Mitochondrial targeting for photochemotherapy. Can selective tumor cell killing be predicted based on n-octanol/water distribution coefficients?

The concept of mitochondrial targeting for chemo- and photochemotherapy of neoplastic diseases has its origin in the observation that enhanced mitochondrial transmembrane potential is a common tumor cell phenotype. As a result of this enhanced transmembrane potential, a number of cationic dyes accumulate in larger amounts and are retained for longer periods in the mitochondria of tumor cells than in normal cells. Only a relatively small number of (photo)toxic dyes known to localize in energized cell mitochondria are capable of inducing the destruction of tumor cells with desirable degrees of selectivity, however. We investigated how lipophilic character may affect the degree of specificity with which cationic dyes localize in energized cell mitochondria and how mitochondrial specificity may affect tumor cell selectivity. To this end, we used fluorescence microscopy to characterize the subcellular localization of ethyl violet and seven analogs of the prototypical mitochondria-specific dye, rhodamine 123. All cationic rhodamines studied here (-0.62 < log D(ow) < 1.60, where D(ow) represents the n-octanol/water distribution coefficient) were found to show considerable mitochondrial specificity, while the more lipophilic ethyl violet (log D(ow) = 2.37) did not. Ethyl violet was found to localize not only in mitochondria, but also in lysosomes. We also compared the degree of selective tumor cell killing induced by ethyl violet and two phototoxic rhodamines, i.e., the dibromo derivatives of rhodamine 123 and its n-octyl ester analog. While ethyl violet induces the destruction of human uterine sarcoma (MES-SA) cells and normal green monkey kidney cells (CV-1) with comparable efficiency, the mitochondria-specific dibromorhodamines were found to induce the destruction of MES-SA cells with considerable selectivity. Our findings are consistent with the premise that mitochondrial localization per se does not provide successful selective tumor cell killing using mitochondrial targeting. Our results reinforce the hypothesis that while most cationic dyes can be expected to localize at least to some extent in energized cell mitochondria, only those showing virtually absolute mitochondrial specificity can actually mediate the destruction of tumor cells with desirable selectivity. These findings also support the hypothesis that the probability of success of mitochondrial targeting in photochemotherapy of neoplastic diseases is bound to be higher when the D(ow) associated with the drug candidate falls within approximately two orders of magnitude of that of rhodamine 123.

Effect of the lipophilic/hydrophilic character of cationic triarylmethane dyes on their selective phototoxicity toward tumor cells.

The observation that enhanced mitochondrial transmembrane potential is a prevalent tumor cell phenotype has provided the conceptual basis for the development of mitochondrial targeting as a novel therapeutic strategy for both chemo- and photochemotherapy of neoplastic diseases. Because the plasma transmembrane potential is negative on the inner side of the cell and the mitochondrial transmembrane potential is negative on the inner side of this organelle, extensively conjugated cationic molecules (dyes) displaying appropriate structural features are driven electrophoretically through these membranes and tend to accumulate inside energized mitochondria. As a result of the higher mitochondrial transmembrane potential typical of tumor cells, a number of cationic dyes preferentially accrue and are retained for longer periods in the mitochondria of these cells compared to normal cells. This differential in both drug loading and retention brings about the opportunity to attack and destroy tumor cells with a high degree of selectivity. Only a small subset of the cationic dyes known to accumulate in energized mitochondria mediate the destruction of tumor cells with a high degree of selectivity, and the lack of a reliable model to describe the structural determinants of this tumor specificity has prevented mitochondrial targeting from becoming a more reliable therapeutic strategy. We describe here a systematic study of how the molecular structure of closely related cationic triarylmethanes affects the selectivity with which these dyes mediate the photochemical destruction of tumor cells. Based on our observations of how the lipophilic/hydrophilic character of these dyes affects tumor selectivity, we propose a simple model to assist in the design of new drugs tailored specifically for imaging and selective destruction of neoplastic tissue via mitochondrial targeting.

Effect of molecular structure on the performance of triarylmethane dyes as therapeutic agents for photochemical purging of autologous bone marrow grafts from residual tumor cells.

Extensively conjugated cationic molecules with appropriate structural features naturally accumulate into the mitochondria of living cells, a phenomenon typically more prominent in tumor than in normal cells. Because a variety of tumor cells also retain pertinent cationic structures for longer periods of time compared with normal cells, mitochondrial targeting has been proposed as a selective therapeutic strategy of relevance for both chemotherapy and photochemotherapy of neoplastic diseases. Here we report that the triarylmethane dye crystal violet stains cell mitochondria with efficiency and selectivity, and is a promising candidate for photochemotherapy applications. Crystal violet exhibits pronounced phototoxicity toward L1210 leukemia cells but comparatively small toxic effects toward normal hematopoietic cells (murine granulocyte-macrophage progenitors, CFU-GM). On the basis of a comparative examination of chemical, photochemical, and phototoxic properties of crystal violet and other triarylmethane dyes, we have identified interdependencies between molecular structure, and selective phototoxicity toward tumor cells. These structure-activity relationships represent useful guidelines for the development of novel purging protocols to promote selective elimination of residual tumor cells from autologous bone marrow grafts with minimum toxicity to normal hematopoietic stem cells.

Effect of self-association and protein binding on the photochemical reactivity of triarylmethanes. Implications of noncovalent interactions on the competition between photosensitization mechanisms type I and type II.

We have explored the photochemical behavior of cationic triarylmethane dye monomers and dimers free in solution and noncovalently bound to bovine serum albumin (BSA) and examined how self-association and the formation of host-guest complexes involving biopolymers and photosensitizers affect the competition between the photosensitization type I and type II mechanisms. Our results have clearly indicated that tri-para-substituted triarylmethane dyes bind efficiently to albumin as monomers and dimers and, interestingly, that the formation of dye aggregates in aqueous solutions is actually assisted by the protein. Protein-assisted dye aggregation takes place under conditions of high biopolymer loading (high [dye]/[protein] ratios), as attested by the appearance of a hypsochromically shifted absorption band (H-band) that overlaps with the spectral shoulder of the respective dye monomer. As predicted by the molecular exciton theory, the intersystem crossing efficiency in H-type dimers is expected to be higher than in the respective dye monomers, and photoinduced electron transfer events are intrinsically favored in dye aggregates as a result of the physical contact between donor and acceptor. We have found that when triarylmethanes are noncovalently bound to BSA their photoreactivity undergoes a remarkable enhancement, and that the photooxidation mechanism type I is particularly favored in the macromolecular environment. A comparative examination of the behavior of triarylmethane dyes with that of methylene blue have shown that in the case of methylene blue the binding phenomenon also favor the type I mechanism.

Mitochondrial effects of triarylmethane dyes.

The mitochondrial effects of submicromolar concentrations of six triarylmethane dyes, with potential applications in antioncotic photodynamic therapy, were studied. All dyes promoted an inhibition of glutamate or succinate-supported respiration in uncoupled mitochondria, in a manner stimulated photodynamically. No inhibition of N,N,N',N'-tetramethyl-p-phenylenediamine (TMPD) supported respiration was observed, indicating that these dyes do not affect mitochondrial complex IV. When mitochondria were energized with TMPD in the absence of an uncoupler, treatment with victoria blue R, B, or BO, promoted a dissipation of mitochondrial membrane potential and increase of respiratory rates, compatible with mitochondrial uncoupling. This effect was observed even in the dark, and was not prevented by EGTA, Mg2+ or cyclosporin A, suggesting that it is promoted by a direct effect of the dye on inner mitochondrial membrane permeability to protons. Indeed, victoria blue R, B, and BO promoted swelling of valinomycin-treated mitochondria incubated in a hyposmotic K+-acetate-based medium, confirming that these dyes act as classic protonophores such as FCCP. On the other hand, ethyl violet, crystal violet, and malachite green promoted a dissipation of mitochondrial membrane potential, accompanied by mitochondrial swelling, which was prevented by EGTA, Mg2+, and cyclosporin A, demonstrating that these drugs induce mitochondrial permeability transition. This mitochondrial permeabilization was followed by respiratory inhibition, attributable to cytochrome c release, and was caused by the oxidation of NAD(P)H promoted by these drugs.

The efficiency of malachite green, free and protein bound, as a photon-to-heat converter.

Dye assisted laser inactivation of proteins has been found to be a methodology that can achieve high selectivity. Despite the fact that the methodology is successful, knowledge of the detailed inactivation mechanism would allow full optimization of this technique. Here, pulsed-laser photoacoustic calorimetry is used to study the photophysical properties, principally the heat release behavior, of protein bound malachite green. We found that when bound to bovine serum albumin the dye is a good photon-to-heat converter, but approximately 2.6% of the absorbed photon energy (lambda(exc) = 624 nm) is not released as heat in less than 10 mus. This observation suggests that a mechanism other than simple heat-induced inactivation may be the principle process; a long lived excited triplet state of malachite green (or species derived from it) is postulated to play a major role.

Peroxidase-promoted aerobic oxidation of 2-nitropropane: mechanism of excited state formation.

Using sensitized emission, the horseradish peroxidase-catalyzed aerobic oxidation of the toxic pollutant 2-nitropropane to nitrite and acetone is shown to produce the latter in the electronically excited triplet state. In turn, this chemiexcitation implies a hydroperoxide precursor. Taking into account the stoichiometry of the reaction and available isotopic data it is inferred that the hydroperoxide reacts with a second molecule of the substrate (aci form). While triplet acetone formed from isobutanal (enol form) is generated within the enzyme, in the present case triplet acetone is formed in the bulk solution.