variants on four phytopathogenic strains.Detection of AHL production using CV026, measurement of biofilm formation with crystal violet staining, and impact of Sso Pox enzymes on PCWDE on (A) D. dadantii , (B) D. chrysanthemi , (C) P. carotovorum (DSM), and (D) P. carotovorum (Vegepolys).Statistical analysis was performed using nonparametric Mann-Whitney test, ns: nonsignificant, on Gr a phP ad Prism 7.04.

Figur e 2 .
Figur e 2. Sso P ox B alters B. glumae phenotypes.(A) Detection of AHL production using CV026.(B) Measurement of biofilm formation with crystal violet staining.Statistical difference was measured according to Mann-Whitney test on Gr a phP ad Prism 7.04, ns: nonsignificant (C) Onion maceration, n = 15 slices for each condition.* * * P -value < .001according to Mann-Whitney test on GraphPad Prism 7.04.

Figur e 3 .
Figur e 3. Sso P ox A decreases biofilm formation in human pathogenic bacteria.Detection of AHL production using CV026 and measurement of biofilm formation with crystal violet staining with (A) A. baumannii .Experiment performed in n = 3 r eplicates.* * * * P -v alue < .0001according to Student's t -test on Gr a phP ad Prism 7.04.and with (B) B. cepacia .Experiment performed in n = 6 replicates.* * * * P -value < .0001according to Student's t -test on Gr a phP ad Prism 7.04.
C).These results demonstrate the impact of QS disruption by Sso Pox A tr eatment to tac kle P. aeruginosa virulence in an in vitro model.Pseudomonas aeruginosa strains wer e lar gel y used for QQE pr obing.Using this bacterial model, se v er al studies hav e demonstr ated that QQE may r educe v arious virulence factors such as motility, biofilm formation, and virulence against model organisms ( Caernorhabditis elegans , Drosophila melanogaster , and mice) (Hr aiec h et al. 2014 , Sikdar and Elias 2020 ) .

Figur e 4 .
Figur e 4. Sso P ox A decreases biofilm formation and cytotoxicity of P. aeruginosa .(A) Detection of AHL production using CV026.(B) Measurement of biofilm formation with crystal violet staining.Experiment performed in n = 3 r eplicates.* * P -v alue < .01 according to Student's t -test on Gr a phP ad Prism 7.04.(C) Cytotoxicity of P. aeruginosa cells and supernatants to w ar d J774.1A macr opha ges with Sso Pox A treatment or not.Experiment was performed in n = 3 r eplicates.* * P -v alue < .01,* * * P -value < .001according to Student's t -test on GraphPad Prism 7.04.

Figur e 5 .
Figur e 5. Sso P ox A alters QS of A. salmonicida.(A) Detection of AHL production using C. violaceum CV026.(B) Measurement of biofilm formation with crystal violet staining.ns: nonsignificant, according to Mann-Whitney test on Gr a phP ad Prism 7.04.(C) Relativ e expr ession of genes involved in the Asa system after lactonase treatment.The experiment was performed in n = 6 r eplicates.ns: nonsignificant, * P -v alue < .05,according to Multiple t -tests on Gr a phP ad Prism 7.04.

Figur e 6 .
Figur e 6. Sso P ox alters biofilm formation and bioluminescence in V. harveyi.(A) Detection of AHL production using C. violaceum CV026.(B) Measurement of biofilm formation with crystal violet staining.The experiment was performed in n = 4 replicates.ns: nonsignificant, * * * * P -value < .0001according to Student's t -test on GraphPad Prism 7.04.(C) Pictures of bioluminescence production.

Table 1 .
Sso Pox variants and concentration used for each strain.

Table 2 .
Sequences of primers used for RT-qPCR experiments.
Statistical analyses were performed with GraphPad Prism 8 softwar e, m ultiple t -test anal yses wer e used with statistical significance of 0.05.The experiment was performed twice, with three r eplicates eac h time ( n = 6).