Cation Effect on Fluorescent Sensing of Pyrophosphate by a Bis(Zn–DPA) Probe

Pyrophosphate (P2O7 , PPi) is a key metabolic byproduct produced by some enzymes, such as nucleic acid polymerases, adenylyl cyclase, and aminoacyl–tRNA synthetases. These enzymes are of high interest because they are closely related to not only cellular proliferation and growth but also several diseases including cancers. Moreover, an abnormal concentration of PPi in body fluid often causes pseudogout, osteoarthritis, and urolithiasis. Therefore, there is growing interest in monitoring the activities of these enzymes. The importance of PPi in bioanalytical and clinical aspects urged chemists to develop selective and sensitive detection methods for PPi. In order to analyze the activities of these enzymes, it is highly desirable to measure the concentration of PPi precisely under complex physiological conditions involving various ions and a large excess of competing biological phosphates such as adenosine 50triphosphate (ATP). However, the majority of PPi probes, including bis(Zn– DPA) (DPA = dipicolylamine) complexes such as 1 2Zn (Scheme 1), was tested in a media of low ionic strength, while many enzymatic reactions require high ionic strength. Therefore, a study on the effects of media toward selectivity and sensitivity is necessary to develop a highly selective and sensitive probe for PPi that is suitable for its real-time monitoring. In that regard, it is worth noting that Jolliffe’s macrocyclic host molecules (e.g., 2 Zn2, Scheme 1) showed higher selectivity for PPi in the biologically relevant Krebs buffer system than in low ionic strength buffers. Nevertheless, the underlying principles for why they show high selectivity are still unclear. Therefore, we investigated how the buffer composition affected molecular recognition of PPi and ATP. Normal Krebs buffer (pH 7.4) contains a number of ions such as sodium (Na, 137 mM), potassium (K, 5 mM), magnesium (Mg, 1.2 mM), calcium (Ca, 2.8 mM), chloride (Cl, 150 mM), sulfate (SO4 , 1 mM), and phosphate (PO4 , 1 mM) as well as non-ionic glucose (10 mM). Because aforementioned glucose and anions would have insignificant effects on the molecular recognition of PPi, it is without doubt that cations (especially Mg + and Ca) play critical roles in modulating association constants. Therefore, we conducted a series of titration experiments in several media containing individual cations. Since probe 1 2Zn showed considerable signal enhancement in the presence of either PPi or ATP as depicted in Figure 1, 1 2Zn is a suitable probe for determining association constants (Figures S1–S3, Supporting information). Table 1 summarized the results obtained by fitting the fluorescence intensity at 470 nm (I470) against the analyte concentrations (Figure 2, Figures S4–S6, see the Appendix S1 for detailed methods for fitting). It is noticeable that log Ka,PPi – log Ka,ATP (difference of two individual log Ka’s) and log KPPi/ KATP (obtained by competition experiment) did not fully agree with each other (Table S1). However, we believe that the latter one is more significant because it directly reflects the ability to detect PPi in the presence of large amounts of ATP. Addition of 100 mM NaCl slightly reduced the association constant between 1 2Zn and analytes (Table 1), which supports the fact that intermolecular forces predominantly rely on ion-ion interactions. However, there is only a little change in association constants, indicating that sodium and chloride ions would have negligible effects on binding of 1 2Zn to PPi and ATP. Surprisingly, alkaline earth metal ions, even at 5 mM concentration, efficiently suppressed the association between 1 2Zn and analytes without inducing a substantial increase in the association constant ratio (log Ka,PPi/Ka,ATP). This phenomenon is clearly different from the binding behavior of Jollife’s peptoid hosts, since many of them showed enhanced selectivity in Krebs buffer system. Therefore, we thought that the selectivity difference between ours and Jolliffe’s would be attributed to not only the action of divalent cations but also the molecular structure of host molecules. In order to understand how the alkaline earth metal ions modulate the association constants between 1 2Zn and analytes, we calculated the association constants by assuming that the ions and 1 2Zn competitively bind to analytes, and the results are summarized in Table S2. The new association constants obtained from the solution containing Mg ions agreed with previous ones within the error range. However, the association constants in the presence of Ca Note DOI: 10.1002/bkcs.11494 J. Oh and J.-I. Hong BULLETIN OF THE KOREAN CHEMICAL SOCIETY