The Lysozyme Inhibitor Thionine Acetate Is Also an Inhibitor of the Soluble Lytic Transglycosylase Slt35 from Escherichia Coli

molecules

Article The Lysozyme Inhibitor Thionine Acetate Is Also an Inhibitor of the Soluble Lytic Transglycosylase Slt35 from Escherichia coli

Aysha B. Mezoughi 1 , Chiara M. Costanzo 2, Gregor M. Parker 1, Enas M. Behiry 1, Alan Scott 1, Andrew C. Wood 1, Sarah E. Adams 1, Richard B. Sessions 3 and E. Joel Loveridge 1,2,*

1 School of Chemistry, Cardiff University, Park Place, Cardiff CF10 3AT, UK; [email protected] (A.B.M.); [email protected] (G.M.P.); [email protected] (E.M.B.); [email protected] (A.S.); [email protected] (A.C.W.); [email protected] (S.E.A.) 2 Department of Chemistry, Swansea University, Singleton Park, Swansea SA2 8PP, UK; [email protected] 3 School of Biochemistry, University of Bristol, University Walk, Bristol BS8 1TD, UK; [email protected] * Correspondence: [email protected]

Abstract: Lytic transglycosylases such as Slt35 from E. coli are enzymes involved in bacterial cell wall remodelling and recycling, which represent potential targets for novel antibacterial agents. Here, we investigated a series of known glycosidase inhibitors for their ability to inhibit Slt35. While glycosidase inhibitors such as 1-deoxynojirimycin, castanospermine, thiamet G and miglitol had

no effect, the phenothiazinium dye thionine acetate was found to be a weak inhibitor. IC50 values and binding constants for thionine acetate were similar for Slt35 and the hen egg white lysozyme. Molecular docking simulations suggest that thionine binds to the active site of both Slt35 and Citation: Mezoughi, A.B.; Costanzo, lysozyme, although it does not make direct interactions with the side-chain of the catalytic Asp and C.M.; Parker, G.M.; Behiry, E.M.; Glu residues as might be expected based on other inhibitors. Thionine acetate also increased the Scott, A.; Wood, A.C.; Adams, S.E.; potency of the beta-lactam antibiotic ampicillin against a laboratory strain of E. coli. Sessions, R.B.; Loveridge, E.J. The Lysozyme Inhibitor Thionine Acetate Keywords: Is Also an Inhibitor of the Soluble lytic transglycosylase; thionine acetate; enzyme inhibition; antibacterial Lytic Transglycosylase Slt35 from Escherichia coli. Molecules 2021, 26, 4189. https://doi.org/ 10.3390/molecules26144189 1. Introduction Bacterial cells are surrounded by a peptidoglycan sacculus on the outside of the cy- Academic Editor: F. J. (Frank) Dekker toplasmic membrane, which is essential for maintaining cell strength and integrity [1]. Peptidoglycan is composed of glycan strands of repeating N-acetylglucosamine (NAG) Received: 14 June 2021 and N-acetylmuramic acid (NAM) residues linked by β-1,4-glycosidic bonds, cross-linked Accepted: 6 July 2021 via peptide side-chains [2]. Escherichia coli turns over about 50% of its sacculus per gen- Published: 9 July 2021 eration, to allow cell growth and division, insertion of proteins into the cell wall and other processes [3]. The muropeptides released from the sacculus are recycled by the cell. Publisher’s Note: MDPI stays neutral The sacculus degradation and remodelling activities are catalysed by enzymes including with regard to jurisdictional claims in glycosidases, amidases, endopeptidases and carboxypeptidases [4,5]. published maps and institutional afﬁl- Lytic transglycosylases are important peptidoglycan-degrading glycosidases [6–8], iations. that catalyse cleavage of the β-1,4-glycosidic bond between NAM and NAG residues. Un- like lysozymes, they do so by catalysing an intramolecular transglycosylation reaction that forms 1,6-anhydromuropeptides [9], which can be recycled to form fresh peptidoglycan but which may also act as inducers of β-lactamase production [10] and as virulence fac- Copyright: © 2021 by the authors. tors [11–13]. Many bacterial species, including human pathogens such as Bacillus anthracis, Licensee MDPI, Basel, Switzerland. Staphylococcus aureus, Neisseria meningitides and Neisseria gonorrhoeae, inhibit the activity of This article is an open access article both lysozyme and lytic transglycosylases by acetylation of C-6 hydroxyl moieties of NAM distributed under the terms and residues in their peptidoglycan [14,15]. This is used to control lytic transglycosylase activity conditions of the Creative Commons and so prevent autolysis [7]. Vertebrate lysozymes are also inhibited by the proteinaceous Attribution (CC BY) license (https:// inhibitor Ivy, which is produced by certain Gram-negative bacteria [16]. Inhibition of the creativecommons.org/licenses/by/ 4.0/). membrane-bound lytic transglycosylase MltB from P. aeruginosa by Ivy was suggested to

Molecules 2021, 26, 4189. https://doi.org/10.3390/molecules26144189 https://www.mdpi.com/journal/molecules Molecules 2021, 26, 4189 2 of 13

Molecules 2021, 26, 4189 eties of NAM residues in their peptidoglycan [14,15]. This is used to control lytic transgly-2 of 13 cosylase activity and so prevent autolysis [7]. Vertebrate lysozymes are also inhibited by the proteinaceous inhibitor Ivy, which is produced by certain Gram-negative bacteria [16]. Inhibitionregulate the of autolyticthe membrane-bound activity of lytic lytic transglycosylases transglycosylase in Gram-negativeMltB from P. aeruginosa bacteria thatby Ivy are wasunable suggested to O-acetylate to regulate their the peptidoglycan autolytic activi [17ty]. of However, lytic transglycosylases despite preventing in Gram-nega- autolysis, tivespeciﬁc bacteria inhibition that are of LTsunable may to also O-acetylat increasee their the potency peptidoglycan of β-lactam [17]. antibioticsHowever, despite against preventingbacteria, as autolysis, exempliﬁed specifi by thec inhibition action of of the LTs bulgecins may also [18 increase–25]. the potency of β-lactam antibioticsIminosugars, against alkaloidsbacteria, as and exemplifie their syntheticd by the analogues action of (Figure the bulgecins1) can act [18–25]. as glycosidase inhibitorsIminosugars, by mimicking alkaloids the and glycosyloxocarbenium their synthetic analogues ion (Figure intermediate 1) can act and as related glycosidase tran- inhibitorssition states by inmimicking enzymatic the glycoside glycosyloxocarbe hydrolysisnium [26 ].ion 1-Deoxynojirimycin, intermediate and related a D-glucose transi- tionanalogue, states in is enzymatic the best known glycoside iminosugar. hydrolysis The[26].alkaloid 1-Deoxynojirimycin, castanospermine a D-glucose is also ana- a β- logue,glucosidase is the best inhibitor known [27 iminosugar.]. The NAG-thiazoline The alkaloid derivative castanospermine thiamet is G also is a potenta β-glucosidase inhibitor inhibitorof β-acetylglucosaminidases [27]. The NAG-thiazoline [28],while derivative miglitol thiamet (Glyset) G is is a usedpotent as inhibitor a drug for of treatment β-acetyl- glucosaminidasesof type II diabetes [28], [29 ].while Thionine miglitol acetate, (Glyset) a planar is used cationic as a drug phenothiazinium for treatment of dye, type has II diabetesbeen shown [29]. toThionine bind to acetate, lysozyme a planar [30]. cati Thisonic compound phenothiazinium is also used dye, inhas electrochemical been shown to bindand photochemicalto lysozyme [30]. biosensors This compound [31,32] is and also has used antibacterial in electrochemical activity towardand photochemical pathogenic biosensorsbacteria [33 [31,32]]. and has antibacterial activity toward pathogenic bacteria [33].

FigureFigure 1. StructuresStructures of of some some reported glycosid glycosidasease inhibitors and thionine acetate.

InIn this this study, study, we evaluated thionine acetate and some known glycosidase inhibitors againstagainst hen hen egg egg white white lysozyme lysozyme and and Slt35, Slt35, using using a a combination combination of of enzyme enzyme activity activity assays, assays, dissociationdissociation constant constant determination determination and molecular docking analysis.

2.2. Results Results and and Discussion Discussion 2.1. Enzyme Activity Assay 2.1. Enzyme Activity Assay 2.1.1. Effect of the Buffer 2.1.1. Effect of the Buffer A turbidimetric assay [34] was used to determine the activity of hen egg white lysozymeA turbidimetric and Slt35, assay by monitoring [34] was used peptidoglycan to determine solubilisation the activity of when hen egg the white enzyme lyso- is zymeadded and to Micrococcus Slt35, by monitoring lysodeikticus peptidoglycell suspension.can solubilisation Although this when assay the has enzyme previously is added been toperformed Micrococcus in sodiumlysodeikticus phosphate cell suspension. buffer for theAlthough same enzymes this assay [17 has], we previously found Tris-maleate been per- formedto give in superior sodium enzyme phosphate activity: buffer maximum for the same 0.018 enzymes A min −[17],1 for we phosphate found Tris-maleate (pH 6.4) and to give0.068 superior A min− enzyme1 for Tris-maleate activity: maximum (pH 5.8), 0.018 both containingA min−1 for 100phosphate mM NaCl. (pH 6.4) Acetate, and 0.068 MES Aand min HEPES−1 for buffersTris-maleate gave very(pH poor5.8), activity:both containing maximum 100 0.020 mM A NaCl. min−1 Acetate,for sodium MES acetate and HEPES(pH 5.5 ),buffers 0.003 gave A min very−1 forpoor MES activity: (pH 6.0)maximum and 0.003 0.020 A A min min−1−1for for HEPESsodium (pHacetate 7.0), (pH all 5.5),containing 0.003 A 100 min mM−1 for NaCl. MES (pH 6.0) and 0.003 A min−1 for HEPES (pH 7.0), all containing 100 mM NaCl. 2.1.2. Effect of Salts

2.1.2. TheEffect effect of Salts of NaCl, KCl, MgCl2 and CaCl2 on enzyme activity was determined. The activityThe of effect Slt35 of was NaCl, not KCl, determined MgCl2 and in the CaCl absence2 on enzyme of salts, activity due to very waslow determined. activities The and activitydifﬁculty of in Slt35 handling was not and determined storing the in protein. the abse Monovalentnce of salts, cations due to did very not low have activities a signiﬁcant and difficultyeffect on Slt35in handling activity, and whereas storing 10 the mM protei divalentn. Monovalent cations gave cations optimal did activity not have (Figure a signifi-2A). cantThe thermostabilityeffect on Slt35 activity, of Slt35 whereas has been 10 shown mM divalent to increase cations when gave it is optimal bound toactivity calcium (Figure ions, 2A).but notThe sodium, thermostability potassium of Slt35 or magnesium has been shown ions at to 10 increase mM [35 when]. The it melting is bound temperature to calcium ◦ ◦ ions,of Slt35 but increased not sodium, from potassium 48.0 C in or the magnesium absence of ions salts at to1055.5 mM [35].C in The the presencemelting tempera- of 1 mM tureCaCl of2, suggestingSlt35 increased that thefrom EF-hand 48.0 °C calcium in the abse bindingnce of site salts of Slt35to 55.5 plays °C in an the important presence role of in1 mMprotein CaCl stability2, suggesting [35]. However, that the theEF-hand functional calcium role binding of this site site has of notSlt35 been plays established. an important The effect of salts on lysozyme activity was also assessed (Figure2B). Potassium was found to be the optimal cation in this case, as other cations caused a reduction in lysozyme activity at higher concentrations.

MoleculesMolecules 2021 2021, 26, 26, 4189, 4189 3 3of of 13 13

rolerole in in protein protein stability stability [35]. [35]. However, However, the the functional functional role role of of this this site site has has not not been been estab- estab- Molecules 2021, 26, 4189 lished.lished. The The effect effect of of salts salts on on lysozyme lysozyme activity activity was was also also assessed assessed (Figure (Figure 2B). 2B). Potassium Potassium3 of 13 waswas found found to to be be the the optimal optimal cation cation in in this this case, case, as as other other cations cations caused caused a a reduction reduction in in lysozymelysozyme activity activity at at higher higher concentrations. concentrations.

−1 −1 FigureFigure 2. 2. Effect Effect of of of salt salt salt concentration concentration concentration on on on ( A(A ()A )Slt35 )Slt35 Slt35 and and and ( B(B) ( B)lysozyme )lysozyme lysozyme activity. activity. activity. For For For Slt35, Slt35, Slt35, 7 7μ μM 7Mµ enzyme Menzyme enzyme and and and0.6 0.6 mg 0.6 mg mgml ml−ml 1cell cell ◦ −1 − suspensioncellsuspension suspension were were wereused used at used at 37 37 °C at °C 37in in 25 25C mM inmM 25 Tris-maleate Tris-maleate mM Tris-maleate (pH (pH 5.8). 5.8). (pH For For 5.8). lysozyme, lysozyme, For lysozyme, 0.1 0.1 μ μMM 0.1enzyme enzymeµM enzyme and and 0.6 0.6 andmg mg ml 0.6ml− mg1cell cell mlsuspen- suspen-1 cell sionsuspensionsion were were used used were at at 37 used 37 °C °C atin in 3725 25◦ mMC mM in Tris-maleate 25Tris-maleate mM Tris-maleate (pH (pH 6.0). 6.0). (pH 6.0).

2.1.3.2.1.3. Effect Effect of of pH pH TheThe activity activity of ofof Slt35 Slt35Slt35 showed showedshowed a aa bell-sha bell-shapedbell-shapedped dependence dependence onon on pH,pH, pH, withwith with maximummaximum maximum activity activ- activ- ityatity atpH at pH pH 5.8 5.8 5.8 (Figure (Figure (Figure3A). 3A). 3A). Lysozyme Lysozyme Lysozyme displayed displa displayedyed a a mucha much much broader broader broader pHpH pH rangerange range giving giving maximum maximum activity,activity, and and was was generally generally more more tolerant tolerant of of changes changes in in the the pH, pH, although although activity activity was was de- de- creasedcreased below below pH pH 5 5 and and above above pH pH 8 8(Figure (Figure 3B).3 3B).B). ItIt It has hashas been beenbeen noted notednoted that thatthat the thethe pH pHpH dependence dependencedependence ofof lysozyme lysozyme activity activity is is strongly strongly affected affected by by the the ionic ionic strength strength [36], [[36],36], which which was was different different for for thethe two two enzymes enzymes here. here. The The accepted accepted mechanism mechanism of of catalysis catalysis [37] [37] [37 ]by by the the hen hen egg eggegg white whitewhite lysozymelysozyme requires requires Asp52 Asp52Asp52 to toto be bebe deprotonated deprotonateddeprotonated and andand Glu35 Glu35Glu35 to toto be bebe protonated. protonated.protonated. The TheThe p ppKKKaa avalues values ofof these these residues residuesresidues are areare 3.7 3.73.7 and and 6.2 6.2 respectively respectively [[38],38 [38],], accountingaccounting accounting for for for the the the reduction reduction reduction in in in activity activity activity at atlowat low low and and and high high high pH. pH. pH. TheThe The suggestedsuggested suggested mechanismmechanism mechanism fo for for r Slt35 Slt35 requires requires requires Glu162 Glu162Glu162 to toto be bebe protonated, protonated, butbut there there is is no no no requirement requirement requirement for for for an an an equivalent equivalent equivalent deprotonated deprotonated deprotonated residue. residue. residue. The The The side-chain side-chain side-chain p pKKa p aofK ofa freeoffree freeglutamate glutamate glutamate is is 4.5, 4.5, is suggesting 4.5,suggesting suggesting that that Glu162that Glu162 Glu162 has has its its has p pKK itsa avalue value pKa valueperturbed perturbed perturbed to to a a higher higher to a highervalue value byvalueby its its local bylocal its environment, environment, local environment, as as seen seen as in in seen the the inhen hen the egg egg hen white white egg whitelysozyme lysozyme lysozyme [38] [38] and [and38] many andmany many other other other en- en- zymesenzymeszymes [39]. [39]. [ 39The The]. Theloss loss lossof of activity activity of activity at at lower lower at lower pH pH is pHis then then is thenlikely likely likely to to be be todue due be to dueto unfavourable unfavourable to unfavourable pro- pro- tonationprotonationtonation of of a a ofdifferent different a different residue, residue, residue, with with with a a p apKK pa aKbelow belowa below 6, 6, that 6, that that may may may affect affect affect substrate substrate substrate binding binding or or thethe overall overall arrangement arrangement of of the the active active site. site.

−1 Figure 3. Effect of pH on (A) Slt35 and (B) lysozyme activity. For Slt35, 7 μM enzyme and 0.6 mg mL − 1cell suspension Figure 3.3. EffectEffect ofof pHpH onon ((AA)) Slt35Slt35 andand (B(B)) lysozyme lysozyme activity.activity. ForFor Slt35, Slt35, 7 7µ μMM enzyme enzyme and and 0.6 0.6 mg mg mL mL−1 cellcell suspensionsuspension −1 were used at 37 °C in 25 mM Tris-maleate containing 10 mM CaCl2. For the lysozyme, 0.1 μM enzyme and 0.6 mg mL − 1 werewere usedused atat 37 37◦ C°C in in 25 25 mM mM Tris-maleate Tris-maleate containing containing 10 mM10 mM CaCl CaCl. For2. For the the lysozyme, lysozyme, 0.1 µ0.1M μ enzymeM enzyme and 0.6and mg 0.6 mL mg− 1mLcell cell suspension were used at 37 °C in 25 mM Tris-maleate containing2 100 mM KCl. suspensioncell suspension were were used used at 37 at◦ C37 in °C 25 in mM 25 mM Tris-maleate Tris-maleate containing containing 100 mM100 mM KCl. KCl.

2.1.4. Effect of the Substrate Slt35 showed maximum activity at a substrate concentration of 0.4 mg mL−1 (Figure4A ), −1 in keeping with previous studies, where substrate concentrations of 0.4–0.6 mg mL were used [17]. Above this, Slt35 activity reduced signiﬁcantly, demonstrating substrate inhibi- Molecules 2021, 26, 4189 4 of 13

2.1.4. Effect of the Substrate Molecules 2021, 26, 4189 4 of 13 Slt35 showed maximum activity at a substrate concentration of 0.4 mg mL−1 (Figure 4A), in keeping with previous studies, where substrate concentrations of 0.4–0.6 mg mL−1 were used [17]. Above this, Slt35 activity reduced significantly, demonstrating substrate tioninhibition [40]. The [40]. optimal The optimal lysozyme lysozyme activity activity was observed was observed at substrate at substrate concentrations concentrations above 0.6above mg/mL 0.6 mg/mL (Figure (Figure4B). Unlike 4B). Slt35,Unlike the Slt35, lysozyme the lysozyme was not inhibitedwas not byinhibited further by increases further inincreases the substrate in the concentration,substrate concentration, in the range in used the range here. Dueused to here. the highDue concentrationto the high concen- of the enzymetration of used the enzyme here, lysozyme used here, did lysozyme not follow did Michaelis–Menten not follow Michaelis–Menten kinetics, so the kinetics, data were so insteadthe data ﬁt were to the instead Morrison fit to equation the Morrison for tight equation binding for of tight the substrate binding [of41 ],the giving substrate an appar- [41], −1 entgivingKm anof 0.034apparent mg mLKm of 0.034cell suspension. mg mL−1 cell This suspension. is in keeping This with is in previous keeping studies,with previous which −1 foundstudies,K mwhichvalues found in the K regionm values of 0.01–0.05in the region mg mLof 0.01–0.05for M. mg lysodeikticus mL−1 for cellM. lysodeikticus suspension [ 42cell], andsuspension a Kd value [42], of and 14 µa MKd forvalue chitotriose of 14 μM [43 for]. chitotriose As Slt35 displayed [43]. As Slt35 both substratedisplayed inhibitionboth sub- andstrate tight inhibition substrate and binding, tight substrate we were binding, unable to we obtain were good unable ﬁts to for obtain the experimental good fits for data. the Fittingexperimental to an expression data. Fitting for to substrate an expression inhibition for basedsubstrate on Michaelis–Menteninhibition based on kinetics Michaelis– [40] −1 −1 gaveMenten a value kinetics of 7.54 [40] mggave mL a valuefor Kofm 7.54and mg 0.0012 mL− mg1 for mL Km andfor 0.0012KI. mg mL−1 for KI.

Figure 4.4. (A): Effect of the substrate concentration on (A) Slt35 and (B) lysozyme activity. ForFor Slt35,Slt35, thethe 7 µμM enzyme was used at 37 ◦°C in 25 mM Tris-maleate (pH 5.8) containing 10 mM CaCl2. A fit to an expression for substrate inhibition [40] used at 37 C in 25 mM Tris-maleate (pH 5.8) containing 10 mM CaCl2. A fit to an expression for substrate inhibition [40] is shown.is shown. For For lysozyme, lysozyme, 0.1 0.1µM μM enzyme enzyme was was used used at 37at 37◦C °C in in 25 25 mM mM Tris-maleate Tris-maleate (pH (pH 6.0) 6.0) containing containing 100 100 mM mM KCl. KCl. A fitA tofit to the Morrison equation for the tight-binding substrate [41] is shown. the Morrison equation for the tight-binding substrate [41] is shown. 2.1.5. Effect of Temperature 2.1.5. Effect of Temperature To investigate the effect of temperature on enzymatic activity, Slt35 and the lysozyme To investigate the effect of temperature on enzymatic activity, Slt35 and the lysozyme were incubated at different temperatures for 10 min before they were added to the sub- were incubated at different temperatures for 10 min before they were added to the substrate. strate. The maximum lytic activity of Slt35 was recorded when the enzyme was incubated The maximum lytic activity of Slt35 was recorded when the enzyme was incubated below below 15 °C, while the enzyme lost about half of its activity at 25–40 °C, and at 50 °C the 15 ◦C, while the enzyme lost about half of its activity at 25–40 ◦C, and at 50 ◦C the enzyme enzyme precipitated (Figure 5A). Increasing the incubation temperature as high as 60 °C precipitated (Figure5A). Increasing the incubation temperature as high as 60 ◦C did did not show any significant effect on lysozyme activity (Figure 5B). Although all kinetic not show any significant effect on lysozyme activity (Figure5B). Although all kinetic measurements in this work were performed at the physiologically relevant temperature measurements in this work were performed at the physiologically relevant temperature of 37of 37◦C, °C, the the enzyme enzyme was was always always stored stored at <4 at ◦<4C, °C, and and all incubationsall incubations with with potential potential inhibitors inhib- wereitors were performed performed on ice on to ice avoid to avoid denaturation denatura oftion the of enzyme the enzyme prior prior to measurements. to measurements. We didWe did not not expect expect significant significant denaturation denaturation of of the the enzyme enzyme to to occur occur during during the the1 1 minmin assay at 37 ◦°C;C; initial rates were linear.linear. Others have alsoalso performedperformed experimentsexperiments with Slt35Slt35 atat temperatures of 25–37 ◦°CC[ [18,44,45].18,44,45].

Molecules 2021, 26, 4189 5 of 13 Molecules 2021, 26, 4189 5 of 13 Molecules 2021, 26, 4189 5 of 13

Figure 5. Effect of temperature on (A) Slt35 andand ((B)) lysozymelysozyme activity.activity. ForFor Slt35,Slt35, thethe 77 µμMM enzymeenzyme andand 0.60.6 mgmg mLmL−−1 cell −1 Figure 5. Effect of temperature on (A) Slt35 and (B) lysozyme activity. For Slt35,2 the 7 μM enzyme and 0.6 mg mL cell suspensionsuspension were used in 25 mM Tris-m Tris-maleatealeate (pH 5.8) containing 10 mM CaCl2.. For For the the lysozyme, lysozyme, the the 0.1 0.1 μµM enzyme enzyme and 0.6suspension mg mL−1 1werecell suspension used in 25 weremM Tris-mused inaleate 25 mM (pH Tris-maleate 5.8) containing (pH 6.0)10 mM containing CaCl2. For 100 the mM lysozyme, KCl. The theenzyme 0.1 μ Mwas enzyme incubated and 0.6 mg mL −1 cell suspension were used in 25 mM Tris-maleate (pH 6.0) containing 100 mM KCl. The enzyme was incubated for0.6 10mg min mL at cell the suspensiontemperatures were indicated used in before25 mM being Tris-maleate added to (p theH 6.0) substrate. containing Activity 100 mM was KCl. monitored The enzyme at 37 ◦°C.was incubated forfor 1010 minmin atat thethe temperaturestemperatures indicatedindicated beforebefore beingbeing addedadded toto thethe substrate.substrate. ActivityActivity waswas monitoredmonitored atat 3737 °C.C. 2.2. Enzyme Inhibition Studies 2.2. Enzyme Inhibition Studies The same turbidimetric assay was used to evaluate the inhibition of lysozyme and The same turbidimetric assay was used to evaluate the inhibition of lysozyme and Slt35.The 1-Deoxynojirimycin, same turbidimetric castanospermine, assay was used thiamet to evaluate G and the miglitol inhibition did notof lysozyme inhibit either and Slt35. 1-Deoxynojirimycin, castanospermine, thiamet G and miglitol did not inhibit either enzymeSlt35. 1-Deoxynojirimycin, at concentrations up castanospermine, to 10 mM. Thionine thiamet acetate G and was miglitol also evaluated did not for inhibit inhibition either enzyme at concentrations up to 10 mM. Thionine acetate was also evaluated for inhibition ofenzyme Slt35 andat concentrations HEWL. Increasing up to 10the mM. concentratio Thioninen acetate of thionine was also acetate evaluated reduced for the inhibition rate of of Slt35 and HEWL. Increasing the concentrationconcentration of thionine acetate reduced the rate of the lytic reaction for both enzymes (Figure 6). The IC50 was found to be 89.3 ± 3 μM for the lytic reaction for both enzymes (Figure6). The IC 50 was found to be 89.3 ± 3 µM for lysozymethe lytic reaction and 66.0 for ± 0.1 both μM enzymes for Slt35. (Figure 6). The IC50 was found to be 89.3 ± 3 μM for lysozyme and 66.0 ± 0.10.1 μµMM for for Slt35. Slt35.

Figure 6. Inhibition of (A) Slt35 and (B) lysozyme by thionine acetate. For Slt35, 7 μM enzyme and 0.6 mg mL−1 cell sus- −1 −1 pensionFigure 6. were InhibitionInhibition used at of of37 ( (A°CA)) Slt35in Slt35 25 andmM and (Tris-mB (B) )lysozyme lysozymealeate bufferby by thionine thionine (pH 5.8) acetate. acetate. containing For For Slt35, 10 Slt35, mM 7 μ 7CaClMµ Menzyme2. enzyme For lysozyme, and and 0.6 0.6mg 0.1 mgmL μM mL enzymecell sus-cell pension were used−1 at 37 °C in◦ 25 mM Tris-maleate buffer (pH 5.8) containing 10 mM CaCl2. For lysozyme, 0.1 μM enzyme andsuspension 0.6 mg mL were cell used suspension at 37 C inwere 25 mMusedTris-maleate at 37 °C in 25 buffer mM Tr (pHis-maleate 5.8) containing buffer (pH 10 6.0) mM containing CaCl2. For 100 lysozyme, mM KCl. 0.1 µM −1 enzymeand 0.6 mg and mL 0.6 mg cell mL suspension−1 cell suspension were used were at 37 used °C atin 3725◦ mMC in Tr 25is-maleate mM Tris-maleate buffer (pH buffer 6.0) (pH containing 6.0) containing 100 mM 100 KCl. mM KCl. 2.3. Binding Studies 2.3. Binding Studies 2.3. BindingInterestingly, Studies although 1-deoxynojirimycin did not inhibit Slt35, the saturation trans- Interestingly, although 1-deoxynojirimycin did not inhibit Slt35, the saturation transfer differenceInterestingly, NMR although spectroscopy 1-deoxynojirimycin showed weak didenhancement not inhibit of Slt35, signals the from saturation 1-deoxynoji- transfer fer difference NMR spectroscopy showed weak enhancement of signals from 1-deoxynoji- rimycindifference on NMR selective spectroscopy excitation showed of Slt35 weak (Figure enhancement S1), suggesting of signals that from this 1-deoxynojirimycincompound is ca- pablerimycinon selective of onbinding selective excitation to Slt35. excitation of Co Slt35nversely, of (Figure Slt35 saturation S1),(Figure suggesting S1), transfer suggesting that difference this that compound thisNMR compound did is capablenot detect is ca- of anypablebinding binding of tobinding Slt35. of castanospermine Conversely,to Slt35. Co saturationnversely, to Slt35. saturation transfer difference transfer NMRdifference did not NMR detect did any not binding detect anyof castanospermine bindingThe extinction of castanospermine of to the Slt35. intrinsic to fluorescenceSlt35. of Slt35 and lysozyme in the presence of thionineThe acetate extinction was of used the to intrinsic determine ﬂuorescencefluorescence the binding of constant. Slt35 and The lysozyme effect of in thionine the presence acetate of thionine acetate was used to determine the binding constant. The effect of thionine acetate onthionine the fluorescence acetate was emission used to determine of Slt35 an thed bindinglysozyme constant. was similar The effect (Figure of thionine 7), and acetatethe Kd on the fluorescence emission of Slt35 and lysozyme was similar (Figure 7), and the Kd valueson the of ﬂuorescence thionine acetate emission complexed of Slt35 to and Slt3 lysozyme5 and lysozyme was similar were (Figurefound to7), be and 17.1 the ± 1.0Kd andvalues 18.6 of ± thionine1.2 μM respectively. acetate complexed to Slt35Slt35 andand lysozymelysozyme werewere foundfound toto bebe 17.117.1± ± 1.0 and 18.6 ± 1.21.2 μµMM respectively. respectively.

Molecules 2021, 26, 4189 6 of 13 Molecules 2021, 26, 4189 6 of 13 Molecules 2021, 26, 4189 6 of 13

FigureFigure 7.7. Fluorimetric titration of lysozyme or Slt35 with thionine acetate. FluorescenceFluorescence emission of (A) 3 µμM Slt35 and (B) Figure 7. Fluorimetric titration of lysozyme or Slt35 with thionine acetate. Fluorescence emission of (A) 3 μM Slt35 and (B) 2.3 μM lysozyme in the presence of varying concentrations (indicated) of thionine acetate. Change in maximum fluores- 2.3 µ2.3M lysozymeμM lysozyme in the in presence the presence of varying of varying concentrations concentrations (indicated) (indicated) of thionine of thionine acetate. acetate. Change Change in maximum in maximum ﬂuorescence fluorescence intensity of (C) lysozyme and (D) Slt35 with increasing thionine acetate concentration. intensitycence ofintensity (C) lysozyme of (C) lysozyme and (D) Slt35 and with(D) Slt35 increasing with increasing thionine acetate thionine concentration. acetate concentration. 2.4. Molecular Docking Studies 2.4.2.4. Molecular Molecular Docking Docking Studies Studies Docking was performed using the Bristol University Docking Engine (BUDE) [46], to DockingDocking was was performed performed using using the the Bristol Bristol University University Docking Docking Engine Engine (BUDE) (BUDE) [ 46[46],], to elucidate the binding mode of thionine acetate in the active of both Slt35 and hen egg to elucidateelucidate thethe bindingbinding mode mode of of thionine thionine acetate acetate in in the the active active of of both both Slt35 Slt35 and and hen hen egg egg white lysozyme. The docking suggests that thionine binding to lysozyme is dominated by whitewhite lysozyme. lysozyme. The The docking docking suggests suggests that that thionine thionine binding binding to to lysozyme lysozyme is is dominateddominated by H-bonding interactions between amino groups in the inhibitor and Glu35 and Arg61 res- by H-bonding interactions between amino amino groups groups in in the the inhibitor inhibitor and and Glu35 Glu35 and and Arg61 Arg61 residues in the lysozyme active site (Figure 8). Another, weaker, interaction could occur be- residuesidues inin thethe lysozymelysozyme active site (Figure 88).). Another,Another, weaker,weaker, interactioninteraction couldcould occuroccur between Asn59 and the inhibitor. betweentween Asn59 Asn59 and and the the inhibitor. inhibitor.

Figure 8. The predicted binding mode of thionine acetate with the hen egg white lysozyme (PDB 2VB1 [47]). The enzyme Figure 8. The predicted binding mode of thionine acetate with the hen egg white lysozyme (PDB 2VB1 [47]). The enzyme Figureis shown 8. Thein pink predicted as (A) bindinga surface mode representation of thionine acetateof the binding with the site hen and egg ( whiteB) a cartoon lysozyme representation (PDB 2VB1 [showing47]). The the enzyme amino is is shown in pink as (A) a surface representation of the binding site and (B) a cartoon representation showing the amino shownacid residues in pink (as as (sticks)A) a surface located representation in the active site of the close binding to the site inhibitor. and (B) The a cartoon inhibitor representation is shown as showing orange sticks. the amino The acidpre- acid residues (as sticks) located in the active site close to the inhibitor. The inhibitor is shown as orange sticks. The pre- residuesdicted hydrogen (as sticks) bonding located interactions in the active are site sh closeown as to green the inhibitor. dashes with The inhibitordistances is in shown Å. as orange sticks. The predicted dicted hydrogen bonding interactions are shown as green dashes with distances in Å. hydrogen bonding interactions are shown as green dashes with distances in Å. The lysozyme–thionine binding mode obtained here is similar, but not identical, to The lysozyme–thionine binding mode obtained here is similar, but not identical, to that Theobtained lysozyme–thionine in previous docking binding studie modes obtainedusing the here lower-resolution is similar, but PDB not identical, 6LYZ [30], to that obtained in previous docking studies using the lower-resolution PDB 6LYZ [30], thatwhich obtained showed in previousthionine dockingacetate studiesforming using hydrogen the lower-resolution bonds to the PDBmain-chain 6LYZ [30 carbonyl], which which showed thionine acetate forming hydrogen bonds to the main-chain carbonyl showedgroups of thionine Leu56 and acetate Ile58 formingof lysozyme. hydrogen In our bonds work, tothionine the main-chain acetate is shifted carbonyl more groups into groups of Leu56 and Ile58 of lysozyme. In our work, thionine acetate is shifted more into the centre of the active site; the orientation and approximate position of thionine is similar the centre of the active site; the orientation and approximate position of thionine is similar

Molecules 2021, 26, 4189 7 of 13

of Leu56 and Ile58 of lysozyme. In our work, thionine acetate is shifted more into the centre of the active site; the orientation and approximate position of thionine is similar to tothat that observed observed in in the the previous previous work, work, but but hydrogen hydrogen bonds bonds areare mademade toto differentdifferent residues (Figure S2). To To investigate whether the differences in the predicted binding mode are due to the different docking software used, we repeated all docking using AutoDock [48]. [48]. We We found only small differences between the binding mode of thionine acetate predicted by BUDE or AutoDock whenwhen PDBPDB 2VBI 2VBI was was used used (Figure (Figure S3), S3), and and obtained obtained poses poses very very similar sim- ilarto those to those observed observed previously previously [30] [30] when when PDB PDB 6LYZ 6LYZ was was used used (Figure (Figure S4). ThisS4). This suggests suggests that thatthe smallthe small differences differences in the in lysozyme-thionine the lysozyme-thionine binding binding mode mode observed observed between between our work our workand the and previous the previous studies studies are due are to due the to different the different crystal crystal structures structures used. used. Docking of thionine with Slt35 suggested the presence of hydrogen bonds between the two amino groups of the inhibitor and thethe carbonyls of Glu162 and Gly215 residues (Figure9 ).9). Another Another hydrogen-bonding hydrogen-bonding interaction interaction was was observed observed between between one of one the aminoof the aminogroups groups in thionine in thionine and the and side-chain the side-chain hydroxyl hydroxyl of Thr172 of Thr172 in the activein the active site of site Slt35, of whileSlt35, whilethe side-chain the side-chain of Val168 of Val168 formed formed a hydrophobic a hydrophobic interaction interaction with onewith of one the of aromatic the aromatic rings ringsin thionine. in thionine. As observed As observed for lysozyme, for lysozyme, the Slt35-thionine the Slt35-thionine docking docking result was result very was similar very similarwhen AutoDock when AutoDock was used was (Figure used (Figure S5). S5).

Figure 9. Predicted binding binding mode mode of of thionine thionine acetate acetate with with Slt35 Slt35 (PDB (PDB 1QUS 1QUS [49]). [49]). The The enzyme enzyme is isshown shown in inblue blue as as(A ()A a) asurface surface representation representation of of the the binding binding site site and and (B (B) )a a cartoon cartoon representation representation showing showing th thee amino amino acid acid residues residues (as sticks) located in the active site close to the inhibitor. The inhibitor is shown as orange sticks. The predicted hydrogen bonding located in the active site close to the inhibitor. The inhibitor is shown as orange sticks. The predicted hydrogen bonding interactions are shown as green dashes with distances in Å. interactions are shown as green dashes with distances in Å.

Thionine does not make direct interactions with the side-chain of the key catalytic Glu162 of Slt35 and adopts a a quite different different binding mode mode to to bulgecin bulgecin A A (Figure (Figure S6) S6) [50]. [50]. Based on the interactions of bulgecin A with Slt35, an interaction between the cationic nitrogen of thethe inhibitorinhibitor andand the the carboxylate carboxylate side side chain chain would would be be expected, expected, to mimicto mimic the the in- interactionteraction between between the the carboxylate carboxylate of Glu162 of Glu162 and theand oxocarbenium the oxocarbenium ion intermediate ion intermediate [49,50]. [49,50].However, However, such an such interaction an interaction is not required is not required for inhibition, for inhibition, as long as as the long active as the site active is ob- sitestructed. is obstructed. Indeed, thionine Indeed, sitsthionine across sits the across mouth the of themouth Slt35 of active the Slt35 site (Figureactive site9, Figure (Figure S6 9,), Figureblocking S6), binding blocking of binding the large of peptidoglycan the large peptidoglycan chain. As chain. the structure As the structure of thionine of thionine is quite isdifferent quite different to that ofto peptidoglycan,that of peptidoglycan, and the an actived the siteactive of site Slt35 of contains Slt35 contains numerous numerous other otherresidues residues capable capable of interacting of interacting with a cationwith a [ 49cation,50], the[49,50], different the different orientation orientation of the inhibitor of the inhibitoris not surprising. is not surprising. Our docking studies suggest that the binding affinity afﬁnity of Slt35 for thionine could be improvedimproved by by extending the the thionine core core at positions C3 and C13 (i.e., the carbon atoms between thethe aminoamino groupsgroups and and the the ring ring bridgeheads; bridgeheads; Figure Figure1). Extension1). Extension at these at these positions posi- tionswith chainswith chains or rings orcontaining rings containing polar groups polar wouldgroups allowwould additional allow additional hydrogen hydrogen bonding bondingand/or ion-pairand/or ion-pair interactions interactions within thewithin active the site,active increasing site, increasing afﬁnity affinity and boosting and boost- the inginhibition the inhibition of the enzyme. of the enzyme.

Molecules 2021, 26, 4189 8 of 13

2.5. Antibacterial Testing Thionine acetate showed no antibacterial activity against Escherichia coli JM109 by disk diffusion assays. However, when inoculated onto the same disk as ampicillin, thionine acetate did give a small increase in the inhibition zone diameter (Table1), suggesting synergism with the antibiotic. This is consistent with previous studies, where the lytic transglycosylase inhibitor bulgecin was shown to greatly increase the potency of beta-lactam antibiotics [18,19,22], and further supports the view that lytic transglycosylase inhibitors may be used clinically as combination therapies with beta-lactam antibiotics [7,8,51,52]. Bulgecin A is also an inhibitor of metallo-β-lactamases [53].

Table 1. Effect of thionine acetate on the inhibition zone diameter of ampicillin against E. coli.

Mass of Thionine Acetate/µg Inhibition Zone Diameter/mm 0 25 ± 1 125 25 ± 2 250 28 ± 1 500 30 ± 1 1000 31 ± 1

3. Materials and Methods 3.1. Chemical Reagents and Enzymes A pET28a plasmid harbouring the gene encoding Slt35 with an N-terminal His-tag, between NdeI and BamH1 restriction sites, was purchased from Epoch Life Science (USA). The hen egg white lysozyme was purchased from Sigma-Aldrich. Thionine acetate and other chemicals were purchased from Sigma-Aldrich or Fisher Scientiﬁc unless otherwise stated.

3.2. Protein Production and Purification Escherichia coli XL-1 Blue and BL21(DE3) cells were transformed with a pET28a plasmid harbouring the slt35 gene using standard techniques. A single colony of freshly- transformed E. coli BL21(DE3) cells was inoculated in LB medium containing kanamycin (50 µg mL−1) and grown overnight at 37 ◦C and 180 rpm. This overnight culture was used to seed (1:50 dilution) fresh kanamycin-selective LB medium and growth was continued at ◦ 37 C and 180 rpm until the OD600 was 0.6–0.8. slt35 expression was induced using 120 mg L−1 isopropyl-β-D-thiogalactopyranoside and cultures were incubated at 18 ◦C overnight. The cells were harvested via centrifugation in a Sorvall RC 6 Plus centrifuge (Thermo Fisher Scientific, Inc, MA, USA) at 6080× g, and stored at –20 ◦C. When required, cells were resuspended in a minimal volume of 50 mM Tris-HCl buffer at pH 7.0 supplemented with 500 mM NaCl, 10% glycerol and 0.05% Tween-20, thawed and lysed by sonication on ice. Cell debris was removed by centrifuging at 17,065× g in a Sorvall RC 6 Plus centrifuge (Thermo Fisher Scientific, Inc, MA, USA). The supernatant solution was loaded onto a HiTrap Chelating column (Ni-NTA resin), which was then washed with the same buffer and bound protein eluted using a gradient of 0–500 mM imidazole in the same buffer. Frac- tions containing pure Slt35 were combined and dialysed overnight against 50 mM Tris-HCl buffer at pH 7.0 supplemented with 300 mM NaCl, 10% glycerol and 0.05% Tween-20.

3.3. Enzyme Activity Whole Micrococcus lysodeikticus cells were suspended in assay buffer and incubated at 37 ◦C before adding the puriﬁed enzyme (ﬁnal concentration 0.1 µM for lysozyme and 7 µM for Slt35). The decrease in turbidity was monitored at 600 nm for 1 min at 37 ◦C, and the initial rate determined. A range of different buffers, pH, salt concentrations, substrate concentrations and temperatures were investigated. All investigations used three independent repeats of each condition, and results are expressed as the average ± standard deviation. The optimum assay buffer was found to be 25 mM Tris-maleate Molecules 2021, 26, 4189 9 of 13

(pH 5.8) containing 10 mM CaCl2 for Slt35 and 25 mM Tris-maleate (pH 6.0) containing 100 mM KCl for lysozyme. A 0.6 mg mL−1 cell suspension was used in all subsequent assays. For the substrate dependence of the enzyme activity, the data for the lysozyme were ﬁtted by nonlinear regression using SigmaPlot 10 to the Morrison equation for tight binding of the substrate [41]: q 2 (ET + ST + Km) − (ET + ST + Km) − 4ETST v = vmax (1) 2ET The data for Slt35 were ﬁtted by nonlinear regression using SigmaPlot 10 to an expression for substrate inhibition based on Michaelis–Menten kinetics [40]:

v S v = max (2) S2 Km + S + KI

3.4. Enzyme Inhibition Slt35 or lysozyme was preincubated with the potential inhibitor for 10 min at 0 ◦C before being added to the assay mixture as above (Section 3.3). Three independent repeats were performed for each inhibitor. Initial rate data were normalised between the rate in the absence of the inhibitor and the rate in the absence of the enzyme, and reported as a percent of inhibition (where the normalised rate in the absence of the inhibitor = 0% inhibition and normalised rate in the absence of the enzyme = 100% inhibition). The data were ﬁtted by nonlinear regression using SigmaPlot 10 (Ligand Binding, sigmoidal dose-response, variable slope): max − min %inhibition = min + (3) 1 + 10(logEC50−x)·Hillslope

3.5. Saturation Transfer Difference NMR Saturation transfer difference NMR was performed on a Bruker AVANCE III 600 MHz (1H) NMR spectrometer with a QCI cryoprobe at 25 ◦C. Test compounds (to 5 mM) and Slt35 (to 20 µM) were added to 50 mM potassium phosphate (450 µL ﬁnal volume) with 50 µL of D2O. The spectra were produced using excitation sculpting for solvent suppression, a relaxation delay of 6 s and a saturation time of 5.9 s containing a train of 50 millisecond E-BURP shaped pulses at 0.5 ppm. For each experiment a corresponding control spectrum was taken whereby the saturation was performed in the absence of the enzyme; no signal enhancement was observed in this case.

3.6. Determination of Kd Dissociation constants were determined by fluorimetric titration, using an LS-22 Luminescence Spectrometer (Perkin-Elmer, Buckinghamshire, UK) attached to a Julabo F25 water bath (Julabo, Seelbach, Germany). A total volume of 2 mL of 3.0 µM enzyme in the same buffer as that used for the turbidometry assay was titrated by increasing the concentration of inhibitors. Fluorescence spectra of the enzyme and enzyme–inhibitor solution was recorded after each addition. Three independent repeats were performed for each inhibitor. The excitation wavelength and slit width were 280 and 10 nm respectively, and the emission wavelength and slit width were 290–450 and 5 nm respectively. The maximum emission of lysozyme was at 348 nm while that of Slt35 was at 344 nm. The fluorescence intensity at maximum wavelength was corrected for dilution and for a small inner-filter effect, and converted to a fluorescence difference:

initial f inal ∆If = If − If (4) Molecules 2021, 26, 4189 10 of 13

The ﬂuorescence difference data were then ﬁtted by nonlinear regression using SigmaPlot 10 (Ligand Binding, one site saturation):

Bmax·x ∆If = (5) Kd + x

3.7. Antibacterial Activity

An overnight culture of E. coli JM109 in the LB medium was diluted to an OD500 of 0.5 and 100 µL aliquots that were spread onto fresh LB-agar (2% agar) plates. Ampicillin and thionine acetate were dissolved in water to give 3 mg/mL and 100 mg/mL solutions respectively. Aliquots of these solutions, to give 30 µg ampicillin and 125–1000 µg thionine acetate, were applied to 5 mm diameter ﬁlter paper discs, which were placed onto the surface of the agar. Plates were incubated at 37 ◦C for 24 h and inhibition zones observed. Each compound was tested in triplicate.

4. Conclusions A range of glycosidase inhibitors were found to have no inhibitory activity against either the hen egg white lysozyme or Slt35, whereas thionine acetate exhibited a high micromolar inhibition of both enzymes. Although the structure of thionine does not mimic the structure of peptidoglycan enzymes substrate, molecular docking analysis suggested that it is able to block the active sites of both enzymes. The activity of Slt35 was much more sensitive to environmental conditions such as the salt concentration, pH and temperature than that of the lysozyme and Slt35 was inhibited by high concentrations of the substrate while lysozyme was not. However, fluorescence results revealed strong binding affinities of thionine to both enzymes and inhibition constants were similar for the two enzymes. This work could provide groundwork towards the design of effective inhibitors for LTs based on modifications to the thionine structure.

Supplementary Materials: The following are available online, Figure S1. Saturation transfer difference NMR spectrum (A) and solvent-suppressed 1D 1H NMR spectrum (B) of 20 µM Slt35 and 5 mM 1-deoxynojirimycin, in 50 mM potassium phosphate buffer (pH 7.0) at 25 ºC. Signals at 7–8 ppm in B are from residual imidazole; these signals are not visible in the saturation transfer difference spectrum (A). Figure S2. Comparison of the predicted binding mode of thionine with hen egg white lysozyme (PDB 2VB1) according to BUDE (thionine shown as orange sticks), and thionine manually docked into hen egg white lysozyme (PDB 6LYZ) based on the results of Shanmugaraj et al. (thionine shown as blue-green sticks). The enzyme is shown in pink as a surface representation of the binding site (left), and a cartoon representation showing the amino acid residues (as sticks) located in the active site close to the inhibitor (right). The catalytic Glu35 and Asp52 residues are shown with the backbone ribbon coloured yellow. The predicted hydrogen bonding interactions are shown as green dashes with distances in Å. Figure S3. Comparison of the predicted binding modes of thionine with hen egg white lysozyme (PDB 2VB1) according to BUDE (thionine shown as orange sticks) and AutoDock (thionine shown as green sticks). The enzyme is shown in pink as a surface representation (left) or as a cartoon representation with key active-site residues shown as sticks (right). The catalytic Glu35 and Asp52 residues are shown with the backbone ribbon coloured yellow. Enzyme-ligand hydrogen bonds are shown in green with the distances indicated. Figure S4. Comparison of the predicted binding mode of thionine with hen egg white lysozyme (PDB 6LYZ) according to AutoDock (thionine shown as orange sticks), and thionine manually docked into hen egg white lysozyme (PDB 6LYZ) based on the results of Shanmugaraj et al. (thionine shown as blue-green sticks). The enzyme is shown in pink as a surface representation (left) or as a cartoon representation with key active-site residues shown as sticks (right). The catalytic Glu35 and Asp52 residues are shown with the backbone ribbon coloured yellow. Enzymeligand hydrogen bonds are shown in green with the distances indicated. Figure S5. Comparison of the predicted binding modes of thionine with Slt35 (PDB 1QUS) according to BUDE (thionine shown as orange sticks) and AutoDock (thionine shown as green sticks). The enzyme is shown in blue as a surface representation (left) or as a cartoon representation with key active-site residues shown as sticks (right). The catalytic Glu162 residue is shown with the backbone ribbon coloured yellow. Enzyme-ligand hydrogen bonds are shown in green with the distances Molecules 2021, 26, 4189 11 of 13

indicated. Figure S6. Comparison of the predicted binding mode of thionine (orange sticks) with Slt35 (PDB 1QUS) according to BUDE, and the experimental crystal structure of Slt35 with bulgecin A (grey sticks) bound (PDB 1D0L). The enzyme is shown in blue as a surface representation (left) or as a cartoon representation with key active-site residues shown as sticks (right). The catalytic Glu162 residue is shown with the backbone ribbon coloured yellow. Author Contributions: Conceptualization, A.B.M. and E.J.L.; methodology, A.B.M., R.B.S. and E.J.L.; software, R.B.S.; investigation, A.B.M., C.M.C., G.M.P., E.M.B., A.S., A.C.W. and S.E.A.; writing— original draft preparation, A.B.M. and C.M.C.; writing—review and editing, S.E.A., A.C.W., R.B.S. and E.J.L.; supervision, A.C.W., S.E.A. and E.J.L.; project administration, E.J.L.; funding acquisition, E.J.L. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the UK Biotechnology and Biological Sciences Research Council (BBSRC), grant number BB/L01758X/1; the Life Sciences Research Network Wales, grant number NRNRG4Mar039; the Libyan Embassy (London), PhD studentship 10007 to A.B.A.M.; the Royal Society of Chemistry, Undergraduate Research Bursary to G.M.P.; Cardiff University and Swansea University. The APC was funded by Swansea University. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Data are contained within the article or Supplementary Materials. Conﬂicts of Interest: The authors declare no conﬂict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. Sample Availability: The plasmid containing the slt35 gene is available from the authors.

References 1. Vollmer, W.; Blanot, D.; de Pedro, M.A. Peptidoglycan structure and architecture. FEMS Microbiol. Rev. 2008, 32, 149–167. [CrossRef][PubMed] 2. Barreteau, H.; Kovaˇc,A.; Boniface, A.; Sova, M.; Gobec, S.; Blanot, D. Cytoplasmic steps of peptidoglycan biosynthesis. FEMS Microbiol. Rev. 2008, 32, 168–207. [CrossRef][PubMed] 3. Goodell, E.W. Recycling of murein by Escherichia coli. J. Bacteriol. 1985, 163, 305–310. [CrossRef][PubMed] 4. Sharma, A.K.; Kumar, S.; Harish, K.; Dhakan, D.B.; Sharma, V.K. Prediction of peptidoglycan hydrolases- a new class of antibacterial proteins. BMC Genom. 2016, 17, 411. [CrossRef][PubMed] 5. Höltje, J.V. From growth to autolysis: The murein hydrolases in Escherichia coli. Arch. Microbiol. 1995, 164, 243–254. [CrossRef] 6. Van Heijenoort, J. Peptidoglycan hydrolases of Escherichia coli. Microbiol. Mol. Biol. Rev. 2011, 75, 636–663. [CrossRef][PubMed] 7. Scheurwater, E.; Reid, C.W.; Clarke, A.J. Lytic transglycosylases: Bacterial space-making autolysins. Int. J. Biochem. Cell Biol. 2008, 40, 586–591. [CrossRef][PubMed] 8. Dik, D.A.; Marous, D.R.; Fisher, J.F.; Mobashery, S. Lytic transglycosylases: Concinnity in concision of the bacterial cell wall. Crit. Rev. Biochem. Mol. Biol. 2017, 52, 503–542. [CrossRef][PubMed] 9. Höltje, J.V.; Mirelman, D.; Sharon, N.; Schwarz, U. Novel type of murein transglycosylase in Escherichia coli. J. Bacteriol. 1975, 124, 1067–1076. [CrossRef][PubMed] 10. Jacobs, C.; Frère, J.-M.; Normark, S. Cytosolic Intermediates for Cell Wall Biosynthesis and Degradation Control Inducible β-Lactam Resistance in Gram-Negative Bacteria. Cell 1997, 88, 823–832. [CrossRef] 11. Adin, D.M.; Engle, J.T.; Goldman, W.E.; McFall-Ngai, M.J.; Stabb, E. V Mutations in ampG and lytic transglycosylase genes affect the net release of peptidoglycan monomers from Vibrio fischeri. J. Bacteriol. 2009, 191, 2012–2022. [CrossRef][PubMed] 12. Boudreau, M.A.; Fisher, J.F.; Mobashery, S. Messenger functions of the bacterial cell wall-derived muropeptides. Biochemistry 2012, 51, 2974–2990. [CrossRef] 13. Sorbara, M.T.; Philpott, D.J. Peptidoglycan: A critical activator of the mammalian immune system during infection and homeosta- sis. Immunol. Rev. 2011, 243, 40–60. [CrossRef][PubMed] 14. Clarke, A.J.; Dupont, C. O-acetylated peptidoglycan: Its occurrence, pathobiological significance, and biosynthesis. Can. J. Microbiol. 1992, 38, 85–91. [CrossRef][PubMed] 15. Pfeffer, J.M.; Strating, H.; Weadge, J.T.; Clarke, A.J. Peptidoglycan O Acetylation and Autolysin Profile of Enterococcus faecalis in the Viable but Nonculturable State. J. Bacteriol. 2006, 188, 902–908. [CrossRef] 16. Monchois, V.; Abergel, C.; Sturgis, J.; Jeudy, S.; Claverie, J.M. Escherichia coli ykfE ORFan gene encodes a potent inhibitor of C-type lysozyme. J. Biol. Chem. 2001, 276, 18437–18441. [CrossRef][PubMed] 17. Clarke, C.A.; Scheurwater, E.M.; Clarke, A.J. The vertebrate lysozyme inhibitor Ivy functions to inhibit the activity of lytic transglycosylase. J. Biol. Chem. 2010, 285, 14843–14847. [CrossRef][PubMed] Molecules 2021, 26, 4189 12 of 13

18. Templin, M.F.; Edwards, D.H.; Höltje, J. V A murein hydrolase is the specific target of bulgecin in Escherichia coli. J. Biol. Chem. 1992, 267, 20039–20043. [CrossRef] 19. Imada, A.; Kintaka, K.; Nakao, M.; Shinagawa, S. Bulgecin A, a bacterial metabolite which in concert with beta-lactam antibiotics causes bulge formation. J. Antibiot. 1982, 35, 1400–1403. [CrossRef][PubMed] 20. Bonis, M.; Williams, A.; Guadagnini, S.; Werts, C.; Boneca, I.G. The Effect of Bulgecin A on Peptidoglycan Metabolism and Physiology of Helicobacter pylori. Microb. Drug Resist. 2012, 18, 230–239. [CrossRef] 21. Skalweit, M.J.; Li, M. Bulgecin A as a β-lactam enhancer for carbapenem-resistant Pseudomonas aeruginosa and carbapenem- resistant Acinetobacter baumannii clinical isolates containing various resistance mechanisms. Drug Des. Devel. Ther. 2016, 10, 3013–3020. [CrossRef] 22. Loveridge, E.J.; Jones, C.; Bull, M.J.; Moody, S.C.; Kahl, M.W.; Khan, Z.; Neilson, L.; Tomeva, M.; Adams, S.E.; Wood, A.C.; et al. Reclassification of the specialized metabolite producer Pseudomonas mesoacidophila ATCC 31433 as a member of the Burkholderia cepacia complex. J. Bacteriol. 2017, 199.[CrossRef][PubMed] 23. Williams, A.H.; Wheeler, R.; Thiriau, C.; Haouz, A.; Taha, M.; Boneca, I.G. Bulgecin A: The Key to a Broad-Spectrum Inhibitor That Targets Lytic Transglycosylases. Antibiotics 2017, 6, 8. [CrossRef][PubMed] 24. Tomoshige, S.; Dik, D.A.; Akabane-Nakata, M.; Madukoma, C.S.; Fisher, J.F.; Shrout, J.D.; Mobashery, S. Total Syntheses of Bulgecins A, B, and C and Their Bactericidal Potentiation of the β-Lactam Antibiotics. ACS Infect. Dis. 2018, 4, 860–867. [CrossRef] [PubMed] 25. Dik, D.A.; Madukoma, C.S.; Tomoshige, S.; Kim, C.; Lastochkin, E.; Boggess, W.C.; Fisher, J.F.; Shrout, J.D.; Mobashery, S. Slt, MltD, and MltG of Pseudomonas aeruginosa as Targets of Bulgecin A in Potentiation of β-Lactam Antibiotics. ACS Chem. Biol. 2019, 14, 296–303. [CrossRef][PubMed] 26. De la Fuente, A.; Martin, R.; Mena-Barragán, T.; Verdaguer, X.; García Fernández, J.M.; Ortiz Mellet, C.; Riera, A. Stereoselective synthesis of 2-acetamido-1,2-dideoxyallonojirimycin (DAJNAc), a new potent hexosaminidase inhibitor. Org. Lett. 2013, 15, 3638–3641. [CrossRef][PubMed] 27. Scofield, A.M.; Witham, P.; Nash, R.J.; Kite, G.C.; Fellows, L.E. Castanospermine and other polyhydroxy alkaloids as inhibitors of insect glycosidases. Comp. Biochem. Physiol. Part A Physiol. 1995, 112, 187–196. [CrossRef] 28. Liu, T.; Xia, M.; Zhang, H.; Zhou, H.; Wang, J.; Shen, X.; Yang, Q. Exploring NAG-thiazoline and its derivatives as inhibitors of chitinolytic β-acetylglucosaminidases. FEBS Lett. 2015, 589, 110–116. [CrossRef][PubMed] 29. Borges de Melo, E.; da Silveira Gomes, A.; Carvalho, I. α- and β-Glucosidase inhibitors: Chemical structure and biological activity. Tetrahedron 2006, 62, 10277–10302. [CrossRef] 30. Shanmugaraj, K.; Anandakumar, S.; Ilanchelian, M. Probing the binding interaction of thionine with lysozyme: A spectroscopic and molecular docking investigation. Dye. Pigment. 2015, 112, 210–219. [CrossRef] 31. Zasonská, B.; Cadkovˇ á, M.; Kováˇrová, A.; Bílková, Z.; Korecká, L.; Horák, D. Thionine-Modified Poly(glycidyl methacrylate) Nanospheres as Labels of Antibodies for Biosensing Applications. ACS Appl. Mater. Interfaces 2015, 7, 24926–24931. [CrossRef] [PubMed] 32. Kang, X.B.; Pang, G.C.; Liang, X.Y.; Wang, M.; Liu, J.; Zhu, W.M. Study on a hydrogen peroxide biosensor based on horseradish peroxidase/GNPs-thionine/chitosan. Electrochim. Acta 2012, 62, 327–334. [CrossRef] 33. Wainwright, M.; Phoenix, D.A.; Laycock, S.L.; Wareing, D.R.; Wright, P.A. Photobactericidal activity of phenothiazinium dyes against methicillin-resistant strains of Staphylococcus aureus. FEMS Microbiol. Lett. 1998, 160, 177–181. [CrossRef][PubMed] 34. Hash, J.H. Measurement of bacteriolytic enzymes. J. Bacteriol. 1967, 93, 1201–1202. [CrossRef][PubMed] 35. Van Asselt, E.J.; Dijkstra, B.W. Binding of calcium in the EF-hand of Escherichia coli lytic transglycosylase Slt35 is important for stability. FEBS Lett. 1999, 458, 429–435. [CrossRef] 36. Davies, R.C.; Neuberger, A.; Wilson, B.M. The dependence of lysozyme activity on pH and ionic strength. Biochim. Biophys. Acta Enzymol. 1969, 178, 294–305. [CrossRef] 37. Vocadlo, D.J.; Davies, G.J.; Laine, R.; Withers, S.G. Catalysis by hen egg-white lysozyme proceeds via a covalent intermediate. Nature 2001, 412, 835–838. [CrossRef] 38. Bartik, K.; Redfield, C.; Dobson, C.M. Measurement of the individual pKa values of acidic residues of hen and turkey lysozymes by two-dimensional 1H NMR. Biophys. J. 1994, 66, 1180–1184. [CrossRef] 39. Harris, T.K.; Turner, G.J. Structural basis of perturbed pKa values of catalytic groups in enzyme active sites. IUBMB Life 2002, 53, 85–98. [CrossRef] 40. Shen, P.; Larter, R. Role of substrate inhibition kinetics in enzymatic chemical oscillations. Biophys. J. 1994, 67, 1414–1428. [CrossRef] 41. Morrison, J.F. Kinetics of the reversible inhibition of enzyme-catalysed reactions by tight-binding inhibitors. Biochim. Biophys. Acta Enzymol. 1969, 185, 269–286. [CrossRef] 42. Neville, W.M.; Eyring, H. Hydrostatic Pressure and Ionic Strength Effects on the Kinetics of Lysozyme. Proc. Natl. Acad. Sci. USA 1972, 69, 2417–2419. [CrossRef][PubMed] 43. Kumagai, I.; Sunada, F.; Takeda, S.; Miura, K. Redesign of the substrate-binding site of hen egg white lysozyme based on the molecular evolution of C-type lysozymes. J. Biol. Chem. 1992, 267, 4608–4612. [CrossRef] 44. Yamaguchi, T.; Blázquez, B.; Hesek, D.; Lee, M.; Llarrull, L.I.; Boggess, B.; Oliver, A.G.; Fisher, J.F.; Mobashery, S. Inhibitors for Bacterial Cell-Wall Recycling. ACS Med. Chem. Lett. 2012, 3, 238–242. [CrossRef][PubMed] Molecules 2021, 26, 4189 13 of 13

45. Lee, M.; Hesek, D.; Llarrull, L.I.; Lastochkin, E.; Pi, H.; Boggess, B.; Mobashery, S. Reactions of All Escherichia coli Lytic Transglycosylases with Bacterial Cell Wall. J. Am. Chem. Soc. 2013, 135, 3311–3314. [CrossRef][PubMed] 46. McIntosh-Smith, S.; Price, J.; Sessions, R.B.; Ibarra, A.A. High performance in silico virtual drug screening on many-core processors. Int. J. High Perform. Comput. Appl. 2014, 29, 119–134. [CrossRef][PubMed] 47. Wang, J.; Dauter, M.; Alkire, R.; Joachimiak, A.; Dauter, Z. Triclinic lysozyme at 0.65 Å resolution. Acta Crystallogr. Sect. D 2007, 63, 1254–1268. [CrossRef][PubMed] 48. Trott, O.; Olson, A.J. AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efﬁcient optimization, and multithreading. J. Comput. Chem. 2010, 31, 455–461. [CrossRef][PubMed] 49. Van Asselt, E.J.; Dijkstra, A.J.; Kalk, K.H.; Takacs, B.; Keck, W.; Dijkstra, B.W. Crystal structure of Escherichia coli lytic transglycosylase Slt35 reveals a lysozyme-like catalytic domain with an EF-hand. Structure 1999, 7, 1167–1180. [CrossRef] 50. Van Asselt, E.J.; Kalk, K.H.; Dijkstra, B.W. Crystallographic Studies of the Interactions of Escherichia coli Lytic Transglycosylase Slt35 with Peptidoglycan. Biochemistry 2000, 39, 1924–1934. [CrossRef] 51. Kraft, A.R.; Prabhu, J.; Ursinus, A.; Höltje, J.-V. Interference with Murein Turnover Has No Effect on Growth but Reduces β-Lactamase Induction in Escherichia coli. J. Bacteriol. 1999, 181, 7192–7198. [CrossRef][PubMed] 52. Korsak, D.; Liebscher, S.; Waldemar, V. Susceptibility to Antibiotics and β-Lactamase Induction in Murein Hydrolase Mutants of Escherichia coli. Antimicrob. Agents Chemother. 2005, 49, 1404–1409. [CrossRef][PubMed] 53. Simm, A.M.; Loveridge, E.J.; Crosby, J.; Avison, M.B.; Walsh, T.R.; Bennett, P.M. Bulgecin A: A novel inhibitor of binuclear metallo-β-lactamases. Biochem. J. 2005, 387.[CrossRef][PubMed]