Fluorescent cationic probes of mitochondria. Metrics and mechanism of interaction.

Mitochondria strongly accumulate amphiphilic cations. We report here a study of the association of respiring rat liver mitochondria with several fluorescent cationic dyes from differing structural classes. Using gravimetric and fluorometric analysis of dye partition, we find that dyes and mitochondria interact in three ways: (a) uptake with fluorescence quenching, (b) uptake without change in fluorescence intensity, and (c) lack of uptake. For dyes that quench upon uptake, the extent of quenching correlates with the degree of aggregation of the dye to dimers, as predicted by theory (Tomov, T.C. 1986. J. Biochem. Biophys. Methods. 13:29-38). Also predicted is the relationship observed between quenching and the mitochondria concentration when constant dye is titrated with mitochondria. Not predicted is the relationship observed between quenching and dye concentration when constant mitochondria are titrated with dye. Because a limit to dye uptake exists, in this case, the degree of quenching decreases as dye is added. A Langmuir isotherm analysis gives phenomenological parameters that predict quenching when it is observed as a function of dye concentration. By allowing for a decrease in membrane potential, caused by incorporation of cationic dye into the lipid bilayer, a modification of the Tomov theory predicts the dye titration data. We present a model of cationic dye-mitochondria interaction and discuss the use of these as probes of mitochondrial membrane potential.     

Pontentiometric cyanine dyes are sensitive probes for mitochondria in intact plant cells : kinetin enhances mitochondrial fluorescence

Selected fluorescent dyes were tested for uptake by mitochrondria in intact cells of barley, maize, and onion. The cationic cyanine dye 3,3′-diheptyloxacarbocyanine iodide [DiOC₇(3)] accumulated in mitochondria within 15 to 30 minutes without appreciable staining of other protoplasmic constituents. The number, shape, and movement of the fluorescent mitochondria could be seen readily, and the fluorescence intensity of the mitochondria could be monitored with a microscope photometer. Fluorescence was eliminated in 1 to 5 minutes by the protonophore carbonyl cyanide m-chlorophenylhydrazone (CCCP) indicating that maintenance of dye concentration was dependent on the inside-negative transmembrane potential maintained by functional mitochondria. Fluorescence of prestained mitochondria was enhanced within 5 to 10 minutes after addition of 0.1 millimolar kinetin to cells. The fluorescence in kinetintreated cells was dissipated by CCCP. These results suggest that kinetin interacted with respiratory processes resulting in higher potential across the mitochondrial membrane.     

Effect of Bcl-2 overexpression on mitochondrial structure and function

Overexpression of the antiapoptotic Bcl-2 protein enhances the uptake of fluorimetric dyes sensitive to mitochondrial membrane potential, suggesting that Bcl-2 changes the mitochondrial proton gradient. In this study, we performed calibrated measurements of mitochondrial respiration, membrane potential, deltapH, and intramitochondrial [K+] in digitonin-permeabilized PC12 and GT1-7 neural cells that either do not express human Bcl-2 (control transfectants) or that were transfected with and overexpressed the human bcl-2 gene to evaluate whether Bcl-2 alters mitochondrial inner membrane ion transport. We found that although Bcl-2-overexpressing cells exhibit higher fluorescence responses to membrane potential, pH, and K+-sensitive dyes, this increased response is due to an enhanced accumulation of these dyes and not an increased mitochondrial membrane potential, deltapH, or [K+]. This result is supported by the presence of equal respiratory rates in Bcl-2+ and Bcl-2- cells. Possible structural alterations in Bcl-2+ mitochondria that could account for increases in fluorescent dye uptake were evaluated using flow cytometry particle sizing and light scattering determinations. These experiments established that Bcl-2-overexpressing mitochondria present both increased volume and structural complexity. We suggest that increased mitochondrial volume and structural complexity in Bcl-2+ cells may be related to many of the effects of this protein involved in the prevention of cell death.     

Safranine O as a fluorescent probe for mitochondrial membrane potential studied on the single particle level and in suspension

The permeant cationic dye safranine O is often used to measure mitochondrial membrane potential due to the dependence of both its absorption and fluorescence on mitochondrial energization, which causes its oligomerization inside mitochondria. In the present study we have used fluorescent correlation spectroscopy (FCS) to record the fluorescence changes on a micro level, i.e. under conditions permitting resolution of contributions from single particles (molecules of the dye and stained mitochondria). We have shown that the decrease in fluorescence signal from a suspension of energized mitochondria stained with a high safranine concentration (10 μM) is explained by the decrease in dye concentration in the medium in parallel with the accumulation of the dye inside the mitochondria, which results in fluorescence quenching. With 1 μM safranine O, the fluorescence rise after energization is caused by the accumulation of the dye up to a level not sufficient for full fluorescence quenching and also by the higher intensity of mitochondrial fluorescence on immersion of the dye in the hydrophobic milieu. Besides the estimation of the inner mitochondrial membrane potential, this approach also assesses the concentration of fluorescent particles. The non-monotonic dependence of the FCS parameter 1/G(τ→0) on the concentration of mitochondrial protein suggests heterogeneity of the system with respect to fluorescence of particles. An important advantage of the described method is its high sensitivity, which allows measurements with low concentrations and quantities of mitochondrial protein in samples (less than 10 μg).     

A ratiometric fluorescent probe for assessing mitochondrial phospholipid peroxidation within living cells

Mitochondrial oxidative damage contributes to a wide range of pathologies, and lipid peroxidation of the mitochondrial inner membrane is a major component of this disruption. However, despite its importance, there are no methods to assess mitochondrial lipid peroxidation within cells specifically. To address this unmet need we have developed a ratiometric, fluorescent, mitochondria-targeted lipid peroxidation probe, MitoPerOx. This compound is derived from the C11-BODIPY(581/591) probe, which contains a boron dipyromethane difluoride (BODIPY) fluorophore conjugated via a dienyl link to a phenyl group. In response to lipid peroxidation the fluorescence emission maximum shifts from ～590 to ～520nm. To target this probe to the matrix-facing surface of the mitochondrial inner membrane we attached a triphenylphosphonium lipophilic cation, which leads to its selective uptake into mitochondria in cells, driven by the mitochondrial membrane potential. Here we report on the development and characterization of MitoPerOx. We found that MitoPerOx was taken up very rapidly into mitochondria within cells, where it responded to changes in mitochondrial lipid peroxidation that could be measured by fluorimetry, confocal microscopy, and epifluorescence live cell imaging. Importantly, the peroxidation-sensitive change in fluorescence at 520nm relative to that at 590nm enabled the use of the probe as a ratiometric fluorescent probe, greatly facilitating assessment of mitochondrial lipid peroxidation in cells.     

Membrane potential of Plasmodium falciparum gametocytes monitored with rhodamine 123

The membrane potential of Plasmodium falciparum gametocytes was monitored with the cationic permeant fluorescent dye rhodamine 123 (R123) as a probe. Epifluorescence microscopy revealed that R123 at 1 microgram/ml rather selectively partitioned into structure resembling large mitochondria. Treatment of R123-loaded gametocytes with various inhibitors including those of respiration resulted in disappearance of fluorescence from what appeared to be the mitochondria, but not from the cytosol. These results indicate that P. falciparum gametocytes have the mitochondrion maintaining an inside negative membrane potential.     

Oxazole yellow homodimer YOYO-1-labeled DNA: a fluorescent complex that can be used to assess structural changes in DNA following formation and cellular delivery of cationic lipid DNA complexes.

To improve transfection efficiency following delivery of plasmid expression vectors using lipid-based carriers, it is crucial to define structural characteristics of the lipid/DNA complexes that optimize transgene expression. Due to its strong affinity for DNA and high quantum yield, the fluorescent DNA intercalator YOYO-1 was used as a tool to assess changes in DNA that occur following lipid binding and cell delivery. In this study, the stability of the dye/DNA complex following binding of poly-L-lysine or monocationic lipids is characterized. More than 98% of the fluorescence measured for a defined DNA/YOYO-1 complex was lost when DNA was condensed using poly-L-lysine. This loss in fluorescence could be attributed to displacement of bound dye. In contrast, more than 30% of the fluorescence of the dye-labeled DNA was retained after formation of cationic lipid/DNA complexes. Significantly, the results illustrate differences in structural changes cationic lipids and PLL exert on plasmid DNA. The fluorescent lipid/DNA complex was used to assess DNA delivery to murine B16/BL6 cells in vitro. An assay relying on fluorescence resonance energy transfer between bound YOYO-1 and propidium iodide was used to distinguish between DNA attached to the cell surface and internalized DNA.     

Changes in mitochondrial function in the cells and cytoplasts of a pig embryo kidney culture as affected by the action of actinomycin D

Vital staining of PE kidney cells by fluorescent cationic dye, ethyl rhodamine, (accumulated inside mitochondria by the membrane potential and hence reflecting their functional state) shows rather close level of the fluorescence intensity both in cells and in their cytoplasts for 10 hours of culture survival in the standard medium. In cells and cytoplasts cultivated in the medium with 0.2 mg/ml of actinomycin D inhibiting RNA synthesis, fluorescence intensity of mitochondria sharply decreases after 10 hours as against the control patterns. It is concluded that mitochondria possess a significant degree of autonomy of the nucleus and it is supposed that a considerable part of mitochondrial RNA is under the nucleus control.     

YidC-driven membrane insertion of single fluorescent Pf3 coat proteins

The membrane insertion of single bacteriophage Pf3 coat proteins was observed by confocal fluorescence microscopy. Within seconds after addition of the purified and fluorescently labeled protein to liposomes or proteoliposomes containing the purified and reconstituted membrane insertase YidC of Escherichia coli, the translocation of the labeled residue was detected. The 50-amino-acid-long Pf3 coat protein was labeled with Atto520 and inserted into the proteoliposomes. Translocation of the dye into the proteoliposome was revealed by quenching the fluorescence outside of the vesicles. This allowed us to distinguish single Pf3 coat proteins that only bound to the surface of the liposomes from proteins that had inserted into the bilayer and translocated the dye into the lumen. The Pf3 coat protein required the presence of the YidC membrane insertase, whereas mutants that have a membrane-spanning region with an increased hydrophobicity were autonomously inserted into the liposomes without YidC.     

Binding of a cationic phenazinium dye in anionic liposomal membrane: a spectacular modification in the photophysics

Interaction of a cationic phenazinium dye, phenosafranin (PSF), with the anionic liposomal vesicle/bilayer of dimyristoyl-l-α-phosphatidylglycerol (DMPG) has been demonstrated using steady state and time resolved fluorescence and fluorescence anisotropy techniques. The charge transfer emission spectrum of PSF shows a dramatic modification in terms of fluorescence yield together with an appreciable hypsochromic shift in the lipid environment. The blue shift indicates a lowering in polarity inside the vesicle as compared to that in bulk water. The fluorescence and fluorescence quenching studies and micropolarity determination reveal that the cationic fluorophore has a profound binding interaction with the anionic DMPG membrane. Anisotropy study indicates the imposition of a motional restriction on the probe inside the bilayer. The electrostatic interaction between the cationic dye and the anionic lipid membrane has been argued to be the reason behind all these observations. The results could be useful in analyzing membrane organization and heterogeneity in natural membranes exploiting PSF or alike compounds as fluorescent probes.     

Do single mitochondria contain zones with different membrane potential?

Observations of Lan Bo Chen’s group using a mitochondria-selective fluorochrome 5,5’,6,6’- tetrachloro- 1,1’,3,3’- tetraethylbenzimidazolocarbocyanine iodide (JC-1) indicate that mitochondria in situ may have zones of different electrochemical potential along their length. This was indicated by the formation of J-aggregates of this dye at distinct sites along a single mitochondrion. Also, intensity variations along single mitochondria were found with diamino-styryl-pyridinium methiodide (DASPMI), another fluorochrome that selectively stains mitochondria depending on their electrochemical potential. DASPMI exchanges easily with the cytoplasm and changes its quantum yield when bound to mitochondrial membranes. Therefore, fluorescence intensity is primarily controlled by the membrane environment rather than by mass accumulation. Two possible explanations of intramitochondrial fluorescence intensity variations have to be discussed: variations in the amount of mitochondrial inner membrane per unit of projection area (or voxel), and differences in the electrochemical gradient. This problem has been approached by comparing fluoro-micrographs of mitochondria in endothelial cells stained with either JC-1 or DASPMI with electron micrographs of the same mitochondria after fixation with glutardialdehyde and osmium tetroxide and ultrathin sectioning. JC-1 red fluorescence (revealing J-aggregate formation) as well as high-intensity staining with DASPMI correlate roughly with the local thickness of mitochondria; no differences in the crista organization are revealed for those areas or mitochondria exhibiting red JC-1 fluorescence and those with green fluorescence. The distance between red fluorescing areas in a single mitochondrion seem to be caused by competition for dye molecules placed in between centres of JC-1 aggregation. Isolated mitochondria are of uniform small size and spherical shape; therefore, no differences in shape interfere with JC-1 staining. Thus JC-1 may be an appropriate indicator of membrane potential in isolated mitochondria. In living cells mitochondria often are large and elongated, and thus the situation is not straightforward to interpret. However, evidence is provided that there are submitochondrial zones, which differ in membrane potential from one adjacent area to another, because DASPMI staining of intramitochondrial zones reveals differences in fluorescence intensity and preferred photodamage of these areas. In some cases separation of the zones of higher membrane potential by cristae traversing the whole diameter of a mitochondrion has been observed. Local photobleaching of stained mitochondria results in a loss of fluorescence along the total length of a mitochondrion. However, this type of bleaching develops over tens of seconds, not in the very short time range (e.g. ms) expected from the discharge of all the membranes if they were electrically coupled.     

Influx and efflux kinetics of cationic dye binding to respiring mitochondria.

We have investigated the kinetics of interaction of cationic fluorescent lipophiles (dyes) rhodamine 123, rhodamine 6G, tetramethyl rhodamine ethyl ester, safranine O, 1,1'-diethyloxacarbocyanine, 1,1'-diethyloxadicarbocyanine, and 1,1'-diethylthiadicarbocyanine iodide with isolated respiring rat-liver mitochondria (RLM). Dye flux across the RLM inner membrane was measured by following the kinetics of fluorescence signal change after mixing of dye and RLM. The time course of fluorescence was analysed in terms of a kinetic model of the binding and transport processes involved. The rate constants of dye influx and efflux were extracted from the observed effect on the apparent time constant of fluorescence change to equilibrium intensity upon mixing dye with increasing concentrations of RLM. From the influx rate constants obtained, the apparent permeability constants for dye influx (at zero potential) across the membrane were calculated and ranged from 3 to 140 x 10(-4) cm/s. The influx rate constant was found to be linearly related to relative dye lipophilicity, as predicted by the model. As another test of the model, from the ratio of the influx and efflux rate constants, the apparent trans-membrane potential, psi, was calculated and found generally to agree with reported values, but to depend on the lipophilicity of the dye used. Not predicted by the simple model was a dissymmtry observed in the influx and efflux time constants for fluorescence change to equilibrium intensity. Inferences are made relating to the utility of these dyes as probes of psi.     

Efficiency of mitochondrial uncoupling by modified butyltriphenylphosphonium cations and fatty acids correlates with lipophilicity of cations: Protonophoric vs leakage mechanisms

In order to determine the share of protonophoric activity in the uncoupling action of lipophilic cations a number of analogues of butyltriphenylphosphonium with substitutions in phenyl rings (C₄TPP-X) were studied on isolated rat liver mitochondria and model lipid membranes. An increase in the rate of respiration and a decrease in the membrane potential of isolated mitochondria were observed for all the studied cations, the efficiency of these processes was significantly enhanced in the presence of fatty acids and correlated with the octanol-water partition coefficient of the cations. The ability of C₄TPP-X cations to induce proton transport across the lipid membrane of liposomes loaded with a pH-sensitive fluorescent dye increased also with their lipophilicity and depended on the presence of palmitic acid in the liposome membrane. Of all the cations, only butyl[tri(3,5-dimethylphenyl)]phosphonium (C₄TPP-diMe) was able to induce proton transport by the mechanism of formation of a cation-fatty acid ion pair on planar bilayer lipid membranes and liposomes. The rate of oxygen consumption by mitochondria in the presence of C₄TPP-diMe increased to the maximum values corresponding to conventional uncouplers; for all other cations the maximum uncoupling rates were significantly lower. We assume that the studied cations of the C₄TPP-X series, with the exception of C₄TPP-diMe at low concentrations, cause nonspecific leak of ions through lipid model and biological membranes which is significantly enhanced in the presence of fatty acids.     

The fluorescent cationic dye rhodamine 6G as a probe for membrane potential in bovine aortic endothelial cells.

The membrane potential of cultured bovine aortic endothelial cells was assessed by a fluorescent probe as an alternative to direct methods. We used the fluorescent cationic dye rhodamine 6G, a lipophilic probe with high permeability in cell membranes. A linear relationship was obtained between fluorescence intensity (F.I.) and membrane potential (Em) as a function of the extracellular Na(+) concentration in the presence of the ionophore gramicidin. From the equation derived from the linear relationship F.I. = -0.004 Em + 0. 03 (P < 0.001), the fluorescence measurements could be converted to membrane potential. The resting plasma membrane potential obtained was -65 +/- 7 mV. Nigericin (27 microM), ouabain (1 mM), and bradykinin (20 nM) induced a decrease in F.I. (depolarization), while ATP (25-100 microM) induced an increase in F.I. (hyperpolarization). Mitochondrial membrane potential inhibitors myxothiazol (3 microM) and oligomycin (4 microM) did not influence F. I. measured in the cultured bovine aortic endothelial cells. The results indicate that rhodamine 6G can be used as a sensitive and specific dye in studies of substances that affect the membrane potential of endothelial cells.     

Interaction of fluorescent berberine alkyl derivatives with respiratory chain of rat liver mitochondria

The cationic fluorescent dyes, berberines, have been observed to inhibit NAD-linked respiration in rat liver mitochondria. Low concentrations inhibit electron transport in the NAD-ubiquinone span after penetration into mitochondria. More hydrophobic alkyl derivatives proved to be stronger inhibitors showing more rapid onset of inhibition. The inhibition was totally dependent on the energization of the membrane; however, the addition of a hydrophobic anion stimulated the inhibition effects in uncoupled mitochondria. Substantially higher concentrations of berberines are needed for the inhibition of the oxidation of succinate. The excess of dye interacting with surface dipoles in the energized state can inhibit the energy transduction through the complexbc ₁. On the basis of the difference in the rate of fluorescence response when berberines are added to coupled mitochondria and the corresponding inhibition effects, the presence minimally of two binding sites was suggested. The dye bound on the outer surface is highly fluorescent and inhibits the energy transduction if added in excess. The remaining dye interacting with NADH dehydrogenase does not fluoresce. The accumulation of alkylberberine in mitochondria results in additional effects in the region of cytochromeb the nature of which is not fully understood.     

Analysis of the membrane potential of rat- and mouse-liver mitochondria by flow cytometry and possible applications

Washed and purified rat- or mouse-liver mitochondria exhibiting high membrane integrity and metabolic activity were studied by flow cytometry. The electrophoretic accumulation/redistribution of cationic lipophilic probes, rhodamine 123, safranine O and a cyanine derivative, 3,3'-dihexyloxadicarbocyanine iodide, during the energization process was studied and was consistent with the generation of a negative internal membrane potential. An exception to this was nonylacridine orange which spontaneously bound to the mitochondrial membrane by hydrophobic interactions via its hydrocarbon chain. Energized purified mitochondria stained with potentiometric dyes exhibited both higher fluorescence and population homogeneity than the non-energized or deenergized (nigericin plus valinomycin) mitochondria. By contrast, under non-energized or deenergized conditions, the mitochondrial population exhibited fluorescence intensity heterogeneity related to the residual membrane potential; two subpopulations were evident, one of low fluorescence which may be related to the autofluorescence of the mitochondria (plus non-specific dye binding) and a second population which exhibited high fluorescence. Flow cytometry of the unpurified, simply washed, rat-liver mitochondria stained with rhodamine 123, a classically used dye, provided evidence of their heterogeneity in terms of light-scattering properties and membrane-potential-related fluorescence. One third of the washed mitochondria were found to be non-functional by such assays. The fluorescence of purified rat-liver mitochondria due to the membrane potential built up by endogenous substrates indicates heterogeneity of the mitochondrial population with respect to levels of endogenous substrates. The low-angle light scattering increases upon energization and provides some original information about the shape and modification of the inner mitochondrial conformation accompanying the energization. The heterogeneity of the rat liver mitochondrial population, from a structural, metabolic (existence of endogenous substrates) and functional (active and non-active mitochondrial population dispersion) point of view could thus be demonstrated by flow-cytometry analysis. Two animal models were examined with regard to the alteration of the mitochondrial membrane potential under the effects of drugs (rat-liver mitochondria), and the effects of ammonium toxicity (mouse-liver mitochondria). These results are promising and open new perspectives in the study of mitochondriopathies.     

Berberine derivatives as cationic fluorescent probes for the investigation of the energized state of mitochondria

The interaction of rat liver and bovine heart mitochondria with a series of fluorescent, cationic berberine derivatives varying in the length of alkyl chain has been investigated. An increase in the hydrophobicity of the derivative was accompanied by a larger value of the partition coefficient and by binding to a more hydrophobic region of the inner mitochondrial membrane. It was found that berberines could be used as sensitive indicators of processes which take place on the outer surface of the mitochondrial membrane; the greatest (15-fold) increase in fluorescence was obtained with 13-methylberberine in the energized state of mitochondria. The fluorescence increase was due to the increase in fluorescence quantum yield although a small increase in the amount of bound derivative could also be detected upon energization. The fluorescence was linearly dependent on the magnitude of the membrane potential. In parallel with an observed fluorescence enhancement a considerable decrease in rotational mobility was found. We suggest that berberines move in the inner membrane according to the polarity of the membrane potential; consequently, deeper immersion in the less polar region in the energized state brings about a larger fluorescence increase. More hydrophobic derivatives inhibited NAD-linked respiration in rat liver mitochondria but exerted no effect on succinate oxidation up to 10 μM concentration.     

Staining of Plasmodium yoelii-infected mouse erythrocytes with the fluorescent dye rhodamine 123

The cationic permeant fluorescent dye rhodamine 123 (R123) was used to stain Plasmodium yoelii-infected mouse erythrocytes. Fluorescence microscopic observations demonstrated that the parasite, but not the matrix of the infected erythrocyte, accumulated the dye. Differences in fluorescence intensity could not be found at the various developmental stages of the parasite; however, quantitation of the cell-associated dye revealed an increase in R123 uptake with parasite development. The retention of the parasite-associated dye, as measured by fluorescence microscopy and spectrophotometry after extraction of R123 with butanol, was markedly reduced by treatment of the infected erythrocytes with a proton ionophore, carbonyl cyanide m-chlorophenylhydrazone (CCCP), and an inhibitor of proton ATPase, dicyclohexylcarbodiimide (DCCD). These results indicate that the accumulation and retention of R123 in P. yoelii reflect the parasite membrane potential and suggest that the parasite plasma membrane has a membrane potential-generating proton pump.     

Discrimination of depolarized from polarized mitochondria by confocal fluorescence resonance energy transfer

Mitochondrial depolarization promotes apoptotic and necrotic cell death and possibly other cellular events. Polarized mitochondria take up cationic tetramethylrhodamine methylester (TMRM), which is released after depolarization. Thus, TMRM does not label depolarized mitochondria. To identify both polarized and depolarized mitochondria in living cells, cultured rat hepatocytes, and sinusoidal endothelial cells were co-loaded with green-fluorescing MitoTracker Green FM (MTG) and red-fluorescing TMRM for imaging by laser scanning confocal microscopy. Like TMRM, MTG is a cationic fluorophore that accumulates electrophoretically into polarized mitochondria. Unlike TMRM, MTG binds covalently to intramitochondrial protein thiols and remains bound after depolarization. In cells labeled only with MTG, excitation with blue (488 nm) light yielded green but almost no red fluorescence. After subsequent loading with TMRM, green MTG fluorescence became quenched. Instead, blue excitation yielded red fluorescence. Mitochondrial de-energization restored green fluorescence and abolished red fluorescence. Conversely, when MTG was added to TMRM-labeled cells, red fluorescence excited by blue light was enhanced, an effect again reversed by de-energization. These observations of reversible quenching of donor fluorescence and augmentation of acceptor fluorescence signify fluorescence resonance energy transfer (FRET). In undisturbed hepatocytes, spontaneous depolarization of a subfraction of mitochondria was an ongoing phenomenon. In conclusion, confocal FRET discriminates individual depolarized mitochondria against a background of hundreds of polarized mitochondria.     

Rhodamine 123 as a probe of transmembrane potential in isolated rat-liver mitochondria: spectral and metabolic properties

The spectral and metabolic properties of Rhodamine 123, a fluorescent cationic dye used to label mitochondria in living cells, were investigated in suspensions of isolated rat-liver mitochondria. A red shift of Rhodamine 123 absorbance and fluorescence occurred following mitochondrial energization. Fluorescence quenching of as much as 75% also occurred. The red shift and quenching varied linearly with the potassium diffusion potential, but did not respond to delta pH. These energy-linked changes were accompanied by dye uptake into the matrix space. Concentration ratios, in-to-out, approached 4000:1. A large fraction of internalized dye was bound. At concentrations higher than those needed to record these spectral changes, Rhodamine 123 inhibited ADP-stimulated (State 3) respiration of mitochondria (Ki = 12 microM) and ATPase activity of inverted inner membrane vesicles (Ki = 126 microM) and partially purified F1-ATPase (Ki = 177 microM). The smaller Ki for coupled mitochondria was accounted for by energy-dependent Rhodamine 123 uptake into the matrix. Above about 20 nmol/mg protein (10 microM), Rhodamine 123 caused rapid swelling of energized mitochondria. Effects on electron-transfer reactions and coupling were small or negligible even at the highest Rhodamine 123 concentrations employed. delta psi-dependent Rhodamine 123 uptake together with Rhodamine 123 binding account for the intense fluorescent staining of mitochondria in living cells. Inhibition of mitochondria ATPase likely accounts for the cytotoxicity of Rhodamine 123. At concentrations which do not inhibit mitochondrial function, Rhodamine 123 is a sensitive and specific probe of delta psi in isolated mitochondria.