Potassium‐ion‐selective fluorescent sensors to detect cereulide, the emetic toxin of B. cereus, in food samples and HeLa cells

Potassium-Ion-SelectiveFluorescent Sensors To Detect Cereulide, the Emetic Toxin of B. cereus,inFood Samples and HeLa Cells Jos8Garc&a-Calvo,[a] Saturnino Ibeas,[a] Eva-Clara Antjn-Garc&a,[a] Tom#sTorroba,*[a] Gerardo Gonz#lez-Aguilar,[b] Wilson Antunes,[c] Elo&sa Gonz#lez-Lavado,[d] and Mjnica L. Fanarraga*[d] This paperisdedicated to the memory of the late Dr.Stefano Marcaccini 1. Introduction Cereulide is an ionophore toxin produced by specific strains of Bacillus cereus.[1] B. cereus causes two different types of food disease:diarrheal and emetic syndrome.[2] Thediarrheal form is developedafter the ingestion of food contaminated with spores or vegetativecells able to produce enterotoxins during vegetativegrowth in the small intestine. The emeticsyndrome is usually associated with the ingestion of foodstuffs contaminated with the preformed toxin, cereulide, produced during the growth of bacteria in food;[3] therefore, intoxicationwith cereulide could be prevented by the early detection of the toxin in food. B. cereus can produce highly resistant structures called endospores, which are capable of survivingtocooking temperatures. Bacillus endospores presentinfoodstuffs can rapidly germinate and produce their toxins,especially during food chilling or food heating.[4] B. cereus is aubiquitous agent in nature, and their spores can be found as anaturalcontaminant in several food products,especially meats, vegetables, and milk.[5] The emeticsyndrome is generally associatedwith rice or other starch-rich products such as pasta and potatobased products.[6] The emetictoxin is acyclic depsipeptide[7] with the structure [d-O-Leu-d-Ala-l-O-Val-l-Val]3,which is closely related to the structure of valinomycin, anothernaturally occurring cyclic depsipeptideand potassium-cation ionophore with the tetramer structure d-a-hydroxyisovaleryl-d-valyl-llactyl-l-valyl [d-O-Hyi-d-Val-l-O-Lac-l-Val]3(Figure 1).[8] The cerWe report the developmentofnew chemical probesfor cereulide, atoxic metabolite produced by specific strains of Bacillus cereus,through displacement of potassium cations from apreformed specific complex and asubsequentchange in the fluorescence emission. For this purpose, we designed fluorescent probesfor potassium cationsthat were suitable for displacement assays with cereulide from organic extracts. The fluorescence detection of natural cereulide in rice samples was achieved by using synthetic cereulide as areference and apotassium fluorescent reporter,and this was found to be useful as aportable and fast method for the in situ detection of cereulide in food extracts.Tostudy the fate of cereulide in live cells, we designed aprocedure that was suitable for live-cell microscopy imaging of HeLa cells by comparing the cellularlocation of the potassium fluorogenic probe, which stained intracellular endolysosomes, in the absence and presence of cereulide;we concluded that in the presence of cereulide, the fluorescence of the probe was decreasedbecause of complexation of the potassiumions by cereulide. Figure 1. Chemical structuresofcereulide and valinomycin. [a] J. Garc&a-Calvo, Dr.S.Ibeas,E.-C. Antjn-Garc&a, Prof. T. Torroba Department of Chemistry,Faculty of Science University of Burgos, 09001Burgos (Spain) E-mail:[email protected] [b] Dr.G.Gonz#lez-Aguilar Centre for Applied Photonics, INESC TEC Rua do Campo Alegre, Dpto. F&sica 4169-007 Porto (Portugal) [c] Dr.W.Antunes LaboratjriodeBromatologia edeDefesa Bioljgica (LBDB) do Ex8rcito Av.Dr. Alfredo Bensafflde, 1849-012 Lisboa (Portugal) [d] E. Gonz#lez-Lavado, Prof. M. L. Fanarraga Grupo de Nanomedicina-IDIVAL Universidad de Cantabria, Santander 39011(Spain) E-mail:[email protected] Supporting Information and the ORCID identification number(s) for the author(s) of this article can be found underhttps://doi.org/10.1002/ open.201700057. T2017 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA. This is an openaccessarticleunder the termsofthe Creative Commons Attribution-NonCommercial-NoDerivs License, which permits useand distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. ChemistryOpen 2017,6,562 –570 T2017 The Authors. PublishedbyWiley-VCH Verlag GmbH &Co. KGaA, Weinheim562 DOI:10.1002/open.201700057 eulide toxin is usually identified by liquid chromatography and mass spectrometry,which require dedicated facilities.[9] Cereulide acts as apotassium-cation ionophoreand is able to disrupt the transmembrane potential in mitochondria of eukaryotic cells;this leads to mitochondrial degeneration and subsequently to cell death.[10] Owing to cereulide toxicity,r apid and portable detection methods to screen the presence of preformed toxin in foodstuffs are required to preventthe occurrence of food-borne outbreaks by emetictoxin. 2. Results and Discussion Fluorescent probes are agood alternative for the detection of toxins or chemical threats.[11] Owingtothe ability of cereulide to complex potassium cations, asuitable methodfor its detection from extracts of biological or food samples should be adisplacementmechanismofthe potassium cation from apreformed specific complex that would give rise to achange in the fluorescenceemission.Following this idea, we designed afluorescent probe for potassium that is suitablefor displacement assays with cereulide in organic solvents. Because cereulide is not soluble in water,arequisite of the fluorogenic probe is that it have good performance in organic or mixed organic–aqueous solvents. Therefore, we selected aperylenemonoimideasthe fluorescentreporter[12] and aK +-selective phenylaza[18]crown-6-lariat-ether[13] or atriazacryptand[14] as the recognition units and bonded both moieties in every case by Suzuki coupling (Scheme 1). With the simplerand more accessible JG76 fluorescent probe,wedesigned the fluorescencedisplacement assays on the basis of Figure 2. The displacementassays were initially performed by using commercial valinomycin instead of lessavailablecereulide. In this way,the JG76 probe was weakly fluorescentinethanol, but in the presence of apotassium chloride solution, adramatic increase in the fluorescencewas observed owing to potassium-ion complexation. As expected, by the addition of avalinomycin solution,the previoussolution becameweakly fluorescent.The differences in fluorescenceare described in Figure 2b. On the other hand, probe JG103 showed very low solubility in ethanol and avery small increase in the fluorescenceinthe presence of potassium cations in all of the tested solvents; therefore, we considered that its utility for the detection of cereulide was very limited and its study was discontinued. Potassium cationsare usually mixed with other ions in biological samples, so we checked carefullythe action of diverse cations and anionsthat could interfereinthe displacementassays with JG76. With respecttoalkali-metal and alkaline-earth metal cations, the JG76 probe showed alarge increaseinfluorescence in the presence of potassium cations with total selectivity with respect to commonly found physiological cations(i.e. Li+,Na +, Mg2+,Ca 2+,and NH4+)aswell as less-common, albeit physiological cations (i.e. Sr2+,Rb +,Cs +)but also showedalargeincrease in fluorescenceinthe presence of the highly toxic Be2+ and Ba2+cations, which are not normally found in biological fluids (Figure 3) except in toxicological assays, for which the JG76 probe could be of interest. The assays were performed by dissolving compound JG76 in EtOH to aconcentration of 50 mmand the cations in water to aconcentration of 5mm;then,the cation solution (10 mL) was added to the JG76 solution (0.5 mL), and changes in the color and fluorescencewere studied. With respect to heavy-metal cations,someacidic cations such as Sn2+and Pb2+showed ahigh increase in fluorescenceinthe presence of JG76, and to alesser extent,some other cations,including Zn2+,Cu 2+,Fe 3+, Scheme1.Synthesis of JG76 and JG103. Figure 2. a) Qualitative andb)quantitative fluorescence displacement assays of JG76,potassium cations, and valinomycin. ChemistryOpen 2017,6,562 –570 www.chemistryopen.org T2017 The Authors. PublishedbyWiley-VCH Verlag GmbH &Co. KGaA, Weinheim563 Sc3+,Al 3+,Hg 2+,Au 3+,and Pd2+,also showed an increasein fluorescenceunder the same conditions (Figure 4) and in the presenceofalarge excess amount of strong acids (pH<5) or oxidants. None of the common anions [i.e. F@,Cl @,Br @,I @, BzO@(Bz=benzoyl), NO3@,H 2PO4@,HSO4@,AcO@,CN @,SCN@] gave asignificant change in color or fluorescence(see the Supporting Information, Figure S98). In apH7 buffered solution [JG76,25mmin EtOH/H2O(7:3 v/v), 20 mm4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)],most of the interferentcations including Fe3+,Sn 2+, Be2+,and Cu2+(5 equiv) did not promote achange in the fluorescenceofJG76, but the sensitivity to K+,Ba 2+,and Pb2+remained unchanged,and they showedahigh increase in the fluorescenceofJG76;therefore, the only real interferents were the nonphysiological Ba2+and Pb2+cations (Figure5). Job’s plot analysis, by representing XJG76(Fo@F)versus XJG76 (where XJG76 =mole fraction of JG76), afforded the stoichiometry of the complex, from the maximumposition at XJG76 =0.5, as a1:1 complex. The calculated quantum yield of the JG76 probe in ethanol is F[JG76] (EtOH)=0.04:0.01 and for the correspondingpotassium complex, F[JG76–K+](EtOH)= 0.17:0.01;the increase is F/F0=4, whichissufficient for the experimental assays. The lifetimeofJG76inEtOH was found to be 3.6 ns, and there was no changeinthe lifetimefor the potassium complex. The limit of detection for K+,obtained by linear regression of atitration of a5mmsolution of JG76 in EtOH and K+by increasing the K+concentration from 0.01 to 0.2 mmand by measuring the increaseinthe fluorescence emission, was calculated as 0.06 mm,with aprobability of false positive and false negative lower than 5%.The JG76–K+complex was studied by preparinga2mmsolution of K(CF3SO3)in EtOH and by increasing the concentration of JG76 from 0to 10 mmwithout changing the concentrationofK +in solution. The equilibrium constant (K)ofthe JG76–K+complex was calculated by fitting the fluorescencetitration plot in ethanol by nonlinearleast-squareregression;the results were compared by performing first the titrationofK +with JG76 andthen that of JG76 with K+,and the same results were obtained. The fitting calculation of the complexation constants was repeated three times to give logK(JG76+K+)=6.34:0.04. To validate the results, Job’s plot analysisand the equilibrium constant of the JG76–K+complex were also studied andcalculated in benzylalcohol, BnOH, and comparable resultswereobtained: a1:1 complex and logK(JG76–K+)=6.11:0.03 (Figure 6). We next checked the suitability of the system for the practical detection of valinomycin by calculating the limit of detection of valinomycin in ethanol. Therefore, 0.75 equivalents of K+ CF3SO3@was added to a5mmsolution of JG76 in ethanol. Then, the concentrationofvalinomycin was increased by successive additions, and the fluorescencespectrum was registered. Regression of the titration plot of 5 mmJG76and 3.75 mmK+in EtOH with valinomycin by decreasing the fluorescence emission gave adetection limit of 0.54 mm,with Figure 3. Color and fluorescence assays of JG76 and alkali-metal and alkaline-earthmetal cations (1 equiv). Figure 4. Color and fluorescence assays of JG76 and selected heavy-metal cations(1equiv). Figure 5. a) Color and fluorescence assays of JG76 and selected cations in abuffered pH 7solution. b) Quantitative fluorescence assays of JG76 and selectedcations in ethanol(left) and abuffered pH 7solution (right). Figure 6. Fitted titration plot of a2mmsolution of K+and JG76 by fluorescenceemission in EtOH and BnOH.Inset:Job’s plot of JG76and K+by fluorescenceemission in EtOH and BnOH. ChemistryOpen 2017,6,562 –570 www.chemistryopen.org T2017 The Authors. PublishedbyWiley-VCH Verlag GmbH &Co. KGaA, Weinheim564 aprobability of false positiveand false negative lower than 5% (see Figure S123). The JG76 fluorescent probe was then ready for titrationwith synthetic cereulide, which was prepared according to the procedure of Biesta-Peters et al.[15] The last step, consisting of the cyclization of the linear H-(l-Val-d-O-Leu-d-Ala-l-O-Val-l-Val-dO-Leu-d-Ala-l-O-Val-l-Val-d-O-Leu-d-Ala-l-O-Val)-OH depsipeptide, is shown in Scheme 2. Purifiedcereulide was checked by NMR spectroscopy and HRMS (MALDI), and the results were identical to the reported data of cereulide. We first checked the suitability of the system for the practical detectionofsyntheticcereulide by calculating the limit of detection of cereulide in ethanol. In an experiment similar to that previously explained, K+CF3SO3@(0.75 equiv) was added to a5mmsolution of JG76 in ethanol. Then, the concentration of cereulide was increased by successive additions, and the fluorescencespectrum was registered. Regression of the titration plot of 5 mmJG76 and 3.75 mmK+in EtOH with cereulide by decreasing the fluorescence emission gave adetection limit of 0.21 mm,with aprobability of false positive and false negative lower than 5% (Figure 7);inthis case, the detection limit for cereulide was lower than the value obtained for valinomycin. Acomplete description of the system was then performed by calculating the equilibrium constants between valinomycin/ cereulide and the potassium cations in competitive equilibria with JG76, for which we previously obtained the binding constants.Equilibrium constantswere first measured in EtOH solution by increasing the concentration of the probe in the presence of the analyte. Weperformedtitrations of constant concentrationsof2mmK+and 20 mmvalinomycin or cereulide with JG76 (0 to 20 mm). In the case of valinomycin, the experiment started with solutions of valinomycin (V) and potassium cations (K) in ethanol sothat the complex (VK) was produced in the equilibrium. Upon adding the JG76probe (abbreviated S), it formed acomplex with free K+and created anew complex (SK) and replaced the previous complex (VK). In this way, the concentration of VK decreased, whereas the concentration of Vincreased. The fluorescenceintensity only depended on JG76 and the JG76–potassium complex. To solve the equations corresponding to the system, apossible approximation could be done, CV@[VK]&CV,inwhich CVis the concentration of valinomycin, and this simplification is more realistic if the initial ratio of V/K is as high as possible. In this way, an equation to describe the equilibria was obtained [Eqs. (1)–(5)]: SþKK1 K! KSK ð1Þ VþKK2 K! KVK ð2Þ K1¼½SKA ½SA½KA¼½SKA ðCS@½SKAÞðCK@½SKA@½VKAÞð3Þ K2¼½VKA ½VA½KA¼½VKA ðCV@½VKAÞðCK@½SKA@½VKAÞð4Þ ½SKA¼ CSþCKþ1þK2CV K1 0/ @ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi CSþCKþ1þK2CV K1 0/ 2@4CSCK r2ð5Þ in which CS,CK,and CVare the concentrations of JG76, K+,and valinomycin, respectively. By taking into account the mass and fluorescence balances for all equilibria and performing the tests several times with different initial ratios of V/K (1:1, 1:0.25, and1:0.1), the best fitting of the obtained equation was found at a1:0.1 ratio, at which the approximation was most valid. By alternating the concentrations of valinomycin and cereulide in several experiments,fitted titration plotswere obtained for both valinomycin and cereulide in EtOH and BnOH (Figure 8). From the averageofthe three titrations, we obtained the following complexation equilibrium constants: logK(valinomycin–K+,EtOH)=5.97:0.01, logK(valinomycin–K+,BnOH)= 4.98:0.01, logK(cereulide–K+,EtOH)=5.99:0.01, K(cereulide–K+,BnOH)=5.01:0.01. With the values of K2and K1,the concentration of [SK]eq could be determined, and with these data the concentration of the specieswas calculated by using Equations (6)–(9). With the data previously obtained, the amountsofreagents and the proportion of the complex existing during the titrationcouldberepresented (Figure 9). Scheme2.Cereulide synthesis. Figure 7. Linear regression of the titration plot of 5 mmJG76 and 3.75 mmK+ in EtOH with increasing amountsofcereulide by decreasing the fluorescence emission of the system. ChemistryOpen 2017,6,562 –570 www.chemistryopen.org T2017 The Authors. PublishedbyWiley-VCH Verlag GmbH &Co. KGaA, Weinheim565 ½SAeq ¼CS@½SKAeq ð6Þ ½VKAeq ¼ CVþCK@½SKAeq þ1 K2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi CVþCK@½SKAeq þ1 K2 0/ 2@4CVðCK@½SKAeqÞ r2 ð7Þ ½VAeq ¼CV@½VKAeq ð8Þ ½KAeq ¼CK@½SKAeq @½VKAeq ð9Þ To validate the method,wecalculated the binding constant of valinomycin with potassium cationsbycircular dichroism following the reported methodology.[16] Starting with aconcentration of 30 mmof valinomycin in EtOH, the concentration of the potassium cations was gradually increased up to 3.75 equivalents. This concentration of valinomycin was selected because of the optimal concentration to follow changes in the dichroism signal, albeit it is too high to perform safe measurements with toxic cereulide.Bychecking the fitting at different wavelengths,the value of the constant appeared to be different depending on the value of the wavelength selected. Within the l=230–250nmrange (at l=238 nm the obtained logK=5.82 value was similar to that previously reported), the calculated range of values for the complexation constant was logK(valinomycin–K+,EtOH)=5.8–6.0, which is compatible with the values obtainedbyfluorescenceassays.Therefore, the developed methodology was validated as ameans to study the fluorescencedetection of cereulide from natural samples. To accomplish this objective, the proposed method involved the competition between cereulide and the JG76probe for potassium.Extraction of natural cereulide was performed on cultures of B. cereus F4810/72 by following the methodology developedfor cookedrice,[17] which consisted of inoculation of rice with 300 Cfu (average value found in rice dishes). Cfu was determined at several time points, and cereulide production was checked at severaltime points by MS (UPLC-TOF). We measured the change in fluorescenceinthe presence of aconstant concentrationofthe probe, aconstant concentration of potassium, and an unknown concentrationofthe cereulide sample (usually within the range of 0.2 to 3.5 mmin acetonitrile), which we titrated with synthetic cereulide. Initial acetonitrile extracts were evaporated, the residue was extracted into dichloromethane/water and concentrated,and the obtained solid was dissolved in EtOH so that the interference of watersoluble interferents was prevented. The EtOH solution was checked by MS (UPLC-TOF) by studyingthe elutiontime of synthetic cereulide, and the LC–MSofthe extracted cereulide samples were measured and calibrated with synthetic cereulide at different concentrations. In this way,the concentration of the cereulide samples from rice remained between1.2 and 1.6 mm[as determined by MS (UPLC)].Wethen measured the background fluorescence due to the matrix of the sample. We performed apreliminary titration of the sample with JG76 to assure that the background fluorescence was aconstant value independentofthe amount of JG76 (Figure 10). With the extracted cereulide samples we measured variations in the emission signal by subtracting the background fluorescenceofthe matrix,measured before adding JG76, and titrated the samples with synthetic cereulide in the presence of JG76 (2 mm)and K+(0.75 mm). With this titration we obFigure 8. Fitted titration plots by fluorescence emission of 2 mmsolutions of K(CF3SO3)and a20mmsolution of valinomycin or cereulide with JG76in EtOH and BnOH:a)K +,valinomycin, and JG76 in EtOH;b)K +,valinomycin, and JG76 in BnOH;c)K +,cereulide, and JG76 in EtOH;d)K +,cereulide, and JG76 in BnOH. Figure 9. Concentrations in equilibrium in EtOHsolution. Top) V(valinomycin), VK (valinomycin–K+), K(K +), and SK (JG76–K+)inequilibrium.Bottom) comparison of the concentrations of the complexes corresponding to valinomycin and cereulide and CK (cereulide–K+)inequilibrium. ChemistryOpen 2017,6,562 –570 www.chemistryopen.org T2017 The Authors. PublishedbyWiley-VCH Verlag GmbH &Co. KGaA, Weinheim566 served that the first points of the rice sample titrationdeviated from those of the reference due to the quantity of natural cereulide present in the sample;therefore, from the initial values of fluorescenceweobtained the original concentrationofnatural cereulide in the sample by comparison with the reference plot from synthetic cereulide (Figure 11). The result of repeating the test three times was (1.0:0.2) mmfor rice samples compared to the concentration of 1.2 mmobtained by mass spectrometry.Inthe case of asample spikedwith 1.75 mmcereulide, the result of the fluorescencetitration was in accordance with the expected value (Figure 12). If there wasnoinitial cereulide in the sample, the curves were coincident (Figure 13). Indeed, the measurements were more precise for the samples spiked with synthetic cereulidethan for the samples containing naturalcereulide, which is due to the factthat natural samples have more than one type of cereulide.[18] The existence of amixture of naturalcereulides (isocereulides) introduces some uncertainty in the experimental measurements, because their potassium-ion-complexationconstants are not known.The fluorescencedetection of natural cereulide in rice samples by using synthetic cereulide and apotassium fluorescent reporter is therefore auseful, portable, and fast method for the in situ detection of cereulidebysimple extraction and fluorescencetitration of the sample. In the case of positivedetection, the method can be complemented by standardmethodology involving LC–MS formore accurate analysis. Taking into account that aLC–MS (Q-TOF)instrumentisnot acheap or portable device relative to abenchtop fluorometer, the reportedmethod may easily prevent the occurrence of foodborne outbreaks by emetictoxin by in situ detectionof cereulide. Another interesting aspectofcereulide detection is visualization of the action of cereulide in live cells. For this purpose, we performed cellular localization studies in HeLa cells (human cervical carcinoma cells) with the JG76 fluorogenic probe. HeLa cells, cultured under standard conditions,[19] were incubated with the probe (18 mmin 1%v/v DMSO/culture medium). Cells were fixed with 4% paraformaldehyde before taking images. The nuclei of the fixed cellswere stained with Hoechst dye (bisbenzimide) before high-resolution confocal microscopy imaging was performed. All confocalcell images were pseudocolored (Figure 14). After exposure for 12 h, the probe stained intracellular vesicular structures that resembled endolysosomes (Figure 14, top). After exposure for 24 h, the Figure 10. Comparison between ethanol and rice sample solutions uponincreasing the JG76 concentration. The backgroundfluorescence was aconstant value of 77 au. Figure 11. Fluorescence emission of acorrected sample versus the reference by titration with increasing quantities of cereulide. The concentrationo f JG76 was 2 mm,and the concentrationofK +was 0.75 mm.The concentration of cereulide in the rice sample was 0.95 mm. Figure 12. Fluorescence emission of arice sample spiked with 1.75 mmcereulideversus the referencebytitration with increasing quantities of cereulide. The concentration of JG76 was 2 mm,and the concentration of K+was 0.75 mm. Figure 13. Fluorescence emission of arice sample with no cereulide versus the reference by titration with increasingquantities of cereulide. The concentration of JG76 was 2 mm,and the concentration of K+was 0.75 mm. ChemistryOpen 2017,6,562 –570 www.chemistryopen.org T2017 The Authors. PublishedbyWiley-VCH Verlag GmbH &Co. KGaA, Weinheim567 JG76 probe displayed an endosomal-RER pattern clearly localized within the cytoplasmic and endosomal membranes (arrows) (Figure 14, middle). After 120 hofstaining, the JG76 probe wasalso localized in the cytoplasmic membrane (green arrow) (Figure 14, bottom). HeLa cells did not display detectable signs of toxicity if grown in the presence of the potassiumion JG76 fluorescentprobe for up to 120 h. Then, HeLa cells, cultured under standard conditions, were incubated with the probe (0.1 mgmL@1)for 2h.Samples of cells were fixed with 4% paraformaldehyde before images were taken. The nuclei of the fixed cells werestainedwith Hoechst dye (bisbenzimide) before high-resolution confocal microscopy imaging was performed. All confocal cell images were pseudocolored(excitation at l=488 nm and emission in the green/red/near-red region) (Figure 15). Similar to the previous experiment, the probe stained intracellular vesicular structures (Figure 15, top left). After exposure to JG76 for 2h,the cells were exposed to synthetic cereulide (0.2 mgmL@1)for another 2h(Figure 15, top right). After exposure to JG76 and cereulide for 2h,the near-red emission of the endosomes was diminished (more pink, less green in pseudocolor), andthe staining of the cytoplasmic membrane also diminished. The cytosol appeared to be more stainedwith the probe. At 24 hafter the addition of cereulidetothe cells exposed to JG76, the cellular viability decreased significantly and several cells appeared wrinkled and showed membrane blebbing and cytosolvacuolization. From the images it is clear that the initial fluorescenceof JG76, complexed with potassium ions in the potassium-rich structures of the cells, is quenched by cereulide in HeLa cells through potassium-ion-displacement complexation, and this leaves only the residual fluorescenceofJG76 in the membranesofthe cells as evidence for the action of cereulide. Even more, morphological changes in the HeLa cells, such as the formation of large vacuoles andmembrane blebbing, as aresult of the action of cereulide with time are easily followed by the residual fluorescenceofJG76 on the membranes;this provesthe efficiency of the JG76 probe as acereulide chemical sensor by potassium-ion-complexation displacement.Therefore, the JG76 fluorescent probe can be considered as auseful tool for the visualization of cereulideinlive cells andfor the loFigure 14. Top) Confocal microscopy projection images of the JG76 probe in HeLa cells 12 hafter staining; right) Z-lateral projection images of the locations of the stained structures inside fixed HeLa cells 12 hafter staining. Nuclei are stained with Hoechst dye (blue channel). Middle) Confocal microscopy projection imagesofthe JG76 probe in HeLa cells 24 hafter staining obtained by exciting the probesequentially with l=488, 562, and 638 nm lasers. Different fluorophore emissions are pseudocoloredintheir respective wavelengths (green:500–550 nm;red:570–620 nm, purple:662–737 nm). Bottom)Confocal microscopy projectionimages of the JG76 probe in HeLa cells 120 hafter staining. Figure 15. Topleft) Confocal microscopy projection imagesofthe JG76 probeinHeLa cells 2hafter staining obtained by exciting the probe sequentially with a l=488 nm laser.Nuclei are stained withHoechst dye (blue channel). Topright) Confocal microscopy projection images of the JG76 probeinHeLa cells 2hafter the addition of cereulide to the previous sample. Bottom left) Confocalmicroscopy projection image of the JG76 probeinHeLa cells 24 hafter the addition of cereulide to the first sample showing large dark vacuoles. Bottom right) Fluorescencemicroscopy image of HeLa cells 12 hafter the addition of cereulide, used as areference. Nuclei are stained with Hoechst dye (bluechannel). ChemistryOpen 2017,6,562 –570 www.chemistryopen.org T2017 The Authors. PublishedbyWiley-VCH Verlag GmbH &Co. KGaA, Weinheim568 calization of highly polar potassium-rich structures,incomparison to low-polarity membrane structures, from live cells. 3. Conclusions In conclusion, we developed afluorogenic procedure that was able to detect cereulide, atoxic metabolite produced by specific strains of Bacilluscereus,inrice samples through displacement of potassium cations from apreformed specific complex with asubsequentchangeinthe fluorescenceemission.The designed fluorescent probe for potassium cationswas suitable for displacementassays with cereulidefrom organic extracts so that the fluorescencedetection of natural cereulide in rice samples was achieved,and this was proven to be aportable and fast methodfor the in situ detection of cereulide in food extracts.Tostudy the fate of cereulide in live cells, we designed aprocedure suitable for live-cell microscopy imaging of HeLa cells by comparing the cellular location of the potassium fluorogenic probe, which stainedintracellular vesicular structures that resembled endolysosomes, in the absence and presence of cereulide.Wewere able to conclude that in the presence of cereulide the fluorescenceofthe probe was decreased because of complexation of the potassium ions by cereulide. Experimental Section Synthesis of JG76 Pd(PPh3)4(10 mg, 5mol%) was added to asolution of 16-[4bromo-2-(2-methoxyethoxy)phenyl]-1,4,7,10,13-pentaoxa-16-azacyclooctadecane (JG70;100 mg, 0.16 mmol) in toluene/nBuOH (15 mL/4 mL) under anitrogen atmosphere in a100 mL Schlenk flask. Then, asolution of N-[1-(1-adamantyl)ethyl]-8-pinacolylboronateperylene-3,4-dicarboxylmonoimide (JG75;81mg, 0.16 mmol) in toluene/nBuOH/water (5 mL:1.5 mL:0.5 mL) was added, followed by Na2CO3(174 mg, 1.64 mmol), and the mixture was heated under reflux for 24 h. The mixture was then poured into water (100 mL), extracted with CH2Cl2(3V100 mL), and worked-up, and then the residue was purified by column chromatography (silica gel, CH2Cl2/MeOH 92:8 v/v) to give N-[1-(1-adamantyl)ethyl]-8-[4- (1,4,7,10,13-pentaoxa-16-azacyclooctadecan-16-yl)-3-(2-methoxyethoxy)phenyl]perylene-3,4-dicarboxylmonoimide (JG76;58mg, 40%) as apurple solid. M.p. 135–1368C. 1HNMR (300 MHz, CDCl3): d=8.53 (m, 2H,C ArH), 8.49–8.33 (m, 3H,C ArH), 8.02 (m, 1H,C ArH), 7.59–7.54 (m, 2H,C ArH), 7.25–6.91 (m, 4H,C ArH), 5.10 (q, J=7.1 Hz, 1H,CH), 4.23 (m, 2H,CH 2), 3.80–3.54 (m, 24H, 12CH2), 3.66–3.40 (s, 5H,CH 2+CH3), 1.98 (m, 3H,3CH), 1.85–1.81 (m, 3H,1.5CH2), 1.73–1.62 ppm (m, 12H, 4.5CH2+CH3). 13CNMR (100 MHz, CDCl3): d=173.7 (C=O), 165.7 (CAr), 165.0 (CAr), 136.8 (CAr), 132.6–123.2 (CAr +CHAr), 121.9 (CAr), 121.0 (CAr), 120.1 (CArH), 116.5 (CAr), 72.1–67.0 (CH2), 59.3–58.2 (CH +CH3), 40.5 (CH), 38.9 (CH2), 38.2 (Cq), 37.2 (CH), 34.2 (CH), 32.1–29.5 ppm (CH +CH3). IR (KBr): n ˜=2955, 2924, 2848, 1738 (C=O), 1692 (C=O), 1685 (C=O), 1651, 1590, 1571, 1506, 1457, 1384, 1354, 1248, 1122 [email protected] (MALDI): m/z:calcd for C55H63N2O9:895.4528 [M++H]+;found:895.4535. Synthesis of Triazacryptand-perylenemonoimide JG103 Pd(PPh)3(5.6 mg, 5mol%) was added to asolution of bromotrizacryptand JG101 (59 mg, 0.097 mmol) dissolved in toluene/nBuOH (10 mL:3.3 mL) under anitrogen atmosphere in a100 mL Schlenk flask. Then, boronic ester JG75 (82 mg, 0.10 mmol) dissolved in toluene/nBuOH (3.5 mL:1 mL) was added dropwise. Then, Na2CO3 (102.2 mg, 0.97 mmol) dissolved in water (3 mL) was added, and the mixture was stirred under reflux for 24 h. The mixture was poured into water (30 mL), and the product was extracted with CH2Cl2(3V100 mL). After workup, the solid residue was purified by column chromatography (silica gel, CH2Cl2/MeOH 50:4) from which the triazacryptand-perylenemonoimide JG103 (48 mg, 42%) was obtained as apurple solid. M.p. 193–1958C. 1HN MR (300 MHz, CDCl3): d=8.53 (m, 2H,C ArH), 8.49–8.33 (m, 3H,C ArH), 8.02 (m, 1H, CArH), 7.59–7.54 (m, 2H,C ArH), 7.25–6.91 (m, 4H,C ArH), 5.10 (q, J= 7.1 Hz, 1H,CH), 4.23–3.50 (m, 39H, 18CH2+CH3), 2.26–2.20 (m, 6H,2CH3), 1.98 (m, 3H,3CH), 1.85–1.81 (m, 2H,CH 2), 1.73– 1.62 ppm (m, 13H, 6CH2+CH3). 13CNMR (100 MHz, CDCl3): d= 165.8 (CAr), 165.1 (CAr), 153.2 (CAr), 137.3 (CAr), 132.9, 132.3, 132.2, 130.0, 129.6, 128.9, 128.5, 128.2, 127.0, 126.8, 126.7, 123.5, 122.3, 121.3, 120.4, 120.1, 114,8, 114.4, 114.2, and 110.0 (CAr +CHAr), 71.4, 71.0, 70.8, 70.5, 69.6, 68.2, and 67.3 (CH2), 59.2 (CH3), 58.2 (CH), 53.6–52.5 (CH2), 40.4 (CH), 38.2 (Cq), 37.1 (CH), 31.7, 31.1, 29.2, and 29.0 (CH +CH2), 21.3 (CH2), 14.3 and 13.3 ppm (CH3). IR (KBr): n ˜= 2955, 2922, 2856, 1736 (C=O), 1696 and 1682 (C=O), 1651, 1592, 1557, 1509, 1456, 1351, 1250, 1170, 1119, 1106, 1049, 959, 812, 750, 721, 697, 667 [email protected] (ESI): m/z:calcd for C73H82N4NaO10: 1197.5923 [M++Na]+;found:1197.6001. Acknowledgements We gratefully acknowledgefinancial supportfromthe Ministerio de Econom&ayCompetitividad, Spain (Projects CTQ2015-71353-R and AES-PI16/000496), Junta de Castilla yLejn, Consejer&adeEducacijnyCultura yFondo Social Europeo (ProjectBU232U13), and the European Commission, Seventh Framework Programme (Project SNIFFERFP7-SEC-2012–312411). J. G.-C. thanks Ministerio de Econom&ayCompetitividad for his predoctoral FPU fellowship. Conflict of Interest The authors declare no conflict of interest. Keywords: cereulide ·fluorescent probes ·potassium · sensors ·valinomycin [1] a) L. P. Stenfors Arnesen, A. Fagerlund, P. E. Granum, FEMS Microbiol. Rev. 2008,32,579–606;b)E.Granum, T. Lund, FEMS Microbiol. Lett. 1997, 157,223–228. [2] G. Lecking, M. K. Dommel, S. Scherer,A.Fouet, M. Ehling-Schulz, Microbiology 2009,155,922–931. [3] S. Ceuppens, N. Boon, M. Uyttendaele, FEMSMicrobiol. Ecol. 2013,84, 433–450. [4] M. Kranzler,K.Stollewerk, K. Rouzeau-Szynalski, L. Blayo,M.Sulyok, M. Ehling-Schulz, Front. Microbiol. 2016,7,1640, https://doi.org/10.3389/ fmicb.2016.01640. [5] Y. Cui, Y. Liu, X. Liu, X. Xia, S. Ding, K. Zhu, Toxins 2016,8,156, https:// doi.org/10.3390/toxins8060156. [6] N. A. Logan, J. Appl. Microbiol. 2011,112,417–429. [7] a) N. A. Magarvey,M.Ehling-Schulz, C. T. Walsh, J. Am. Chem. Soc. 2006, 128,10698–10699;b)M.Ehling-Schulz, E. Frenzel, M. Gohar, Front. Microbiol. 2015,6,704, https://doi.org/10.3389/fmicb.2015.00704;c)G. Lecking, E. Frenzel,A.Retschle, S. Marxen, T. D. Stark, T. Hofmann, S. ChemistryOpen 2017,6,562 –570 www.chemistryopen.org T2017 The Authors. PublishedbyWiley-VCH Verlag GmbH &Co. KGaA, Weinheim569 Scherer,M.Ehling-Schulz, Front. Microbiol. 2015,6,1101, https:// doi.org/10.3389/fmicb.2015.01101. [8] Cereulide:a)A.Makarasen, K. Yoza,M.Isobe, Chem. Asian J. 2009,4, 688–698;b)A.Makarasen, T. Nishikawa,M.Isobe, Synthesis 2009, 2184–2204;valinomycin:c)C.Annese, D. I. Abbrescia, L. Catucci, L. D’Accolti, N. Denora, I. Fanizza, C. Fusco,G.LaPiana, J. Pept. Sci. 2013, 19,751–757;d)R.M.Iacobazzi, C. Annese, A. Azzariti, L. D’Accolti, M. Franco, C.Fusco, G. La Piana,V.Laquintana, N. Denora, ACS Med. Chem. Lett. 2013,4,1189–1192. [9] a) M. M. H-ggblom, C. Apetroaie, M. A. Andersson, M. S. Salkinoja-Salonen, Appl. Environ. Microbiol. 2002,68,2479–2483;b)T.Bauer,T.Stark, T. Hofmann, M. Ehling-Schulz, J. Agric. Food Chem. 2010,58,1420– 1428;c)V.Fricker,U.Messelhaeßer, U. Busch, S. Scherer,M.EhlingSchulz, Appl. Environ. Microbiol. 2007,73,1892–1898;d)H.T.Rønning, T. N. Asp, P. E. Granum, Food Addit.Contam. A 2015,32,911 –921;e)L. Delbrassinne, M. Andjelkovic, A. Rajkovic, P. Dubois,E.Nguessan, J. Mahillon, J. VanLoco, Food Anal. Methods 2012,5,969–979. [10] a) M. A. Andersson, P. Hakulinen, U. Honkalampi-Hamalainen, D. Hoornstra, J.-C. Lhuguenot, J. Maki-Paakkanen, M. Savolainen,I.Severin, A.-L. Stammati, L. Turco,A.Weber,A.von Wright, F. Zucco,M.Salkinoja-Salonen, Toxicon 2007,49,351–367;b)E.L.J--skel-inen,V.Teplova, M. A. Andersson, L. C. Andersson, P. Tammela, M. C. Andersson, T. I. Pirhonen, N.-E. L. Saris, P. Vuorela, M. S. Salkinoja-Salonen, Toxicol. in Vitro 2003, 17,737–744. [11] a) J. Garc&a-Calvo, S. Vallejos, F. C. Garc&a, J. Rojo, J. M. Garc&a, T. Torroba, Chem. Commun. 2016,52,11915–11918;b)J.Garc&a-Calvo, P. Calvo-Gredilla, M. IbaÇez-Llorente,T.Rodr&guez, T. Torroba, Chem. Rec. 2016,16, 810–824;c)B.D&az de GreÇu, J. Garc&a-Calvo, J. Cuevas, G. Garc&a-Herbosa, B. Garc&a, N. Busto, S. Ibeas, T. Torroba, B. Torroba, A. Herrera,S. Pons, Chem. Sci. 2015,6,3757–3764; d) B. D&az de GreÇu, D. Moreno,T. Torroba, A. Berg, J. Gunnars, T. Nilsson, R. Nyman, M. Persson, J. Pettersson, I. Eklind, P. W-sterby, J. Am. Chem. Soc. 2014,136,4125–4128. [12] a) S. Kaloyanova, Y. Zagranyarski, S. Ritz, M. Hanulov#,K.Koynov,A.Vonderheit,K.Mellen, K. Peneva, J. Am. Chem. Soc. 2016,138,2881–2884; b) U. Lewandowska, W. Zajaczkowski, L. Chen,F.BouilliHre, D. Wang, K. Koynov,W.Pisula, K. Mellen, H. Wennemers, Angew.Chem. Int. Ed. 2014, 53,12537–12541; Angew.Chem. 2014,126,12745–12749; c) A. Bolag, N. Sakai, S. Matile, Chem. Eur.J.2016,22,9006–9014;d)J.A.Hutchison, H. Uji-i, A. Deres, T. Vosch, S. Rocha, S. Meller,A.A.Bastian, J. Enderlein, H. Nourouzi, C. Li, A. Herrmann, K. Mellen, F. De Schryver,J.Hofkens, Nat. Nanotechnol. 2014,9,131–136;e)U.Lewandowska, W. Zajaczkowski, W. Pisula,Y.Ma, C. Li, K. Mellen,H.Wennemers, Chem. Eur.J.2016, 22,3804–3809. [13] a) T. Schwarze, R. Schneider,J.Riemer,H.-J. Holdt, Chem.Asian J. 2016, 11,241–247;b)T.Schwarze, J. Riemer, S. Eidner,H.-J.Holdt, Chem. Eur. J. 2015,21,11306–11310;c)S.Ast, T. Schwarze, H. Meller,A.Sukhanov, S. Michaelis, J. Wegener,O.S.Wolfbeis, T. Kçrzdçrfer,A.Derkop, H.-J. Holdt, Chem. Eur.J.2013,19,14911–14917. [14] a) B. Sui, X. Yue, M. G. Tichy,T.Liu, K. D. Belfield, Eur.J.Org. Chem. 2015, 1189–1192;b)H.He, M. A. Mortellaro, M. J. P. Leiner, R. J. Fraatz, J. K. Tusa, J. Am. Chem. Soc. 2003,125,1468–1469;see also:c)X.Zhou, F. Su, Y. Tian,C.Youngbull, R. H. Johnson, D. R. Meldrum, J. Am. Chem. Soc. 2011,133,18530–18533; d) W. Namkung, P. Padmawar,A.D.Mills, A. S. Verkman, J. Am. Chem. Soc. 2008,130,7794–7795;e)X.Zhou, F. Su, W. Gao, Y. Tian, C. Youngbull,R.H.Johnson,D.R.Meldrum, Biomaterials 2011,32,8574–8583;f)R.D.Carpenter,A.S.Verkman, Org. Lett. 2010, 12,1160–1163;g)X.Li, X. Gao, W. Shi, H. Ma, Chem. Rev. 2014,114, 590–659;h)T.Hirata, T. Terai, H. Yamamura,M.Shimonishi,T.Komatsu, K. Hanaoka, T. Ueno, Y. Imaizumi, T. Nagano, Y. Urano, Anal. Chem. 2016, 88,2693–2700. [15] E. G. Biesta-Peters, M. W. Reij,R.H.Blaauw,P.H.in’t Veld,A.Rajkovic, M. Ehling-Schulz, T. Abee, Appl. Environ. Microbiol. 2010,76,7466–7472. [16] M. C. Rose, R. W. Henkens, Biochim. Biophys. Acta Gen. Subj. 1974,372, 426–435. [17] a) A. ZuberovicMuratovic, R. Trçger,K.Granelli, K.-E.Hellen-s, Toxins 2014,6,3326–3335;b)M.Yamaguchi, T. Kawai, M. Kitagawa, Y. Kumeda, FoodMicrobiol. 2013,34,29–37;c)M.Decleer,A.Rajkovic, B. Sas, A. Madder,S.DeSaeger, J. Chromatogr.A2016,1472,35–43. [18] a) S. Marxen, T. D. Stark, E. Frenzel, A. Retschle, G. Lecking, G. Perstinger, E. E. Pohl, S. Scherer,M.Ehling-Schulz, T. Hofmann, Anal. Bioanal. Chem. 2015,407,2439–2453; b) S. Marxen, T. D. Stark, A. Retschle, G. Lecking, E. Frenzel, S. Scherer,M.Ehling-Schulz, T. Hofmann, Sci. Rep. 2015,5, 10637, https://doi.org/10.1038/srep10637;c)S.Marxen, T. D. Stark, A. Retschle, G. Lecking, E. Frenzel, S. Scherer,M.Ehling-Schulz, T. Hofmann, J. Agric. FoodChem. 2015,63,8307–8313. [19] a) L. Rodriguez-Fernandez,R.Valiente, J. Gonzalez, J. C. Villegas, M. L. Fanarraga, ACS Nano 2012,6,6614–6625;b)L.Garc&a-Hevia, R. Valiente, R. Mart&n-Rodr&guez,C.Renero-Lecuna, J. Gonz#lez, L. Rodr&guezFern#ndez,F.Aguado, J. C. Villegasa, M. L. Fanarraga, Nanoscale 2016,8, 10963–10973; c) B. Sanz, M. P. Calatayud, T. E. Torres, M. L. Fanarraga, M. R. Ibarra, G. F. Goya, Biomaterials 2017,114,62–70. Received: March 21, 2017 Version of record online June 12, 2017 ChemistryOpen 2017,6,562 –570 www.chemistryopen.org T2017 The Authors. PublishedbyWiley-VCH Verlag GmbH &Co. KGaA, Weinheim570