Novel fluorescent biosensor for pathogenic toxins using cyclic polypeptide conjugates† Omowunmi A. Sadik‡* and Fei Yan Department of Chemistry, State University of New York at Binghamton, P. O. Box 6000, Binghamton, NY 13902-6000, USA Received (in Columbia, MO, USA) 8th December 2003, Accepted 9th March 2004 First published as an Advance Article on the web 1st April 2004 This work describes a two-step conjugate synthesis of a new fluorescent analog of microcystin-LR and its subsequent utilization for the development of an optical biosensor for cyanobacteria toxins. The biosensor concept is based on the competitive binding between the native microcystin and its fluorescent analog at immobilized alkaline phosphatase en- zymes. Recently, there is a great deal of interests in developing novel approaches to counter the effect of toxins, chemicals and biological warfare agents. Such chemistry should enable the design of tools that can be used not only to detect, but also to effectively combat biochemical warfare. Despite the availability of numerous bio- detection techniques, there are no biosensors that can detect, identify and accurately classify microcystin toxins. From studying the generation and identification of toxins derived from pathogenic bacteria (via the inhibition of phosphatase enzyme activity), we have developed a generic approach for covalent modification of microcystin using cysteine coupled with fluorescent isothiocyanate (MC-Cys-FITC) conjugate. Microcystins are both effective and specific poisons having toxicities that are several orders of magnitude greater than most nerve agents. 1,2 A growing number of bacterial pathogens such as Microcystis have been identified as important food/water-borne pathogens. 3–5 The basic structure of microcystin (MC) is a cyclic heptapeptide and its variation gives rise to more than 50 types isolated and characterized to date. 6–10 Figure 1 shows the structures of the two most extensively studied microcystins (i.e. microcystin-LR and microcystin-RR). Micro- cystin-LR accounts for nearly 90% of the total toxicity. Micro- cystins have a common moiety comprising five amino acids, 3-amino-9-methoxy-10-phenyl-2,6,8-trimethyldeca-4,6-dienoic acid (Adda), N-methyldehydroalanine (Mdha), D-alanine, b-linked D-erythro-b-methylaspartic acid, g-linked D-glutamic acid, and two L-amino acids variants. 6–8 The synthesis of MC-Cys-FITC con- jugate was developed in line with the basic action of microcystins as potent inhibitors of phosphatase enzymes. 9–11 The fact that microcystins inhibit protein phosphatase 1 and 2A enzymes (PP1 and PP2A) raised the possibility of a functional assay for microcystins and associated toxins. To our knowledge, a biosensor of this type has not been reported. 12 However, its usefulness should range from fundamental research to practical applications. In order to prepare the MC-Cys-FITC, a fluorescent active group has to be introduced into the molecules because intact MCs have no fluorophore. There are at least four possible sites for chemical modification of microcystin-LR, i.e. the guanidino group in the arginine residue, two carboxyl groups (in the methylaspartate and glutamate residues) and the CNC double bond in the N-methyldehy- droalanine residue. The arginine residue is not the obvious choice for derivatization because it is not available in some species of Microcystis. Also, its location is too close to the Adda residue, which is essential for activity and toxicity of microcystins. 13–15 Microcystin-LR was reported to retain activity after saturation of the Mdha residue, thus suggesting that this site could provide a potential site for derivatization. 15,16 Hence, we have demonstrated the feasibility of a novel fluorescent biosensor for microcystins using chemical modification of MC-LR. 17,18 MC-Cys-FITC conjugate was synthesized in two steps involving a Michael type addition of thiol available in cysteine to a,b- unsaturated carbonyl of Mdha in microcystin. 13 Characterization was achieved using reverse-phase high performance liquid chroma- tography (HPLC), fluorescence experiments, and kinetic studies. We examined the presence of the new conjugate by injecting a sample of the reaction mixture into the HPLC column. We observed the existence of the microcystin adduct with clearly distinguishable retention times (Fig. 2). The native MC-LR eluted in 11.30 min while cysteine was detected in only 7.72 min. A new peak observed at 9.97 min demonstrates the existence of the MC-Cys adduct. Fluorescence experiments verified that the resulting conjugate is fluorescently active. The emission wavelength for FITC is 528 nm and the conjugate had an emission wavelength at 514 nm. No reproducible results were obtained when the carboxyl groups were used as the target for derivatization, probably due to steric hindrance around † Electronic supplementary information (ESI) available: synthetic approach for microcystin conjugates. See http://www.rsc.org/suppdata/cc/b3/ b316057b/ ‡ Temporary address: Department of Chemistry & Chemical Biology, Harvard University, Cambridge, MA 02138, USA. Fig. 1 Structure of microcystins showing the common moiety composed of different amino acid variants. Fig. 2 High-performance liquid chromatogram of the reaction mixture of microcystin-LR with cysteine. Column: Zorbax C18, 150 3 4.6 mm. Mobile phase: MeOH/50 mM phosphate buffer (pH 3) (60 : 40). Flow rate: 1 mL min 21 . Detection: 238 nm. This journal is © The Royal Society of Chemistry 2004 DOI: 10.1039/b316057b 1136 Chem. Commun., 2004, 1136–1137