Reaction-based Indicator displacement Assay (RIA) for the selective colorimetric and fluorometric detection of peroxynitrite

Using the self-assembly of aromatic boronic acids with Alizarin Red S (ARS), we developed a new chemosensor for the selective detection of peroxynitrite.


Introduction
Peroxynitrite (ONOO À )a combination of nitric oxide (NOc) and the superoxide radical anion (O 2 À c)was rst discovered as a biological endogenous oxidant in 1990. 1 It is recognised as a strong oxidant with a very short lifetime ($10 ms) in physiological and pathological processes. Peroxynitrite is a highly reactive molecule involved in cell signal transduction, and leads to apoptosis in HL-60 cells, and PC-12 cells. Many biomolecules are oxidised and/or nitrated by peroxynitrite-derived radicals, including DNA, tyrosine residues, thiols, and unsaturated fattyacid-containing phospholipids. Endogenous peroxynitrite formation, and associated protein nitration, has been implicated in Alzheimer's disease, Parkinson's disease, Huntington's disease, amyotrophic lateral sclerosis, viral myocarditis, septic shock, cardiac allogra, transplant coronary artery disease, idiopathic dilated cardiomyopathy, atrial brillation, hypercholesterolemia, atherosclerosis, hypertension, diabetes, diabetic nephropathy, and traumatic brain injury. 2,3 Thus, the importance of peroxynitrite has captured the attention of many groups who seek effective and selective approaches for its detection. Among the powerful tools available for ONOO À detection are small-molecule probes which are attractive owing to their high sensitivity, easy manipulation, coupled with widely available instrumentation. To the best of our knowledge, limited work has been carried out in the development of small-molecule probes for the selective colorimetric sensing of peroxynitrite. 4,5 The successful application of indicator displacement arrays, by Anslyn, [6][7][8][9][10][11] Severin, 12 Singaram, [13][14][15][16][17][18] and reaction-based smallmolecular uorescent probes for chemoselective ROS detection, developed by Chang, 19,20 inspired us to combine these two approaches and develop RIA: "Reaction-based Indicator displacement Assay" for the selective detection of peroxynitrite (ONOO À ) (Scheme 1).
Phenylboronic acid (PBA, pK a ¼ 8.83), benzoboroxole (BBA, pK a ¼ 7.30), and 2-(N,N-dimethylaminomethyl)phenylboronic acid (NBA, pK a ¼ 6.70) are three representative receptors among common boronic acid motifs (Fig. 1), which have been extensively applied in the area of monosaccharides sensing, 21-23 while, Alizarin Red S (ARS), has been successfully employed as a general optical reporter for investigating the binding of boronic acids with carbohydrates ( Fig. 1). 23,24 Also, some of our previous research has used the binding and analyte-mediated release of Alizarin Red S from hydrogelbound boronic acids. 25 Bearing this previous work in mind, we decided to evaluate the sensing systems (ARS-BAs), which are formed by the conjugation of aromatic boronic acid receptors with Alizarin Red S in situ, for the selective detection of peroxynitrite.

Results and discussion
When buffered at neutral pH, we found only small changes occur i.e. only a slight blue-shi was observed in the UV-Vis absorption and a very small uorescence response due to binding of PBA (200 mM), BBA (200 mM), NBA (200 mM) with ARS (50 mM), respectively (FS1, FS2). Also, the almost identical pink colours of the solutions clearly indicate weak binding between ARS and three boronic acid receptors at pH 7.30 (FS2). Therefore, we decided to use a higher pH working environment to enhance the binding of boronic acid with the catechol unit of ARS.
As can be seen from Fig. 2 in 52.1% MeOH/H 2 O PBS buffer (pH 8.10), compounds NBA, PBA and BBA behaved differently with ARS and in particular for the uorescence, there is ca.
15.0-fold uorescence increase for NBA (F/F 0 ) while only a 3.4-fold increase is observed for PBA and a 2.8-fold increase with BBA (FS3), respectively. In terms of absorption, NBA (200 mM) produced the largest shi in wavelength from l max ¼ 520 nm to l max ¼ 465 nm (55 nm blue-shi) on binding with ARS, in comparison with a 10 nm blue-shi for BBA and a 20 nm blue-shi for PBA was observed. A signicant colour change from pink to orange in the presence of NBA (200 mM) with ARS (50 mM) was observed (FS4), while BBA (200 mM) and PBA (200 mM) did not change signicantly even aer one night of stirring. Based upon recent results, the binding between NBA and ARS (binding constant k ¼ 7200 AE 92 M À1 , FS5), is most likely enhanced by the N-B interaction via a solvent insertion (Scheme 2). From the boron NMR (FS6), the NBA boron is clearly sp 3 in nature and remains sp 3 with added ARS under neutral and alkaline conditions (16 ppm) due to solvent insertion. The binding between ARS and NBA results in two species (one major and one minor with approximately 2 : 1 ratio) as described by Benkovic. 26 We have previously demonstrated the N-B interaction provides protection for boronic acids towards common oxidants resulting in selectivity towards stronger reactive oxygen species such as peroxynitrite (ONOO À ). To a preformed sensing ensemble, peroxynitrite (ONOO À ) mediated oxidation of the aryl boronate ARS adduct led to phenols and release of ARS, thereby giving a uorescence decrease and an accompanying red-shifted absorbance wavelength (Fig. 3).
As can be seen from the dose-dependent absorption response in Fig. 3a, when various concentrations of ONOO À (0-1.0 mM) were added to a solution of ARS-NBA (ARS, 50 mM; NBA, 200 mM), a decrease in the absorption at l max ¼ 465 nm was observed with the appearance of a red-shied band centred at 500 nm (A 500nm /A 465nm ¼ 1.15). In the relationship between A 500nm /A 465nm and concentration of peroxynitrite, there was rapid absorption increase in the presence of peroxynitrite below 0.4 mM while the reaction between ARS-NBA and ONOO À (0.4-1.0 mM) was reduced due to saturation (FS7). In Fig. 3b, the uorescence of the ARS-NBA complex was reduced to a ca.  0.45 of its original value (F (in the presence of ONOO À )/F 0 (in the absence of ONOO À )) over a concentration range of ONOO À from 0 to 1 mM. A linear relationship (R 2 ¼ 0.994) was observed between the relative uorescence ((F À F 0 )/F at l 550nm ) and the peroxynitrite (ONOO À ) concentration (0.4-1.0 mM, FS8).
Furthermore, due to the very short lifetime ($10 ms) and low steady-state concentration (nM range) of peroxynitrite, we monitored the time dependant response towards peroxynitrite (using both UV-Vis and Fluorescence methods). From these time-drive experiments the chemical reaction between probe ARS-NBA and peroxynitrite proceeds rapidly (k 0 ¼ 4.39 s À1 , t 1/2 ¼ 0.14 s, k 2 ¼ 8.78 Â 10 4 s À1 M À1 , FS9 and FS10). Using dosedependent titrations, we obtained a LOD (limit of detection) of 5.4 mM for the sensing of peroxynitrite (FS8). Importantly, these results indicate that the ARS-NBA system with an N-B interaction produce signicant colorimetric and uorometric response towards peroxynitrite.
As reported previously, hydrogen peroxide (H 2 O 2 ), hypochlorite (ClO À ) and peroxynitrite (ONOO À ) react with boronatebased compounds to produce the phenol analogues. 27, 28 We have previously demonstrated that the solvent-insertion interaction between the amine and boron provides a protection for the boronic acid towards oxidation and results in the selective detection of peroxynitrite (ONOO À ) over hydrogen peroxide (H 2 O 2 ) and hypochlorite (ClO À ) and other ROS/RNS species. 29,30 Noticeably, we also observed that H 2 O 2 is much more reactive with increasing pH. However, when we use NBA (2-(N,N-dimethylaminomethyl)phenylboronic acid) with ARS this provides a complex with a strong N-B interaction resulting in the selective sensing of peroxynitrite (ONOO À ) over hydrogen peroxide (H 2 O 2 ) and other ROS/RNS species.
Both the absorption and emission spectra (Fig. 4) clearly demonstrate that the RIA system has excellent selectivity for peroxynitrite (ONOO À ) over other common ROS. Additionally, high concentrations of H 2 O 2 (1 mM) were not able to change the uorescence and absorption intensity of the ARS-NBA complex (ARS, 50 mM; NBA, 200 mM) over 60 min (FS11), indicating, that the oxidation-reduction between boron and H 2 O 2 has been prevented by the N-B interaction. The response of the ARS-NBA sensing system to hydroxyl HOc (500 mM), hypochlorite ClO À (500 mM), and peroxynitrite ONOO À (500 mM) (FS12) indicate that peroxynitrite reacts signicantly more ((F À F 0 )/F ¼ ca. 0.45) than with hydroxyl (ca. 0.29) and hypochlorite ClO À (ca. 0.35). Importantly, both, hydroxyl HOc (500 mM) and hypochlorite ClO À (500 mM) respond less than peroxynitrite ONOO À (500 mM) in the RIA system FS12. However, hypochlorite (100 mM) can quench the uorescence and change the absorption of just the ARS dye (FS13). Since hydroxyl is produced from the Fenton reaction (Fe 2+ + H 2 O 2 ), and is inherently coloured. Therefore, while both hydroxyl HOc, hypochlorite ClO À are not ideal ROS to evaluate the RIA colorimetric system, the system remains peroxynitrite selective (FS12). While, the other, ROS/RNS (NO, 1 O 2 , ROOc, O 2 À c) (Fig. 4) do not change the absorption and emission Scheme 2 Sensing mechanism of ARS-NBA complex probe for peroxynitrite.

Solvents and reagents
Solvents and reagents were reagent grade unless stated otherwise and were purchased from Fisher Scientic UK, Frontier Scientic Europe Ltd, TCI UK, Alfa Aesar and Sigma-Aldrich Company Ltd and were used without further purication.

Fluorescence measurements
Fluorescence measurements were performed on a Perkin-Elmer Luminescence Spectrophotometer LS 50B and Gilden Photonics Ltd. FluoroSENS, utilising Starna Silica (quartz) cuvette with 10 mm path lengths, four faces polished. Data was collected via the Perkin-Elmer FL Winlab soware package. All solvents used in uorescence measurements were HPLC or uorescence grade and the water was deionised.

UV-Vis measurement
UV-Vis measurements were performed on a Perkin-Elmer Spectrophotometer. Absorption, utilising Starna Silica (quartz) cuvette with 10 mm path lengths, two faces polished. Data was collected via the Perkin-Elmer Lambda 20 soware package. Further reprocess of the data was operated in OriginPro 8.0 graphing soware.

Preparation of ONOO À
Peroxynitrite stock solution was prepared by the reaction of hydrogen peroxide with sodium nitrite and stabilised in basic solution, which is frequently used for the in vitro experiments. 31 Owing to its short-lived oxidant species, especially in the physiological pH (half-life, 10 ms), the preparation of peroxynitrite was carried out in the sodium hydroxide solution by chemical synthesis with a quenched ow reactor: using equal ow rates of the following solutions at 0 C: 0.6 M KNO 2 , a solution 0.6 M in HCl and 0.7 M in H 2 O 2 , and 3 M NaOH. 31 The product solution was analysed spectrophotometrically. The stock solution of peroxynitrite was fresh-made, and the pH of the stock solution (pH ¼ 13.5) was adjusted by 12 N hydrochloric acid to pH ¼ 12 and re-analysed each time in the detection experiments. The concentration of peroxynitrite was estimated by using an extinction coefficient of 1670 AE 50 cm À1 M À1 at 302 nm in 0.1 N sodium hydroxide (aq.).

Preparation of À OCl, H 2 O 2 and other ROS/RNS
The concentration of À OCl was determined from the absorption at 292 nm (3 ¼ 350 M À1 cm À1 ); 32 the concentration of H 2 O 2 was determined from the absorption at 240 nm (3 ¼ 43.6 M À1 cm À1 ); nitric oxide (NO) was prepared by treating a sodium nitrite solution (7.3 M) with sulfuric acid (3.6 M) and its stock solution (2.0 mM) was prepared by bubbling NO into deoxygenated deionised water for 30 min; 33 1 O 2 was generated by the reaction of H 2 O 2 with NaClO; ROOc was generated from 2,2 0 -azobis(2amidinopropane) dihydrochloride; superoxide was generated from KO 2 ; hydroxyl radical was generated by Fenton reaction.
The UV-Vis spectra and uorescent titrations with hydrogen peroxide were carried out at 25 C in 1/15 M PBS buffer at pH 7.30 and 52.1% methanol/aqueous PBS buffer solution at pH 8.10.

Conclusion
Using an indicator displacement assay induced by a chemical oxidation, we investigated a new chemo/biosensor for the selective sensing of peroxynitrite, using the self-assembly of boronic acid NBA with ARS as the reporter. The ARS-NBA system displays minimal response towards hydrogen peroxide (H 2 O 2 ) and other ROS/RNS due to the protection provided by a B-N solvent-insertion interaction, but a large absorption and uorescence response with peroxynitrite (ONOO À ). The boronic acid receptor NBA is a good candidate for the selective detection of peroxynitrite as part of dye-displacement arrays and for biological applications, such as drug design and cell labelling experiments. The large initial change of uorescence intensity indicates that the system might also be used as a temporal uorescent probe to map intracellular ONOO À . Systems that produce colour changes are particular interesting since they could be incorporated into a diagnostic test paper for ONOO À , similar to universal indicator papers for pH. More importantly, ratiometric probes increase the dynamic range and permit signal-rationing, thus they provide a built-in correction for monitoring environmental effects. 34,35 Therefore, we believe that this simple but powerful RIA system can be extended into new applications for the sensing of reactive oxygen and nitrogen species (ROS and RNS) both in vitro and in vivo.
Scholarship. The Catalysis And Sensing for our Environment (CASE) network is thanked for research exchange opportunities. 36 Notes and references