H2DCF-DA (dihydrodichlorofluorescein diacetate) is widely used to evaluate ‘cellular oxidative stress’. After passing through the plasma membrane, this lipophilic and non-fluorescent compound is de-esterified to a hydrophilic alcohol [H2DCF (dihydrodichlorofluorescein)] that may be oxidized to fluorescent DCF (2′,7′-dichlorofluorescein) by a process usually considered to involve ROS (reactive oxygen species). It is, however, not always recognized that, being a hydrophilic molecule, H2DCF does not cross membranes, except for the outer fenestrated mitochondrial ones. It is also not generally realized that oxidation of H2DCF is dependent either on Fenton-type reactions or on unspecific enzymatic oxidation by cytochrome c, for neither superoxide, nor H2O2, directly oxidizes H2DCF. Consequently, oxidation of H2DCF requires the presence of either cytochrome c or of both redox-active transition metals and H2O2. Redox-active metals exist mainly within lysosomes, whereas cytochrome c resides bound to the outer side of the inner mitochondrial membrane. Following exposure to H2DCF-DA, weak mitochondrial fluorescence was found in both the oxidation-resistant ARPE-19 cells and the much more sensitive J774 cells. This fluorescence was only marginally enhanced following short exposure to H2O2, showing that by itself it is unable to oxidize H2DCF. Cells that were either exposed to the lysosomotropic detergent MSDH (O-methylserine dodecylamide hydrochloride), exposed to prolonged oxidative stress, or spontaneously apoptotic showed lysosomal permeabilization and strong DCF-induced fluorescence. The results suggest that DCF-dependent fluorescence largely reflects relocation to the cytosol of lysosomal iron and/or mitochondrial cytochrome c.
The molecular mechanisms behind AMD (age-related macular degeneration) are much debated but far from understood, even though it is the most common cause of visual impairment among the elderly in the Western world [1–3]. However, on the basis of recent evidence, oxidative stress within RPE (retinal pigment epithelial) cells, with the ensuing accumulation of lysosomal lipofuscin and thereby depressed phagocytosis/autophagocytosis, is considered to be a major aetiological factor behind AMD [4–8]. Consequently, in relation to various experimental studies on cultures of RPE cells, the extent of cellular oxidative stress is often being evaluated.
A method that is frequently relied upon for this purpose is the DCF (2′,7′-dichlorofluorescein) test [9–11]. H2DCF-DA (dihydrodichlorofluorescein diacetate) is added to cells in culture and the intracellular oxidation of H2DCF (dihydrodichlorofluorescein) to DCF is documented over time. H2DCF-DA is a non-fluorescent lipophilic ester that easily crosses the plasma membrane and passes into the cytosol, where it is rapidly cleaved by unspecific esterases . One of the reaction products is the non-fluorescent alcohol H2DCF. The oxidation of this molecule to the fluorochrome DCF results in green fluorescence when excited with blue light. The brightness of this fluorescence is usually considered to reflect the extent to which ROS (reactive oxygen species) are present, unfortunately without any further definition of what kind of ROS may give rise to the oxidation [13,14]. Furthermore, it is often wrongly assumed that H2DCF is evenly distributed in the cell following the intracellular cleavage of the added diacetate ester, neglecting the inability of hydrophilic molecules to traverse membranes. Moreover, the evaluation of the DCF test commonly involves plate readers or flow cytofluorimeters, which do not allow any careful morphological analysis of individual cells that, if undertaken, might have disclosed unexpected cellular variations. Thus the DCF test often seems to be somewhat uncritically used.
We have shown previously that oxidation of H2DCF to DCF is not a result of exposure to ROS in general, but rather indicates the specific impact of hydroxyl radicals formed during Fenton-type reactions . In the present study, we also stress that cytochrome c, as has been pointed out previously [16–18], operates as an unspecific peroxidase with H2DCF as a target. It is therefore plausible to suppose that LMP (lysosomal membrane permeabilization), with release to the cytosol of redox-active iron, and/or mitochondrial release of cytochrome c is required for the induction of strong cytosolic DCF-mediated fluorescence. We also point out that, in the absence of LMP and apoptosis/necrosis, H2DCF only occurs in the cytosol and the mitochondrial inter-membranous space, where it gives rise to weak cytosolic and somewhat stronger mitochondrial fluorescence that indicates oxidation. Most probably, this faint fluorescence, which regularly seems to be considered to be background and to be overlooked, results from the normal mitochondrial production of H2O2 that diffuses all over the cell, the presence of cytochrome c in the mitochondrial intermembranous space as well as minute amounts of labile iron under transport in the cytosol. Consequently, if LMP and related mitochondrial damage with relocation of cytochrome c has taken place, a positive DCF test might be interpreted as a sign of oxidative stress, even if that would not necessarily be the case. Similarly, a low or undetectable level of DCF-induced fluorescence may not necessarily indicate the absence of ROS, but rather designate stable lysosomal and mitochondrial compartments.
MATERIALS AND METHODS
DMEM (Dulbecco's modified Eagle's medium), Ham's F12 medium, FBS (fetal bovine serum), penicillin and streptomycin were from Invitrogen. AO (Acridine Orange) base was from Gurr, CP22 (1-propyl-2-methyl-3-hydroxypyrid-4-one) and MSDH (O-methylserine dodecylamide hydrochloride) were gifts from Professor Robert Hider, Pharmaceutical Sciences Division, King's College London, London, U.K., and Dr Gene Dubowchik (Bristol-Myers Squibb Pharmaceutical Research Institute, Wallingford, CT, U.S.A.) respectively. All other reagents were from Sigma.
Cells and culture conditions
ARPE-19 (human immortalized retinal pigment epithelial) cells and murine J774 macrophage-like histiocytic lymphoma cells (both obtained from the A.T.C.C., Manassas, VA, U.S.A.) were grown at 37 °C in humidified air with 5% CO2 in DMEM and Ham's F12 (1:1) supplemented with 10% (v/v) FBS, 2 mM L-glutamine, 100 IU/ml penicillin and 100 μg/ml streptomycin. Cells were subcultivated twice a week, seeded in six-well plates with or without coverslips at a concentration of 5×105 cells/well (J774), while the much larger ARPE-19 cells were seeded at 2×105 cells/well with or without coverslips. Cells were subjected to experiments 12, 24 and 48 h after subcultivation.
Conditions for basic studies with H2DCF-DA only
Small amounts of stock solutions, containing 1–10 mM H2DCF-DA in DMSO, were added to fresh complete growth medium. The carefully blended solution was added to the culture dishes following removal of the old medium. Cells on coverslips were in that way exposed to 3–30 μM H2DCF-DA in medium with 0.03–3% DMSO for 30 min under otherwise standard culture conditions. The cells were then rinsed in HBSS (Hanks balanced salt solution), mounted directly in HBSS, and within 5–6 min were subjected to laser-scanning confocal microscopy using a Nikon Eclipse C1 laser-scanning confocal microscope.
Following tests with various concentrations of H2DCF-DA and DMSO, a stock solution of 10 mM H2DCF-DA in DMSO was selected for both cell types and added to medium in such a way that the final concentrations of H2DCF-DA and DMSO were 10 μM and 0.1% respectively. The cells were observed and photographed using the above confocal microscope, a 488 nm argon laser and the Nikon EZ-C1 V3.70 software for image acquisition. DCF-induced fluorescence was detected using a 515/30 nm band pass filter. Since the DCF-induced fluorescence of control cells is rather weak, the medium-sized pinhole and an electronic gain of 6.5 were applied. All photographs were taken at a resolution of 2 megapixels. To allow comparison between experiments, these settings were then kept the same for all experiments involving evaluation of DCF-induced fluorescence, even when the intensity of fluorescence was much enhanced due to the conditions being applied and, consequently, a small pinhole and less amplification would have given better resolution and prevented overexposure of strongly fluorescent cells.
Localization of mitochondria using the mitochondria-specific dye TMRE (tetramethylrhodamine ethyl ester)
Mitochondria were demonstrated using the cationic and lipophilic dye TMRE, which accumulates in the matrix of normal mitochondria. Cells were incubated with TMRE in complete culture medium (100 nM) for 15 min at 37 °C and were observed using the above-mentioned Nikon confocal microscope. Since mitochondrial TMRE fluorescence is strong, optimal documentation conditions were used (the smallest pinhole and a low gain).
Overlays of DCF/TMRE pictures could not be produced because TMRE is a metachromatic fluorophore that provides both red and green fluorescence when activated by a 488 nm laser and the green one is much stronger than that of DCF in control cells. TMRE fluorescence was detected using a 590/50 nm band pass filter.
Studies aiming at various cytosolic labile iron concentrations before exposure to H2DCF-DA
To evaluate the influence on DCF-induced fluorescence of different concentrations of labile iron in the cytosol, as well as in the mitochondrial intermembranous space, cells were initially exposed to H2DCF-DA as above and then mounted in HBSS with 500 μM FAC (ferric ammonium citrate) to enhance labile iron. Separately, cells were incubated with H2DCF-DA together with 100 μM of the iron chelator CP22 to ligate labile iron and then mounted in HBSS in the continuous presence of CP22. CP22 (molecular mass 203.7 Da) is an effective water-soluble bidentate iron chelator with a complex constant of ~1036 . When complexed with CP22, iron is prevented from redox cycling and is unable to support Fenton-type reactions.
Evaluation of DCF-induced fluorescence in relation to oxidative stress
Exposure to H2O2
In order to induce oxidative stress, ARPE-19 and J774 cells were initially exposed to H2DCF-DA as above and then mounted in HBSS with 500 μM H2O2.
To induce apoptosis of the ARPE-19 (very resistant to oxidative stress) and J774 (sensitive) cells, they were exposed to 15 mM and 100 μM H2O2 in HBSS respectively for 30 min and then kept at standard culture conditions for approx. 6 h when they were incubated with H2DCF-DA and studied as described above. In order to emphasize further the differences in lysosomal membrane stability between cell types, both of them were exposed to 100 μM H2O2 in HBSS for 30 min and then kept under standard culture conditions for approx. 6 h, when they were exposed to the lysosomotropic metachromatic fluorochrome AO as described previously [20–24]. A 488 nm argon laser was used for excitation and AO fluorescence was detected using 515/30 nm (green) and 590/50 nm (red) band pass filters.
Exposure to a lysosomotropic detergent
In order to induce lysosomal rupture by means other than oxidative stress, ARPE-19 cells were initially exposed to H2DCF-DA as above. They were then mounted in HBSS with 100 μM of the lysosomotropic detergent MSDH, which is known to induce lysosomal labilization , before being studied over a 15 min period of time. In separate experiments, designed to verify the LMP effect of MSDH, cells were initially exposed to AO as above and then mounted in HBSS with MSDH and followed over time.
Evaluation of DCF-induced fluorescence in apoptotic ARPE-19 cells
Apoptotic and post-apoptotic necrotic cells are always present in cell cultures, especially shortly after subcultivation, because a number of cells do not survive that procedure. Such apoptotic ARPE-19 cells were identified, 12 h following seeding, on the basis of their characteristic morphology that includes chromatin condensation, nuclear fragmentation, plasma membrane blebbing and formation of apoptotic bodies. DCF-induced fluorescence was evaluated following exposure to H2DCF-DA as described above.
Assessment of the capacity of cytochrome c to oxidize H2DCF
Using a modification  of a technique described by Myhre et al. , H2DCF was exposed to cytochrome c with or without the small iron chelator CP22 (at 1 and 20 μM respectively). Briefly, a standard consisting of ferric iron (10 μM) was reduced to its ferrous form by cysteine (100 μM) in a Hepes buffer (pH 7.0; 150 mM). H2O2 (100 μM) was added to initiate oxidative conversion of non-fluorescent H2DCF (5 μM) into fluorescent DCF [H2DCF was obtained by hydrolysing its diacetate ester (H2DCF-DA)]. DMSO (10%) and CP22 (20 μM) were used to demonstrate the involvement of HO• and iron respectively.
The capacity of cytochrome c to oxidize H2DCF was then compared with that of the Fe(II) standard. Fluorescence was measured in a FL600 microplate fluorescence reader (Bio-Tek) at λex 485 nm and λem 530 nm.
DCF-induced fluorescence intensity was measured using the NIH ImageJ software version 1.42q. Following delineation of each cell, their mean fluorescence was calculated and analysis performed using Student's t test.
Normal cells show a weak cytosolic and a somewhat stronger mitochondrial-type DCF-induced fluorescence pattern, whereas apoptotic cells demonstrate strong diffuse fluorescence
As mentioned in the Materials and methods section, initial experiments indicated that a final concentration of 10 μM H2DCF-DA in complete culture medium, obtained by using a stock solution of 10 mM H2DCF-DA in DMSO, gave acceptable results and allowed reasonable documentation of both normal and apoptotic ARPE-19 and J774 cells using the same confocal settings. It should be pointed out that the fluorescence obtained from normal cells was low (particularly in J774 cells). In studies using flow- or plate-reader fluorometry, it would have been considered non-specific background. The final concentration of DMSO did not influence the results as long as it was below 1%. Normal ARPE-19 and stretched-out J774 cells showed a mitochondrial DCF-induced weak fluorescence in combination with a very faint cytosolic fluorescence (Figures 1A and 2A), whereas apoptotic ARPE-19 and J774 cells presented strong DCF-induced fluorescence that was 6–14-fold higher than that of the control cells (Figures 1C, 2B and 2D) whether the apoptosis was spontaneous (ARPE-19 cells; Figure 1C) or due to exposure to oxidative stress (Figures 2B and 2D). It needs to be emphasized that in contrast with the ARPE-19 cells that adhere firmly to their substratum, the J774 cells tend to round up and detach easily, especially when becoming apoptotic. Therefore the rinsing steps that are needed for the H2DCF-DA exposure make it very difficult to find more than occasional apoptotic cells in the mounted cultures that are studied using the confocal microscope (see the legend to Figure 2).
Exposure to H2DCF-DA induces weak mitochondrial and hardly noticeable cytosolic DCF-induced fluorescence in control ARPE-19 cells that only slightly changes following exposure to H2O2, while spontaneously apoptotic cells show strong cytosolic fluorescence
Normal J774 cells show DCF-induced fluorescence similar to that of the ARPE-19 cells, whereas oxidative-stress-induced apoptotic cells display strong cytosolic fluorescence
The initially weak mitochondrial-type DCF-induced fluorescence pattern from normal cells became slightly stronger after a few seconds of exposure to blue light, which is considered to be an oxidative effect and probably due to the formation of singlet oxygen. Owing to the low intensity of fluorescence, the smallest pinhole and a low electronic gain, normally chosen in order to obtain the sharpest possible confocal microscopy pictures, could not be used. Nevertheless, it was clear that the DCF-mediated fluorescence of non-apoptotic cells originated in organelles that looked morphologically like mitochondria. Using the mitochondria-specific stain TMRE on cells not exposed to H2DCF-DA, we found an identical pattern of fluorescence, proving that, apart from the weak cytosolic fluorescence, the DCF-induced fluorescence of normal cells originates in mitochondria (Figure 1B).
In normal cells, DCF-induced fluorescence is a function of cytosolic Fenton-type reactions and mitochondrial enzymatic oxidation by cytochrome c
As mentioned above, the main portion of the DCF-induced fluorescence of normal cells of both types showed a distinct mitochondrial pattern, although better visualized in the thin and flat ARPE-19 cells than in the thicker and more rounded-up J774 cells (Figures 1A and 2A). This was confirmed by the identical DCF- and TMRE-induced fluorescences (although overlay pictures could not be produced because of the green TMRE-induced fluorescence following excitation with blue light; TMRE is a metachromatic fluorophore). Since hydrophilic molecules do not penetrate cellular membranes, it seems safe to assume that the DCF-induced fluorescence originates in the mitochondrial intermembranous space where cytochrome c is present, together with traces of low-mass labile iron compounds under transport to the mitochondrial matrix.
Following exposure to FAC, the fluorescence from both compartments was substantially enhanced (Figure 3A), indicating the presence of H2DCF all over the cytosol as well as in the mitochondrial intermembranous space. This finding shows that oxidation of H2DCF requires the presence of low-mass iron in redox-active form, which was confirmed by the observation that cells exposed to the small iron chelator CP22 presented a clear depression of both cytosolic and mitochondrial DCF-induced fluorescence (Figure 3B).
Exposure to low-mass iron somewhat enhances the cytosolic as well as the mitochondrial DCF-induced fluorescence of ARPE-19 cells, whereas exposure to an iron chelator has the opposite effect
To demonstrate the capacity of cytochrome c to catalyse oxidation of H2DCF to DCF in the presence of H2O2, we applied the in vitro DCF test as described previously . As shown in Figure 4, cytochrome c, as well as hydroxyl radicals, readily induce such oxidation, which is in accordance with previous findings of other groups [16–18].
Cytochrome c oxidizes H2DCF in the presence of H2O2
Strong cytosolic DCF-induced fluorescence is a function of LMP-dependent relocation to the cytosol of redox-active iron, as well as of release of cytochrome c following MMP (mitochondrial membrane permeabilization)
As pointed out previously, short exposure to H2O2 (100 μM) is insufficient to induce LMP, as assayed by the AO-uptake method, in the very oxidative-stress-resistant ARPE-19 cells . The J774 cells are much more sensitive and apoptosis can easily be initiated by exposure for 30 min to 100 μM H2O2 (Figure 5). However, no immediate LMP occurs, not even if cells are exposed to higher concentrations, because peroxidation and ensuing fragmentation of the lysosomal membranes require some time . We hypothesized, on the basis of the results following the addition of FAC to the cells (see above), that LMP might cause DCF-induced fluorescence by allowing lysosomal transition metals (mainly iron) to relocate to the cytosol and meet H2DCF and catalyse its oxidation. LMP also results in MMP with release to the cytosol of cytochrome c with ensuing enzymatic oxidation of H2DCF (see further below).
ARPE-19 and J774 cell lysosomes are differently sensitive to oxidative stress
Following exposure of cells to the metachromatic and lysosomotropic fluorochrome AO, normal ARPE-19 and J774 cells show a large number of distinct and bright red lysosomes when activated with blue light, whereas the cytosol simultaneously shows moderate green fluorescence (Figures 5A and 5C).
As pointed out above, many J774 cells that were exposed to 100 μM H2O2 started to show signs of early apoptosis (nuclear pycnosis and plasma membrane blebbing). They also showed a reduced number of lysosomes and enhanced cytosolic green fluorescence following exposure to AO (Figure 5D). Meanwhile, the ARPE-19 cells, which are resistant to oxidative stress, remained normal when treated similarly (Figure 5B). As mentioned above, apoptotic and necrotic cells show a reduced number of intact lysosomes [21,22,27].
LMP induced without oxidative stress also causes strong diffuse DCF-induced fluorescence
The above-described results, obtained by exposing the cells to H2O2 (see also Figures 1D and 2C), suggested that oxidative stress does not necessarily give rise to more than a slight and almost insignificant increase in DCF-induced fluorescence. In order to substantiate that hypothesis further, we wanted to induce LMP by means other than oxidative stress. As shown previously, lysosomotropic detergents induce LMP [25,28]. By exposing ARPE-19 cells to MSDH at 100 μM and then comparing the resulting time-dependent increase in DCF-induced fluorescence intensity with the decreased number of intact lysosomes, as visualized with the AO-uptake test, it became clear that LMP and enhanced DCF-induced fluorescence were strongly linked (Figures 6A and 6B). This finding confirms that a high DCF-induced fluorescence is not necessarily a consequence of oxidative stress, but rather a function of relocation to the cytosol of lysosomal redox-active iron and mitochondrial cytochrome c. However, it should be kept in mind that lysosomal rupture, with release to the cytosol not only of low-mass iron, but also of cathepsins, certainly secondarily induces enhanced production of superoxide and H2O2 from damaged mitochondria.
Lysosomal rupture induced without oxidative stress greatly enhances cytosolic DCF-mediated fluorescence in ARPE-19 cells
The findings of the present study show that significant cytosolic oxidation of H2DCF to DCF depends on the combined effect of Fenton-type reactions and enzymatic activity of cytochrome c. Consequently, it should not be considered to be a result of the presence of unspecified ROS. Rather, the reaction is apparently dependent on the relocation of transition metals in redox-active form from lysosomes and/or of cytochrome c from mitochondria. Moreover, because H2DCF is a hydrophilic alcohol, it would not, under normal conditions, be expected to traverse cellular membranes, apart from the fenestrated mitochondrial outer membrane, which is known to allow molecules up to 5000 Da to pass .
The observed mitochondrial fluorescence from normal cells, albeit low, was also observed by at least one other group of investigators . It reflects the combined presence of redox-active iron, H2O2 and enzymatically active cytochrome c in the intermembranous space. Haem proteins, such as cytochrome c are iron-containing proteins that act as pseudoperoxidases owing to their content of compound I, which consists of a ferryl Fe(IV) species and a porphyrin radical state of haem . Since the water-soluble low-molecular-mass iron chelator CP22 was found to somewhat depress both the weak mitochondrial as well as the even weaker cytosolic fluorescence of normal cells, it is likely that some low-mass iron in labile form is responsible for the formation of DCF in the cytosol of normal cells and, partly, in the mitochondria. Most probably, such iron is under transport from lysosomes, where it has been liberated following degradation of autophagocytosed ferruginous macromolecules , for storage in ferritin or for synthesis of iron-containing macromolecules. As is well known, many metabolic synthetic steps that involve iron take place both in the cytosol and, above all, in the mitochondrial matrix.
When cells were initially exposed to FAC, the cytosolic and mitochondrial DCF-induced fluorescence was clearly enhanced. This observation indicates that, following exposure to H2DCF-DA, the cleaved reaction product H2DCF is distributed evenly in the cytosol and also enters the mitochondrial intermembranous space. The finding that the cytosolic fluorescence of cells not exposed to FAC is weak, even after a short exposure to H2O2, suggests that the amount of labile cytosolic iron is normally minute and that iron under transport from the lysosomal compartment may well be carried in a non-redox-active form .
Only when H2DCF interacts with significant amounts of redox-active low-mass transition metals or with cytochrome c may we expect a more pronounced general cellular DCF-induced fluorescence. It is now well established that the only cellular organelle that contains such metals (mainly iron) in significant amounts is the lysosomal compartment [22,23,31]. We have shown previously, using the calcein method for detection of ‘labile iron’, that when lysosomes rupture, low-mass iron is relocated to the cytosol . Thus it is reasonable to assume that LMP, rather than oxidative stress, would be the major mechanism behind any obvious DCF-induced fluorescence. It was therefore no surprise to find that the exposure of cells to the lysosomotropic detergent MSDH in amounts that induce LMP and ensuing apoptosis/necrosis initiated strong DCF-induced fluorescence. Under normal conditions, hydroxyl radicals are regularly produced intralysosomally because the H2O2 that is normally present in small amounts in the cell permeates lysosomes and initiates Fenton-type reactions. These radicals are, however, not detected using the DCF method because H2DCF is a hydrophilic non-permeating molecule. Following LMP and ensuing relocation of lysosomal redox-active iron to the cytosol, the very same amount of iron now gives rise to DCF-induced fluorescence since the LMP that occurred allows H2DCF, redox-active iron and H2O2 to meet. Consequently, at the very same degree of ‘oxidative stress’, it is the stability of the lysosomal membranes that influences the outcome of the DCF test. Importantly, when the test is positive, no higher ‘oxidative stress’ necessarily needs to be present than when it is negative.
Importantly, oxidative stress in the form of a brief exposure to H2O2 only slightly changed the fluorescence of both ARPE-19 cells, which are highly resistant to oxidative stress, and J774 cells that are much more sensitive. The J774 cells, however, following a 30 min period of oxidative stress and a period of several hours under standard culture conditions, showed both LMP and a strong DCF-induced fluorescence. This shows that H2O2 by itself does not induce oxidation of H2DCF, unless small amounts of labile iron are present, but also that peroxidation/fragmentation of lysosomal membranes following oxidative stress is a time-dependent process . Since both the studied cell types degrade H2O2 to a similar degree , it is reasonable to assume that their large differences in lysosomal stability against oxidative stress reflect much different intralysosomal concentrations of redox-active transition metals.
In the literature, we find many examples of the DCF test being used to demonstrate ‘oxidative stress’ when in fact it most probably reflects LMP. A typical example is , where it is shown (in Figure 1 of the paper) that transgenic islet cells overexpressing metallothioneins are significantly “less sensitive to oxidative stress” than control cells. Importantly, in this study, the authors found that DCF-induced fluorescence engendered in normal cells following the addition of H2O2 was nearly absent from cells overexpressing metallothionein, despite the fact that the same concentration of peroxide was present . We showed previously that overexpression of metallothioneins or HSP70 (heat-shock protein 70) gives rise to increased autophagy of these proteins that, in turn, creates a temporary reduction of intralysosomal redox-active iron that lasts as long as the influx of metallothioneins persists. That makes lysosomes less prone to undergo LMP following exposure to various forms of oxidative stress [15,34].
Another example is  where it is shown that radiation of cultured human prostatic cancer cells in combination with exposure to H2O2 strongly enhances formation of ROS, as evaluated using the H2DCF-DA method. Pre-treatment with ammonium chloride abolished the formation of DCF-induced fluorescence, although cells were then exposed to the same concentration of H2O2. Moreover, it was demonstrated that radiation plus H2O2 ruptures most lysosomes, while they were preserved in cells pre-treated with ammonium chloride. The results strongly suggest that the appearance of cytosolic DCF-induced fluorescence requires lysosomal rupture with relocation of redox-active iron that in turn mediates oxidation of H2DCF to DCF. Such oxidation may very well take place without any enhanced oxidative stress at all, but rather rely on normal cellular concentrations of H2O2 for the necessary formation of hydroxyl radicals.
A clear understanding of the factors that determine DCFinduced fluorescence is required in order to appreciate what DCF-induced fluorescence really means. Consequently, and based on the results of the present study, we suggest that it should be considered to reflect LMP and MMP with ensuing relocation to the cytosol of redox-active iron and cytochrome c rather than to be the result of some incompletely defined ‘ROS or oxidative stress’.
In summary, strong DCF-induced fluorescence seems to require the simultaneous presence in the cytosol of either H2DCF, H2O2 and Fe(II) or H2DCF and cytochrome c. Relocation to the cytosol of Fe(II) or cytochrome c, in turn, requires LMP or detachment of cytochrome c from the outside of the inner mitochondrial membrane. In apoptosis, LMP is often an upstream event that results in MMP with release of cytochrome c [22,31]. Consequently, induction of DCF-induced fluorescence may indicate LMP and/or MMP, stemming from whatever cause and not just from oxidative stress.
age-related macular degeneration
Dulbecco's modified Eagle's medium
ferric ammonium citrate
fetal bovine serum
Hanks balanced salt solution
lysosomal membrane permeabilization
mitochondrial membrane permeabilization
O-methylserine dodecylamide hydrochloride
reactive oxygen species
retinal pigment epithelial
tetramethylrhodamine ethyl ester
Markus Karlsson performed most of the experiments and participated in planning and writing. Tino Kurz carried out the in vitro DCF assay and participated in planning and writing. Ulf Brunk planned the study, evaluated data and overlooked the writing. Sven Nilsson and Christina Frennesson participated in planning and writing.
We thank Mr Stephen Hampson for linguistic editing before acceptance.
This work was supported by the Crown Princess Margareta's Foundation for the Visually Handicapped and the Linköping University Hospital Research Fund (ALF).