Mitochondrially-Targeted Ratiometric Redox Probes

  • Amandeep KaurEmail author
Part of the Springer Theses book series (Springer Theses)


The oxidative capacity and levels of ROS production throughout the cell are by no means homogeneous, and mitochondria are the cardinal players in cellular redox homoeostasis and signalling.

The oxidative capacity and levels of ROS production throughout the cell are by no means homogeneous, and mitochondria are the cardinal players in cellular redox homoeostasis and signalling. Mitochondrial ROS levels are known to be key to the function of the organelle, particularly in redox signalling processes, which have a variety of physiological roles, including the maintenance of mitochondrial morphology [1, 2], stem cell differentiation [3] and cardiac remodelling [4]. On the other hand, mitochondrial oxidative stress is implicated in diseases associated with ageing [5, 6].

Despite the vast interest in elucidating the role of mitochondrial redox state in cellular signalling and disease, there is a paucity of tools that can report on the ROS levels within the mitochondria. In particular, there is a lack of tools that can reversibly monitor redox changes over time, providing the potential to sense oxidation-reduction fluxes and to distinguish between transient oxidative bursts and chronic oxidative stress. Although the flavin-based redox probe NpFR2, discussed in Chap.  2, demonstrated excellent mitochondrial accumulation, a new generation redox probe that exhibits ratiometric fluorescence properties in addition to mitochondrial localisation would be a valuable addition to the limited ratiometric and reversible mitochondrial redox probes developed to date.

As discussed in Chap.  3, FRET is an elegant strategy to develop ratiometric probes and this was successfully implemented in the case of FCR1, a flavin-based ratiometric redox probe with cytoplasmic localisation. This chapter details the work performed towards employing the FRET strategy for the development of a mitochondrially-targeted ratiometric probe with reversible redox sensing abilities. The sub-cellular localisation of the developed probes was interrogated using confocal microscopy and subsequent biological experiments were undertaken to evaluate the reversible redox sensing properties of the probes. Aspects of the work discussed in this chapter have been published in Antioxidants and Redox Signalling [7].

4.1 Excitation-Ratiometric Redox Probes

Considering the inherent redox responsive abilities of flavin, it was again selected as the redox-active moiety. As outlined in Sect.  3.2, in order to develop a FRET-based ratiometric probe, it is essential to identify a good donor-acceptor FRET pair, such as the coumarin-flavin pair in FCR1. Mitochondrial localisation can then be accomplished either by modifying the scaffold to incorporate a mitochondrial tag, or by employing a different fluorophore with inherent mitochondrial-localising ability (Fig. 4.1), which must also be capable of establishing a good FRET-pair with flavin. Fluorophores with delocalised cationic nature, such as rhodamine and cyanine derivatives, commonly accumulate in the mitochondria. This accumulation is dependent on the negative potential across the mitochondrial membrane [8].
Fig. 4.1

Two approaches towards the design of a mitochondrially-targeted flavin-based ratiometric redox probe—tethering flavin to a fluorophore a that is attached to a mitochondrial tag and b with inherent mitochondrially-localising ability

Rhodamines are a class of fluorophores that satisfy both these conditions. The spectral properties of rhodamine indicated that it was ideal to act as a FRET-acceptor from flavin (Fig. 4.2). With the redox-active component fixed as the FRET donor and rhodamine as the FRET acceptor, the strategy was to develop an excitation-ratiometric redox probe. Amongst the FRET-based ratiometric probes reported to date for a myriad of sensing purposes, emission ratiometric probes dominate the literature, with only a limited number of probes that are excitation-ratiometric [9, 10]. Development of a redox probe with excitation-ratiometric sensing properties would be a valuable addition to the list of probes that have employed this strategy.
Fig. 4.2

Absorbance (dashed line) and emission (solid line) spectra of flavin (black) (10 \( \mu \)M) and rhodamine (red) (10 \( \mu \)M in 100 mM HEPES buffer, pH 7.4) indicating a significant overlap (yellow) of the emission profile of the flavin moiety with the absorbance of the rhodamine

4.2 Flavin-Rhodamine FRET Probe

The rationale behind utilising the excitation-ratiometric flavin-rhodamine FRET pair as a reporter of redox state is that in the oxidised form, there would be a greater extent of spectral overlap between the flavin and rhodamine scaffold, and excitation of the flavin would result in a red rhodamine acceptor emission by FRET. In the reduced form, owing to the formation of the colourless and non-fluorescent bent conformation of flavin, the spectral overlap between flavin and rhodamine is minimised, FRET is suppressed, resulting in negligible emission from the rhodamine acceptor (Fig. 4.3).
Fig. 4.3

The FRET process taking place within an excitation-ratiometric flavin-rhodamine probe in the oxidised and reduced forms

Furthermore, excitation at a different (longer) wavelength of the rhodamine itself for both the oxidised and reduced forms of the flavin would result in red emission, independent of the FRET process. Therefore, a ratio of the emission intensities of rhodamine resulting from excitation at two different wavelengths could be used to gauge the redox state. On the basis of the predicted spectral overlap between flavin and rhodamine, two flavin-rhodamine redox probes FRR1 and FRR2, were designed (Fig. 4.4).
Fig. 4.4

The chemical structures of FRR1 and FRR2

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

Flavin-rhodamine redox probe 1 (FRR1) contains N-ethylflavin tethered to rhodamine B via piperazine, a relatively rigid spacer. N-Ethylflavin is the redox sensing group used in FCR1. In this study, the use of a naturally-existing flavin derivative, tetraacetylriboflavin, was also investigated. Tetraacetylriboflavin, with an acetylated-ribose tail at N-10 and methyl groups at positions 7 and 8 on the isoalloxazine ring, has been reported to possess a higher quantum yield and significantly red-shifted excitation and emission wavelengths compared to N-ethylflavin [11]. Therefore, the influence of these structural variations on the photophysical and redox properties of the probes was investigated. Moreover, it was envisioned that the lipophilic tetraacetylribose group could potentially improve the cellular retention and localisation of the probe. Discussions regarding probe design with Dr. Karolina Jankowska, a postdoctoral researcher in the group, led to the determination of the synthetic approach towards the flavin-rhodamine FRET probes (Scheme 4.1). Synthesis was then carried out by Dr. Jankowska.

Scheme 4.1

Synthesis of a FRR1 and b FRR2

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

4.3 Spectral Characterisation of FRR Probes

Considering the spectral overlap of flavin and rhodamine scaffolds (Fig. 4.2), excitation wavelengths of 460 and 530 nm were chosen to ensure maximum excitation of flavin and rhodamine respectively. Similar to the redox probes investigated thus far, the photophysics of FRR1 and FRR2 were studied in HEPES buffer (100 mM, pH 7.4). These initial studies were performed by Dr. Karolina Jankowska using an excitation wavelength of 460 nm because at this wavelength the flavin moiety exhibits high absorbance whereas the rhodamine scaffold has negligible absorbance, thus allowing for preferential excitation of the flavin component over rhodamine in FRR1 and FRR2. These studies indicated that upon excitation at 460 nm, FRR1 and FRR2 exhibit two emission maxima: the flavin maximum at 510 nm (FRR1) and 525 nm (FRR2) respectively; and the rhodamine maximum observed at 580 nm for both the probes (Fig. 4.5). Furthermore, the rhodamine could be independently excited at 530 nm.
Fig. 4.5

Absorbance (dotted line) and emission (solid line, \( \lambda _{ex} =\) 460 nm) spectra of (a) FRR1 and (b) FRR2 (10 \( \mu \)M in 100 mM HEPES buffer, pH 7.4). Spectra obtained by Dr. Jankowska

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

FRR1 and FRR2 were examined for their response to reduction by sodium dithionite. As envisioned, this caused reduction of the flavin molecule to its non-emissive configuration, resulting in a suppressed FRET mechanism. Consequently, upon excitation at 460 nm the integrated emission intensity of both the probes decreased. The ratio of the peak intensity of both FRR probes at 580 nm upon excitation at 530 and 460 nm increased upon reduction (Fig. 4.6). Re-oxidation of the reduced probes by treatment with H\( _{2} \)O\( _{2} \) could be achieved in 20 min, and resulted in the restoration of the original fluorescence properties.
Fig. 4.6

Fluorescence emission spectra of a, b FRR1 (10 \( \mu \)M) and d, e FRR2 (10 \( \mu \)M) in oxidised (black) and reduced (dashed line) forms upon excitation at 460 nm (a, d) and 530 nm (b, e). Probes were reduced using 200 equivalents of Na\( _{2} \)S\( _{2} \)O\( _{4} \). The ratio of the emission of FRR1 (c) and FRR2 (f) at 580 nm upon excitation at 530 versus 460 nm in oxidised (black) and reduced (grey) forms. All data were acquired in 100 mM HEPES buffer, pH 7.4. Error bars represent standard deviation (n \(=\) 3). Spectra were obtained by Dr. Jankowska

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

More rigorous characterisation of the redox responsive and spectral characterisation of FRR1 and FRR2 were performed by Dr. Jankowska. These studies indicated that FRR1 exhibited a 3-fold increase in the I\( _{530 ex} \)/I\( _{460 ex} \) ratio upon reduction and the ratiometric response of the probe towards reduction and re-oxidation remained unaltered for up to 5 cycles. FRR2 displayed a more pronounced 7-fold change in the I\( _{530 ex} \)/I\( _{460 ex} \) ratio upon sequential reduction and re-oxidation events, with its ratiometric response remaining unchanged for up to 7 cycles. These results demonstrated the potential of FRR1 and FRR2 to function as reversible sensors for redox state. Just as in the case of other redox probes discussed in Chaps.  2 and  3, control experiments were performed on FRR1 and FRR2, which showed that the probes undergo re-oxidation with biological ROS/RNS within 30 min, and the I\( _{530 ex} \)/I\( _{460 ex} \) ratio remained unaltered in the presence of biologically relevant metal ions, and at pH values between 3 and 8.

While, rhodamine based probes are widely used for imaging purposes due to their high fluorescence quantum yields, at high concentrations, the fluorescence intensity decreases considerably, and this concentration based quenching is attributed to the formation of dimers [12]. Therefore, to determine the concentrations of the probes that would be ideal for imaging purposes, a concentration-based fluorescence assay was performed by recording the fluorescence spectra of the probes in HEPES buffer (100 mM, pH 7.4) with incremental probe concentrations (Fig. 4.7).
Fig. 4.7

Fluorescence of FRR1 (a) and FRR2 (b) with varying concentration. Bars represent the integrated fluorescence intensity (\( \lambda _{ex} \) \(=\) 460 nm, \( \lambda _{em}\) \(=\) 480 − 700 nm). All data were acquired in 100 mM HEPES buffer, pH 7.4. Errors bars represent standard deviation (n \(=\) 3)

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

The integrated fluorescence intensities of both FRR1 and FRR2 followed a linear trend for concentrations ranging from 5–80 \( \mu \)M. Concentration-based quenching does not occur below 80 \( \mu \)M, therefore probe concentrations below this value should be suitable for biological imaging, provided there are no cytotoxic effects.

4.4 Electrochemistry

The reduction potentials of FRR1 and FRR2 were investigated by recording cyclic voltammograms with 5 mM concentrations of the probes in MeCN containing TBAB (tetrabutylammonium bromide) as an electrolyte. The cyclic voltammograms depicted two sets of peaks for each probe (Fig. 4.8). The peaks at the lower half-wave potential (E\( ^{\circ } \)) of −544 mV correlates to the value reported for the reduction potential of rhodamine B [13]. The peaks at −186 and −369 mV (E\( ^{\circ } = -\)259 mV vs. SHE) in FRR1 correspond to the reduction potential of the N-ethylflavin component, whereas the peaks at −173 and −413 mv (E\( ^{\circ } = -\)290 mV vs. SHE) observed for FRR2 represents the reduction potential of tetraacetylriboflavin (Fig. 4.9)
Fig. 4.8

Cyclic voltammogram of a FRR1 and b FRR2 (5 mM concentration) in MeCN at 25\(^\circ \)C with a scan rate of 100 mV/s

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

Fig. 4.9

Photophysical behaviour of FRR2 under 488 nm excitation. a Spectra of FRR2 untreated (black), reduced with 2 mM sodium dithionite (red) and re-oxidised with 4 mM H\( _{2} \)O\( _{2} \) (blue). b Fluorescence emission ratio of emission intensity upon excitation with 530 and 488 nm. c Integrated intensity of red fluorescence (\(\lambda _{ex} =\) 488 nm, \(\lambda _{em} =\) 560 − 700 nm) All spectra were acquired at 10 \( \mu \)M concentration of probe in 100 mM HEPES buffer, pH 7.4

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

These electrochemical studies demonstrate that varying the substituent at the N-10 position on the isoalloxazine ring modulates the reduction potential of the flavin. In this case, it was observed that the ribose tail in FRR2 tuned the reduction potential to a more biologically relevant value compared to FRR1. Although the cyclic voltammograms indicate a peak corresponding to the reduction of rhodamine, this reduction potential lies far outside the scope of biological values. The profile of the obtained cyclic voltammograms confirm the chemical reversibility of the oxidation and reduction processes of both FRR1 and FRR2.

4.4.1 Fluorescence Properties at 488 nm Excitation

While the ratiometric response of FRR1 and FRR2 towards redox state has been extensively investigated, the use of these probes depends upon the availability of two appropriate excitation sources, which many standard microscopy and cytometry instruments lack. In order to illustrate the broader applicability of both probes in such protocols, the redox-responsive ability of the probes under a single excitation wavelength was investigated. Upon excitation at 488 nm, FRR1 did not exhibit any fluorescence emission from the flavin and only minimal fluorescence from the rhodamine. Therefore, upon excitation at 488 nm no significant changes were observed in the fluorescence properties of FRR1 upon reduction.

Although in the case of FRR2 (Fig. 4.9a) the peak corresponding to the fluorescence from flavin was greatly diminished, a similar trend in ratiometric response to reduction was observed upon excitation at 488 nm (I\( _{530 ex} \)/I\( _{488 ex} \)) as for 460 nm (I\( _{530 ex} \)/I\( _{460 ex} \)). The decrease in the (I\( _{530 ex} \)/I\( _{488 ex} \)) ratio upon reduction could be reversed upon re-oxidation with H\( _{2} \)O\( _{2} \) (Fig. 4.9b). Notably, the absolute fluorescence intensity (560–590 nm) of FRR2 decreases upon reduction, and this decrease can be reversed by treatment with H\( _{2} \)O\( _{2} \) (Fig. 4.9c). Therefore, FRR2 demonstrates the ability to report on variations in the redox environment, when excited using a single wavelength excitation of 488 nm. Hereafter, microscopy and flow cytometry experiments were performed using a 488 nm excitation wavelength, with a more oxidised probe indicated by a more intense red fluorescence.

4.5 Biological Imaging Experiments

Having established the redox-active behaviour and reversibility of FRR1 and FRR2, the probes were investigated for their cellular localisation, toxicity and ability to report on biological oxidative capacity.

4.5.1 Sub-cellular Localisation

The sub-cellular localisation of the probes was examined in RAW 264.7 macrophages. Using a confocal microscope, cells were excited by a 488 nm laser and the fluorescence emission was recorded from 495–620 nm. Under 488 nm excitation, negligible fluorescence was observed from control cells untreated with the probe. As depicted in Fig. 4.10, significantly higher fluorescence intensity was observed when cells were treated with FRR1 or FRR2 (20 \( \mu \)M, 15 min). Furthermore, the sub-cellular accumulation pattern of both the probes strongly suggested mitochondrial localisation. Therefore, just as with NpFR2, as discussed in Chap.  2, colocalisation experiments were performed employing commercial tracker dyes—Mitotracker DeepRed FM and Lysotracker DeepRed.
Fig. 4.10

Confocal microscopy images of RAW 264.7 cells, untreated, cells treated only with FRR1 (20 \( \mu \)M, 15 min), FRR2 (20 \( \mu \)M, 15 min), Mitotracker Deep Red (100 nM, 15 min) and Lysotracker Deep Red (100 nM, 15 min), in channel 1 (\( \lambda _{ex} =\) 488 nm, \( \lambda _{em} = 495 -\) 600 nm), channel 2 (100 nM, \( \lambda _{ex} =\) 633 nm, \( \lambda _{em} =\) 650 − 750 nm) and merged images of channel 1 and 2

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

Singly stained controls, prepared by treating cells with FRR1, FRR2, Mitotracker DeepRed FM and Lysotracker DeepRed individually, were imaged to ensure that no emission leaked from one channel into the other. As shown in Fig. 4.10, significant fluorescence was observed from cells treated with FRR1 or FRR2 (20 \(\upmu \)M, 15 min) in channel 1 (\(\lambda _{ex}\) 488 nm, \(\lambda _{em}\) 495 − 620 nm), whilst negligible fluorescence observed in channel 2 (\(\lambda _{ex}\) 633 nm, \(\lambda _{em}\) 650 − 750 nm). Cells treated with the tracker dyes fluoresced only in channel 2, thus validating that there is no fluorescence bleed-through from one channel to another.

Cells were then co-stained by treating them with FRR1 or FRR2 (20 \(\upmu \)M, 15 min) and Mitotracker DeepRed FM (100 nM, 15 min). The cells were interrogated for their fluorescence properties in channels 1 and 2, and pseudo-coloured green and red respectively. The fluorescence images from both the channels were then merged using FIJI (National Institutes of Health), an image processing software. As indicated by the yellow regions in the merged image in Fig. 4.11, fluorescence emission of both FRR1 and FRR2 overlaps significantly with the commercial Mitotracker DeepRed FM, confirming clear mitochondrial accumulation of the probes. RAW 264.7 cells co-stained with the probes and the Lysotracker DeepRed (100 nM, 15 min), illustrate very different localisation profiles (Fig. 4.11), suggesting very low association with lysosomes. This was further evidenced by the poor Pearson’s co-localisation coefficients calculated to be 0.25 and 0.13 for FRR1 and FRR2, respectively.
Fig. 4.11

Co-localization images of macrophages (RAW 264.7) treated with FRR1 (20 \( \upmu \)M) or FRR2 (20 \( \upmu \)M), co-stained with Mitotracker deep red (100 nM) and Lysotracker deep red (100 nM). FRR1/FRR2 emission is in channel 1 (\(\lambda _{ex} =\) 488 nm, \(\lambda _{em} =\) 495 − 620 nm) and Mitotracker/Lysotracker emission in channel 2 (\(\lambda _{ex} =\) 633 nm, \(\lambda _{em} =\) 650 − 750 nm). Merged images indicate good co-localisation of Mitotracker with both probes

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

The mitochondrial localisation of FRR1 and FRR2 is further evidenced by the plot profiles (Fig. 4.12), in which a line selection is made across the image and the intensities of fluorescence in each of the channels are then plotted against the distance. A complete overlap of the grey values (fluorescence intensity) from both the channels would indicate the presence of both probe and the tracker dyes at each pixel, suggesting a perfect colocalisation. The plot profile of FRR1 indicated the presence of some peaks (marked with an asterisk, Fig. 4.12c) that did not overlap with the Mitotracker. This indicates that there are some non-mitochondrial regions in the cell where FRR1 localises. However, the plot profile of FRR2 exhibits great extent of overlap with that of Mitotracker (Fig. 4.12). The Pearson’s co-localisation coefficients were determined to be 0.68 and 0.92 for FRR1 and FRR2 respectively which is in agreement with the results obtained from the plot profiles for both the probes.
Fig. 4.12

Plot profiles of FRR1 (a) and FRR2 (d) in comparison to Mitotracker Deep Red (b, e). The black lines in the plot profiles (c, f) represent Mitotracker Deep Red and those for FRR1 (c) and FRR2 (f) are shown in red. Profiles have been generated from the regions marked with a white line in the images. ⁎indicates regions where there is FRR1 fluorescence but no MitoTracker emission

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

In the interests of further validating the mitochondrial localisation of the probes, co-localisation experiments were carried out with the commercially available CellLight® Mitochondria-GFP, BacMam 2.0. This is a cell-transfection based technology in which mithochondrial targeting is achieved by E1 alpha pyruvate dehydrogenase (a mitochondrial enzyme) and visualisation by a fluorescent protein. A fused DNA construct for E1 alpha pyruvate dehydrogenase and emerald green fluorescent protein (emGFP) is packaged within a Baculovirus and transfection of cells with these virus particles result in the production of a fluorescent mitochondrial enzyme. This enables the detection of mitochondria independent of mitochondrial membrane potential. Since this technology cannot be applied to macrophages, DLD-1 human colorectal cancer carcinoma cells were investigated. DLD-1 cells were transfected with CellLight® Mitochondria-GFP, BacMam 2.0 overnight at a concentration of 10 particles per cell (PPC) followed by treatment with FRR1 or FRR2 (20 \( \upmu \)M, 15 min). Excellent mitochondrial colocalisation was observed for both FRR1 and FRR2, as indicated by the yellow regions in the merged image (Fig. 4.13).
Fig. 4.13

Representative confocal microscopy images of DLD-1 cells, cells transfected with CellLight® mitochondria-GFP, BacMam 2.0, co-stained with FRR1 (20 \( \upmu \)M, 15 min), FRR2 (20 \( \upmu \)M, 15 min), in channel 1 (\(\lambda _{ex} =\) 488 nm, \(\lambda _{em} =\) 510 − 540 nm), channel 2 (\(\lambda _{ex}\) = 488 nm, \(\lambda _{em} =\) 560 − 700 nm) and merged images of channel 1 and 2. Scale bars represent 10 \( \upmu \)m

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

To ensure that the probes did not have any toxic effects, cytotoxicity of the probes was evaluated using the standard MTT assay. This was accomplished by treating RAW 264.7 murine macrophage cells with FRR1 and FRR2 at 0–160 \( \upmu \)M concentrations and incubated for a period of 24 h followed by 4 h of treatment with MTT. The IC\( _{50} \) values were determined to be 38 (\( \pm \)1 \( \upmu \)M) and 41 (\( \pm \)2 \( \upmu \)M) for FRR1 and FRR2 respectively. These values are much higher both in terms of concentration and treatment times employed in cellular imaging studies. In addition, time-lapse imaging experiments were carried out to examine the consequences of longer laser exposure on the photophysical behaviour of the probe and cell viability. Following laser irradiation at 488 nm every 30 s, cell morphology and fluorescence emission from RAW 264.7 cells treated with FRR1 or FRR2 (20 \( \upmu \)M, 15 min) were monitored. The results indicated that FRR1 and FRR2 did not exhibit any self-amplication of the fluorescence signal, over 25 min (Fig. 4.14).
Fig. 4.14

Cell morphology and fluorescence intensity of RAW 264.7 macrophages treated with FRR1 or FRR2 (20 \( \upmu \)M) and imaged every 30 s with laser excitation at 488 nm, and an acquisition speed of 400 Hz. a Transmitted light and confocal (\(\lambda _{ex} =\) 488 nm, \(\lambda _{em} =\) 560 − 700 nm) images of cells at 0 and 25 min time-points. b Emission intensities obtained at indicated time-points for cells treated with FRR1 (grey) or FRR2 (black). Bars represent average fluorescence intensity (\(\lambda _{ex} =\) 488 nm, \(\lambda _{em} =\) 560 − 700 nm), errors bars represent standard deviation (n \(=\) 3)

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

Moreover, no significant changes in cell morphology were observed over this time, demonstrating a lack of phototoxic effects. Consistent with the the brighter-green fluorescence from the riboflavin moiety in FRR2 compared to the synthetic N-ethylflavin in FRR1 aided in confocal imaging and image analysis. Therefore, all further biological experiments were carried out with FRR2.

4.5.2 Measuring Redox Changes in LPS-Stimulated Macrophages

The reversibility of redox responsive FRR2 was harnessed to monitor changes in oxidative capacity of macrophages following a lipopolysaccharide (LPS) insult. LPS, a major outer membrane component of gram negative bacteria, binds to toll-like receptor 4 (TLR4), present on the macrophage membrane, responsible for pathogen recognition [14]. This binding stimulates a cascade of events within the macrophage cell which execute the bactericidal function. An important participant in these events are the mitochondria within the macrophages, which are known to produce a burst of ROS/RNS which subsequently kills the bacteria [14]. Therefore, this was considered an interesting biological system to test the reversible redox-active properties of FRR2.
Fig. 4.15

a Response of FRR2 to macrophages (RAW 264.7) stimulated with LPS from 0–6 h (black) and those re-stimulated with LPS after the 2 h time-point (grey). Bars represent the mean fluorescence intensity of red emission (585/42 nm) when excited with a 488 nm laser. Error bars represent standard error of mean, p < 0.05 is considered significant. b Representative flow cytometry dataset. The histograms represent intensity of red fluorescence emissions (585/42 nm) when excited with a 488 nm laser.

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

RAW 264.7 cells were treated with FRR2 (20 \( \upmu \)M, 15 min) followed by stimulation with LPS at a concentration of 1 \( \upmu \)g/mL, over treatment times ranging from 0–2 h. The cells were then washed and resuspended in FACS buffer (PBS supplemented with 1% FCS) containing propidium iodide (1 \( \upmu \)g/mL). The cells were then interrogated for their red fluorescence emission using the 585/42 nm (42 nm band width, centred at 585 nm) filter flow cytometer equipped with a 488 nm laser. For each sample, a population of single and live cells was gated out and the red fluorescence intensity was analysed. As shown in Fig. 4.15, a 5-fold increase in fluorescence intensity was observed by the 30 min treatment timepoint, after which the intensity decreases, demonstrating the burst in ROS production takes place 30 min after the LPS stimulation. By 2 h after LPS stimulation, the fluorescence emission returned to levels observed for unstimulated cells. Similar emission levels were observed at 4 and 6 h after LPS treatment.
Fig. 4.16

Confocal microscope images of RAW 264.7 macrophages cells showing the fluorescence intensity (\(\lambda _{ex} =\) 488 nm, \(\lambda _{em} =\) 560 − 700 nm) upon stimulation with LPS for 0 h (a), 0.5 h (b) and 1 h (c) followed by treatment with FRR2 (20 \( \upmu \)M, 15 min). Scale bars represent 100 \( \upmu \)m. d Bars represent the mean fluorescence intensity of images acquired from macrophages stimulated with LPS for indicated times. Error bars represent standard deviation, \( ^{*} \)p < 0.05, \( ^{**} \)p < 0.01

Reprinted with permission from Antioxidants and Redox Signalling, Volume 24, Issue 13, published by Mary Ann Leibert, Inc., New Rochelle, NY

Following this, a re-stimulation experiment was performed, wherein the macrophages were re-stimulated with LPS 2 h after the initial stimulation. Although there was a significant increase in fluorescence intensity after 0.5 and 1 h, this increase was less drastic than the initial response. This experiment demonstrate the ability of FRR2 to report on the changes in the mitochondrial oxidative capacity of RAW 264.7 cells following an LPS-insult, clearly illustrate the potential of FRR2 as a reversible sensor of redox fluxes within the mitochondria.

This result was further confirmed by live cell imaging experiments. RAW 264.7 macrophages were stimulated with LPS for 0.5 and 1 h followed by treatment with FRR2 (Fig. 4.16). Similar to the results obtained from flow cytometry, a significant increase in intensity was observed 30 min post-stimulation, followed by a decrease thereafter, validating that a maximal burst in the mitochondrial ROS production occurs 30 min after the LPS-insult. The obtained results establish the potential of FRR2 to reversibly report on the variations in mitochondrial redox state.

4.6 Conclusions

Although excitation-ratiometric probes are less commonly designed and developed because of the complications that arise from the simultaneous use of two different excitation wavelengths, they are certainly more beneficial than intensity-based probes. With their excitation-ratiometric output, FRR1 and FRR2 show great potential as fluorescent tools to interrogate fluxes in oxidative capacity within the mitochondria. Mitochondrial targeting of FRET-based probes can be achieved by taking advantage of fluorophores with intrinsic mitochondrial localising abilities. This removes the requirement of attaching a mitochondrial targeting tag, mitigating synthetic complexity. This chapter describes the studies that investigated the properties of two flavin-rhodamine redox probes FRR1 and FRR2, which differ by the substitution at the N-10 position on the flavin scaffold. FRR1 contains a synthetic N-ethylflavin, whilst FRR2 incorporates a naturally existing riboflavin. The electrochemical studies confirmed that the ribose group at the N-10 position on the flavin scaffold of FRR2 was capable of modulating the reduction potential of the probe to a more biologically-relevant range—this aspect of probe design could be applicable towards the development of redox probes with different redox-active potentials.

In addition, co-localisation experiments indicated clear mitochondrial localisation of the probes FRR1 and FRR2. The fluorescence properties indicated a brighter-flavin emission of FRR2 than FRR1, aiding its imaging using a confocal microscope and image analysis thereafter. Furthermore, it is important to ensure that the photophysical properties of a developed probe are not too exotic to limit its application to limited state of the art instrumental set-ups. The broader applicability of probes for much simpler protocols may also prove beneficial. Therefore, the suitability of the probe to measure redox changes using a basic flow cytometer was examined. The results from the LPS stimulation experiment validate it to be a valuable system for the characterisation of a reversible redox probe, particularly for the mitochondria, such as FRR2. The probe demonstrated that a burst in the production of mitochondrial ROS in macrophages occurs following LPS stimulation and that this burst is the greatest 30 min post LPS stimulation. In addition, more rigorous investigations using FRR2 to report on the variations in mitochondrial oxidative capacity in specific biological systems are discussed in Chaps.  6 and  7.


  1. 1.
    P.H.G.M. Willems, R. Rossignol, C.E.J. Dieteren, M.P. Murphy, W.J.H. Koopman, Redox homeostasis and mitochondrial dynamics. Cell Metab. 22, 207–18 (2015)CrossRefGoogle Scholar
  2. 2.
    M.P. Murphy, A. Holmgren, N.-G. Larsson, B. Halliwell, C.J. Chang, B. Kalyanaraman, S.G. Rhee, P.J. Thornalley, L. Partridge, D. Gems, T. Nystrom, V. Belousov, P.T. Schumacker, C.C. Winterbourn, Unraveling the biological roles of reactive oxygen species. Cell Metab. 13, 361–366 (2011)CrossRefGoogle Scholar
  3. 3.
    C.L. Bigarella, R. Liang, S. Ghaffari, Stem cells and the impact of ROS signaling. Development 141, 4206–4218 (2014). (Cambridge, England)CrossRefGoogle Scholar
  4. 4.
    C. Wolke, A. Bukowska, A. Goette, U. Lendeckel, Redox control of cardiac remodeling in atrial fibrillation. Biochim. Biophys. Acta 2015, 1555–1565 (1850)Google Scholar
  5. 5.
    A.K. Biala, R. Dhingra, L.A. Kirshenbaum, Mitochondrial dynamics: orchestrating the journey to advanced age. J. Mol. Cell. Cardiol. 83, 37–43 (2015)CrossRefGoogle Scholar
  6. 6.
    L.F.R. Francesca, Bonomini Rita Rezzani, Metabolic Syndrome, Aging and Involvement of Oxidative Stress Google Scholar
  7. 7.
    A. Kaur, K. Jankowska, C. Pilgrim, S.T. Fraser, E.J. New, Studies of hematopoietic cell differentiation with a ratiometric and reversible sensor of mitochondrial reactive oxygen species (Antioxid, Redox Signal, 2016)Google Scholar
  8. 8.
    R.A. Smith, R.C. Hartley, M.P. Murphy, Mitochondria-targeted small molecule therapeutics and probes. Antioxid. Redox Signal. 15, 3021–3038 (2011)CrossRefGoogle Scholar
  9. 9.
    L. Yuan, W. Lin, Z. Cao, J. Wang, B. Chen, Development of FRET-based dual-excitation ratiometric fluorescent pH probes and their photocaged derivatives. Chemistry (Weinheim an der Bergstrasse, Germany) 18, 1247–1255 (2012)Google Scholar
  10. 10.
    L. Yuan, Q.-P. Zuo, FRET-based mitochondria-targetable dual-excitation ratiometric fluorescent probe for monitoring hydrogen sulfide in living cells. Chem.-An Asian J. 9, 1544–1549 (2014)CrossRefGoogle Scholar
  11. 11.
    M. Insińska-Rak, A. Golczak, M. Sikorski, Photochemistry of riboflavin derivatives in methanolic solutions. J. Phys. Chem. A 116, 1199–1207 (2012)CrossRefGoogle Scholar
  12. 12.
    D. Setiawan, A. Kazaryan, M.A. Martoprawiro, M. Filatov, A first principles study of fluorescence quenching in rhodamine B dimers: how can quenching occur in dimeric species? Phys. Chem. Chem. Phys. 12, 11238–11244 (2010)CrossRefGoogle Scholar
  13. 13.
    T. Takizawa, T. Watanabe, K. Honda, Photocatalysis through excitation of adsorbates. 2. A comparative study of Rhodamine B and methylene blue on cadmium sulfide. J. Phys. Chem. 82, 1391–1396 (1978)CrossRefGoogle Scholar
  14. 14.
    A.P. West, I.E. Brodsky, C. Rahner, D.K. Woo, H. Erdjument-Bromage, P. Tempst, M.C. Walsh, Y. Choi, G.S. Shadel, S. Ghosh, TLR signalling augments macrophage bactericidal activity through mitochondrial ROS. Nature 472, 476–480 (2011)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG 2018

Authors and Affiliations

  1. 1.School of ChemistryUniversity of SydneySydneyAustralia

Personalised recommendations