Sensing Peroxynitrite in Different Organelles of Murine RAW264.7 Macrophages With Coumarin-Based Fluorescent Probes

The elucidation of biological processes involving reactive oxygen species (ROS) facilitates a better understanding of the underlying progression of non-communicable diseases. Fluorescent probes are a powerful tool to study various ROS and have the potential to become essential diagnostic tools. We have developed a series of coumarin fluorescent probes for the selective and sensitive detection of peroxynitrite (ONOO−), a key ROS. Coumarin based probes exhibit good photostability, large Stokes shift and high quantum yields. The three ratiometric probes all contain a boronate ester motif for the detection of ONOO− and a distinctive organelle targeting group. The study of ONOO− generation in a particular organelle will allow more precise disease profiling. Hence, targeting groups for the mitochondria, lysosome and endoplasmic reticulum were introduced into a coumarin scaffold. The three ratiometric probes displayed sensitive and selective detection of ONOO− over other ROS species. All three coumarin probes were evaluated in murine RAW264.7 macrophages for detection of basal and stimulated ONOO− formation.


INTRODUCTION
Macrophages are a diverse population of innate immune cells with roles that include anti-microbial phagocytic activities, wound healing, adipose tissue metabolism and removal of senescent cells (Hirayama et al., 2017). In pathology, macrophages are also linked to the progression of tumors as well as chronic inflammatory diseases and auto-immune conditions such as rheumatoid arthritis Di Benedetto et al., 2019). Many macrophage effector functions are regulated or mediated by the formation of reactive oxygen and nitrogen species including phagocytosis, inflammatory responses, pyroptosis and tumor-associated cytotoxicity (Prolo et al., 2014;Fauskanger et al., 2018;Kelley et al., 2019;Wang et al., 2019). In particular, the powerful oxidant peroxynitrite (ONOO − ) has been shown to play a key role in the control of infections and the regulation of phagocytic activity in macrophages (Prolo et al., 2014). In cellular systems, ONOO − is generated from nitric oxide free radicals (NO·) reacting with superoxide anions (O ·− 2 ) generated by a number of enzymatic pathways including NADPH oxidases. ONOO − can then modify an array of cellular pathways including oxidation and nitration of proteins and lipids. New approaches for site-specific detection of sub-cellular ONOO − will improve our understanding of localized signal transduction and the importance of ONOO − regulation in health and disease.
The development of fluorescent tools to monitor reactive oxygen species (ROS), in particular ONOO − , has lead in recent years to the development of a variety of diagnostic small molecule fluorescent probes (Wu et al., 2017;Wang et al., 2018). Our group has exploited boronic acids and esters as a sensing motif for ONOO − (Odyniec et al., 2018;Weber et al., 2018;Wu et al., 2018). The fluorophore is masked by boronate esters which upon reaction with ONOO − results in fluorophore activation and the emission of a distinctive fluorescence signal (Scheme 1). We have previously used resorufin, fluorescein and other fluorophores as core scaffolds Wu et al., 2019). More recently, we have been interested in the coumarin scaffold since these are widely employed in a variety of applications, notably in the development of anti-cancer, antiinflammatory and neurodegenerative drugs (Jameel et al., 2016) as well as fluorescent tools for the detection of biologically relevant analytes such as ROS (Lo and Chu, 2003;Soh et al., 2008;Yuan et al., 2011). Our motivation to use coumarin as a scaffold was its excellent photostability, large Stokes shift and high quantum yield, making them particularly attractive for imaging applications.
A majority of cellular processes linked to ONOO − production and other ROS takes place inside cellular organelles (Zhu et al., 2016). Different strategies of delivering fluorescent probes to specific organelles have been reviewed by Jiang, Chang, Yuan, and co-workers in 2016 (Xu et al., 2016). Herein, we report the synthesis of three fluorescent probes for ONOO − based on a coumarin scaffold. Each probe was designed to detect ONOO − in a different organelle: mitochondrion, lysosome and endoplasmic reticulum via the incorporation of relevant targeting groups for these organelles and a boronate ester sensing motif into a coumarin scaffold. SCHEME 1 | Fluorophore activation by ONOO − triggered cleavage of (A) Bpin (B) benzyl Bpin.

RESULTS AND DISCUSSION
The Synthesis of Coumarin Based Potential Mitochondria-Targeting (CM), Lysosome-Targeting (CL) and Endoplasmic Reticulum-Targeting (CE) Probes The coumarin probes CM, CL, and CE were easily accessible as outlined in Scheme 2. Briefly, 2,4-dihydrobenzaldehyde's alcohol at the 4 position was selectively substituted using 4-(bromomethyl) benzeneboronic acid pinacol ester. This gave (1a) containing the benzyl Bpin targeting group in 49% yield. This was followed by the addition of Meldrum's acid to achieve the core structure of coumarin bearing a free carboxylic acid group. (1b) Was obtained in 55% yield and required no further purification. (1b) Is a key intermediate in the synthesis of our three potential organelle targeting ONOO − molecular probes. To access CM, (1b) was coupled with (2a) which was prepared from (4-bromobutyl) triphenylphosphonium bromide. Similarly, CE and CL were prepared from (1b) by coupling with (3a) and (4a). All compounds and intermediates were fully characterized by 1 H NMR, 13 C NMR, HRMS, and IR spectroscopy (see Supplementary Material).

Fluorescence Analysis of the Synthesized Probes
In the first instance, we confirmed that upon reaction with ONOO − , CM, CL, and CE generate 7-hydroxycoumarin-3carboxylic acid. Indicating, that the benzyl Bpin sensing unit was successfully cleaved by the ONOO − (Scheme 1B). Figure 1A illustrates the emission spectrum of CM, CL, and CE without ONOO − . All three exhibit moderate fluorescence with maximum emission values at λ max = 400 nm for CM, λ max = 395 nm for CL, and λ max = 405 nm for CE. As expected, 7-hydroxycoumarin-3-carboxylic acid exhibits a maximum emission peak at 447 nm. However, upon addition of ONOO − , the emission spectrum of CM, CL and CE changes, and becomes similar to 7hydroxycoumarin-3-carboxylic acid ( Figure 1B), thus indicating the release of 7-hydroxycoumarin-3-carboxylic acid as the fluorescent species of all three organelle targeting probes. Additionally, the probes have the potential to be used as ratiometric probes due to a significant shift of the emission profile of the probes with and without ONOO − .
We then turned our attention towards ROS selectivity and titrations with ONOO − and H 2 O 2 , plus sensitivity to other ROS. These experiments are to determine the selectivity of CM, CL, and CE for ONOO − and in particular selectivity over H 2 O 2 since boronate esters are good sensing groups for both species. As illustrated in Figures 2-4, all three probes CM, CL, and CE displayed high selectivity towards ONOO − over H 2 O 2 and other ROS. All three probes exhibit the same behavior. Surprisingly, H 2 O 2 has a minimal effect on cleaving the boronic ester moiety demonstrating the strong oxidizing ability of ONOO − for these particular probes. Importantly, the ratiometric nature of probes CM, CL, and CE should allow for the quantitative measurement of ONOO − concentrations. Subsequent titration studies with ONOO − demonstrated that the probes CM, CL, and CE (Figures 5-7) could detect ONOO − (over a range of 1-50 µM) with limits of detection (LoD) determined to be 0.28 µM, 0.26 µM, and 0.36 µM, respectively. Titrations with H 2 O 2 (see Supplementary Material) confirmed the initial finding of the ROS selectivity study that H 2 O 2 was not able to release 7-hydroxycoumarin-3-carboxilic acid since no shift in emission occurs. Therefore, with these probes, we were able to show that CM, CL, and CE were particularly selective for ONOO − , which is particularly beneficial in a cellular environment to selectively detect ONOO − . Therefore, CM, CL, and CE show great promise for the sensitive and selective detection of ONOO − in situ over other ROS species. CM, CL, and CE were also shown to be stable over a pH range from 3 to 9 (Figures S21-S23). In addition, all three probes were shown to produce a fluorescent output at relevant physiological pHs: eight for mitochondria, five for lysosome and seven for endoplasmic reticulum (Figures S24-S26).

Evaluation of CM, CL, and CE for the Detection of ONOO − in RAW 264.7 Macrophages
First, exogenous generation of ONOO − was used to validate whether the probes are cell permeable. SIN-1, an ONOO − donor, acts as a positive control to evaluate the probe's ability to detect ONOO − inside the cell. SIN-1 was incubated for 1 h at a concentration of 100 µM, followed by incubation of the relevant probes CM, CL, and CE for 30 min at a concentration of 20 µM (Figure 8). In the absence of SIN-1, no fluorescence signal was detected for all the probes tested (CM, CE, and CL). Upon treatment with SIN-1, no fluorescence signal was detected in RAW264.7 macrophages loaded with CM and CE probes. We hypothesize that the lack of fluorescence is due to poor cellular uptake of the probes. In contrast, a fluorescence signal was detected in CL-loaded, SIN-1 treated macrophages indicating cellular uptake of the CL probe and reaction with ONOO − .
As such, we focused on the characterization of the potential lysosomal targeting CL probe and evaluated the ability of CL to detect endogenous ONOO − production in RAW264.7 macrophages (Figures 9, 10). To mimic the inflammatory conditions encountered in infections, macrophages were stimulated with lipopolysaccharide (LPS) (1 µg/ml) and interferon (IFN)-γ (50 ng/ml) to trigger the generation of endogenous ONOO − . These conditions have been associated with lysosome-localized ONOO − in murine macrophages (Guo et al., 2018). No basal fluorescence was detected in unstimulated macrophages however LPS and IFN-γ elicited an increase in fluorescence in CL-loaded macrophages (Figure 9). We predicted that pre-incubation with a O ·− 2 scavenger, ebselen, would attenuate the ONOO − -dependent response. It was found that ebselen extinguished the CL-fluorescence response induced by LPS and IFN-γ treatment (Figure 9). Finally, we characterized RAW264.7 macrophages with PMA stimulation (Figure 10), which activates ONOO − production in a different way compared to LPS and IFN-γ. PMA activates protein kinase C which stimulates nicotinamide adenine dinucleotide phosphate oxidase (NOX). NOX are a family of enzyme complexes that generate O ·− 2 from molecular oxygen at the expense of NADPH (Rastogi et al., 2016). Hence, PMA can induce enhanced O ·− 2 production via this pathway.
CL was able to detect ONOO − via PMA stimulation in RAW 264.7 macrophages. Unfortunately, the weak signal produced by CL generated under the various ONOO − cell stimulation conditions, meant that co-localization studies with LysoTracker Green were inconclusive. Indicating that more work will be required to confirm the predicted sub-cellular localization of the CL probe.

Fluorescent Characterization of the Probes
The coumarin-based probes were analyzed using a BMG Labtech CLARIOstar plate reader where compounds were added to a Griener Bio-one 96-well, black-walled microplate with fbottom chimney wells (ThermoFisher). Fluorescent readings were collected using BMG Labtech MARS software. HPLC or fluorescent grade solvents and de-ionized water were used in the measurements. was combined and stirred with aq. NaOCl for 2 min. ROO· was prepared from 2, 2 ′ -azobis (2-amidinopropane) dihydrochloride. AAPH (2,2 ′azobis (2-amidinopropane) dihydrochloride, 10 M) was added and stirred in de-ionized water at 37 • C for 30 min. O ·− 2 was formed from KO 2 (1 eq) and 18-crown-6 (2.5 eq) dissolved in DMSO. HO· was generated via a Fenton reaction: ferrous chloride (1 M) was combined with 10 eq of H 2 O 2 (37.0 wt%). ROS/RNS-sensitivity of each probe were determined by titration at 25 • C in PBS buffer (pH 7.4). Each probe was studied at 5 µM +with different concentrations of ROS/RNS as indicated in the Figure. In vitro Characterization in RAW264.7 Macrophages Cell culture Murine RAW264.7 macrophages were kindly donated by Prof. Masaru Ishii from the Graduate School of Medicine, Osaka University. The cells were grown in complete medium containing high-glucose Dulbecco's modified Eagle medium (DMEM) plus GlutaMAX-I supplemented (ThermoFisher, 31966021) with 10% fetal bovine serum (FBS, ThermoFisher, 16000044), 100 U/mL penicillin and 100 µg/mL streptomycin (ThermoFisher, 15140122) at in a humidified atmosphere with 5% CO 2 . During cell passaging (every 2 days), media was removed, washed with PBS (2 x 4 ml), trypsin (4 ml, ThermoFisher, 15400054) was added and incubated for 3 min at 37 • C, 5% CO 2 . The solution was mixed and 1 ml of cell solution transferred to a new culture dish containing 10 ml of complete medium. Prior to the experiment, cells were plated in a glass bottom dish (35 mm Iwaki, I.C.T., S. L., 3930-035) at a cell density of 5 × 10 5 cells and cultured in 1 ml DMEM medium (without penicillin or streptomycin) for 18 h in a humidified environment at 37 • C in 5% CO 2 . The cells were then treated as described in the following sections. Where indicated, the subsequent treatments were performed in phenol red-free DMEM (ThermoFisher, 21041025) with no added FBS, penicillin or streptomycin (described as "phenol redfree DMEM").
Investigating CM, CE, and CL with SIN-1 The culture media was removed, cells were washed twice with Hanks' Balanced Salt Solution (HBSS) (2 × 2 ml) followed by the addition of 100 µM SIN-1 (Cayman Chemical, 82220) in phenol red-free DMEM and incubated for 30 min or 1 h at 37 • C in 5% CO 2 , as indicated. The SIN-1 containing DMEM culture media was removed, the cells were washed twice with HBSS (2 × 2 ml), then the probe (CM, CL, or CE) (20 µM) in phenol redfree DMEM was added and incubated for 30 min at 37 • C in 5% CO 2 . The cells were then immediately analyzed using fluorescent confocal microscopy.
Investigating CL With Phorbol 12-myristate 13-acetate (PMA) The complete DMEM culture media was removed, cells were washed twice with HBSS (2 × 2 ml), PMA (1 µg/ml, Cayman Chemical, 10008014) in phenol red-free DMEM was added and incubated for 1 h at 37 • C in 5% CO 2 . The PMA-containing phenol red-free DMEM was removed, cells were washed twice with HBSS (2 × 2 ml), CL (20 µM) in phenol red-free DMEM was added and incubated for 30 min at 37 • C in 5% CO 2 . The cells were then immediately analyzed using fluorescent confocal microscopy.

Fluorescent Confocal Microscopy
Fluorescence microscopy images were captured on an Olympus, FLUOVIEW FV10i confocal microscope using 35 mm glass base dishes. Images were captured at a magnification of x60 with the following parameters: λ ex = 405 nm, λ em = 420-460 nm. Laser 405 nm with an intensity of 50% was used. Cells were imaged at 37 • C. Processing and analysis of confocal microscopy images were performed with Image J (https://imagej.nih.gov/ij/).

DATA AVAILABILITY STATEMENT
All data supporting this study are provided as supplementary information accompanying this paper.

AUTHOR CONTRIBUTIONS
MW designed and carried out the synthesis, fluorescence and cell studies. NY assisted in the cell studies. XT helped with the fluorescence analysis. MM and KK provided advice on cell studies and provided facilities. SB, AM, and TJ offered guidance on the project. The manuscript was written by MW with support from TJ and the final version was edited and approved by all the contributing authors.