Search

We used confocal microspectrofluorometry to investigate intracellular distribution of pirarubicin or THP-DOX in parental K562, CEM, and LR73 tumor cells and their corresponding multidrug-resistant (MDR) strains. Each spectrum of a recorded image was considered as a combination of cell autofluorescence and fluorescence of the drug. In the cytoplasm of parental K562, CEM, and LR73 cells, THP-DOX fluorescence emission profile was similar to that of free drug in aqueous buffer. The (I550nm/I600nm) ratio was 0. 50 +/- 0.1. However, in the cytoplasm of resistant cells the 550-nm band profile was modified. The I550nm/I600nm ratio was 0.85 +/- 0.2 in MDR K562 cells, which is significantly different from the ratio in sensitive cells (p<0.01). This appeared first to correspond to accumulation and self-oligomerization of THP-DOX in cytoplasmic organelles of MDR cells. Transfection of LR73 cells with the mdr1 gene conferred this characteristic on the resistant LR73R cells. Bodipy-ceramide, a trans-Golgi probe, was co-localized with the typical fluorescence emission peak at 550 nm observed in the cytoplasm of MDR cells. This organelle has been shown to be more acidic in MDR cells. Moreover, this specific pattern was similar to that observed when anthracycline is complexed with sphingomyelin. The typical fluorescence emission peak at 550 nm decreased in MDR cells incubated simultaneously in the presence of the drug and quinine, verapamil, or S9788. Growth inhibitory effect and nuclear accumulation of THP-DOX data obtained on LR73R and LR73D cell lines showed that only during reversion of resistance by verapamil and S9788 was an increase of nuclear THP-DOX accumulation observed. Our data suggest that characteristics of molecular environment, such as higher pH gradient or lipid structures, would be potential mechanisms of multidrug-resistance via the sequestration of anthracyclines.

To study the pH gradient status through membranes of acidic vesicles, either in sensitive or in multidrug-resistant living cancer cells, we monitored the fluorescence-emission spectra of acridine orange. Successive stainings with a pH-sensitive dye and AO showed that low-pH organelles were stained red by AO. In these compartments, high AO concentrations are driven by the pH gradient through membrane vesicles. The resulting rise in the dye's oligomeric/monomeric ratio induced an increase in the red/green (655-nm/530-nm) emission intensity ratio. Therefore, the accumulation of AO in acidic organelles was appraised by determination of the contribution of the red emission intensity (R%) in each emission spectrum, using laser scanning confocal microspectrofluorometry. In vesicles of multidrug-resistant K562-R cells, R% is significantly higher (72 +/- 10%) than the value (48 +/- 8%) from K562-sensitive cells (p < 0.001). This result is interpreted as a more important accumulation of AO in acidic cytoplasmic structures of resistant cells, which induces a shift from AO monomers (green emission) to self-associated structures (red emission). Equilibration of the pH gradient through acidic organelles was performed by addition of weak bases and carboxylic ionophores. Ammonium chloride (0.1 mM), methylamine (0.1 mM), monensine (10 microM), or nigericine (0.3 microM) all suppressed the initial difference of local AO accumulation between both cell lines. These agents decreased the red emission intensity for the resistant cell line but not for the sensitive one. The same effects were induced by 50 microM verapamil, a pleiotropic drug-resistance modulator. Our data allow the hypothesis of a higher pH gradient through membranes of acidic organelles, which would be a potential mechanism of multidrug resistance via the sequestration of weak bases inside these organelles.

There is a large discrepancy between the changes in drug accumulation and the changes in drug cytotoxicity that accompany development of anthracycline in multidrug-resistant cells. Moreover, although different molecular targets for anthracyclines such as DNA, cell membranes, or enzymes like topoisomerases could be involved, mechanisms by which these compounds exert their cytotoxic and differentiating effects remain unclear. Studies of correlation between the biological effects of anthracyclines and drug uptake have given conflicting conclusions. For example, a decrease in drug cytotoxicity for different incubation temperatures has been observed in spite of the same intracellular anthracycline amount, suggesting that temperature-dependent cytotoxic effects may be mediated by drug interaction with the cell membrane. What we propose in this review are results of our laboratory which are in agreement with an action mechanism targeted to the nucleus. In fact, we have shown by using microspectrofluorometry, that identical nuclear anthracycline concentration induces the same degree of cytotoxicity, independent of cellular MDR phenotype and the anthracycline structure. Thus, we could acquire information on the mechanisms of drug resistance related to drug transport. We could also give evidence that this accumulation is increased when MDR modulators, such as verapamil and S9788 and cyclosporin A or anthracyclines are used. For clinical applications, our studies have already dealt with nuclear concentration measurements of doxorubicin in leukocytes of treated patients, and in vitro measurements of drug efflux from nuclei of acute leukemic cells and its correlation with P-glycoprotein expression. However, in these studies, there was no correlation between anthracycline nuclear accumulation in vitro and P-glycoprotein expression. In addition, from preliminary results, we have shown that some modulators such as quinine do not significantly increase nuclear accumulation of anthracyclines in MDR cells but are able to restore anthracycline sensitivity. Other authors have recently shown that quinine has a relatively weak effect on cellular doxorubicin accumulation in MDR cells but is able to completely restore doxorubicin sensitivity. They concluded that quinine has essentially intracellular targets involved in drug distribution (cytoplasm --> nucleus) from sequestration compartments. Our data contradict this and we believe that such modulator modifies the molecular environment of anthracyclines and/or their binding to a possible cytoplasmic target leading to different cell death. Thus, we conclude that mechanisms by which anthracyclines induce cell death, and ways by which chemotherapy fails in resistant cells remain complex and are related to more than one target.

Confocal microspectrofluorometry allows the analysis of fluorescent molecules such as anthracylines in isolated living cells. An optical microscope fitted with a phase-contrast 100 X water-immersion objective enables simultaneous observation of the sample, focusing of the laser beam on the selected cell fraction (nucleus) and collection of the fluorescence emitted from the sample. The resulting intranuclear spectra are interpreted according to a quantitative model of the fluorescence spectra of both free and DNA-bound anthracycline. The intranuclear drug concentration can thus be determined. This technique has been applied to blast cells collected in patients with acute leukemia. Leukemic cells are aspirated from bone marrow, separated by Ficoll sedimentation and resuspended in RPMI-1640 containing 10% fetal calf serum and 200 nM tetrahydropyranyl-doxorubicin (THP-DOX). After one hour, 20 cells are analyzed and the mean nuclear drug content is determined (C1). Cells are then resuspended in the same medium but without anthracycline for 3 hours and the mean intranuclear drug concentration is then also determined (C3). From C1 and C3 the retention rate (RR) is calculated. Firstly, the accuracy of the method was checked. In 4 AML patients, two different samples aspirated on the same day were divided into two portions. Thus, two measurements were made on each one (4 values per patient). Coefficients of variation were satisfactory (4, 6, 12, and 12%). Secondly, blast cells collected in patients with AML and ALL at diagnosis or in relapse were studied. P-glycoprotein (P-gp) and CD34 expression was also studied using respectively immunohistochemistry land flow cytometry. Results obtained from the first 21 patients showed that there was no correlation between RR and either P-gp or CD34 expression. This could result from the efflux of THP-DOX by other mechanisms and/or low sensitivity of the staining technique.

A phenotype of resistance to the new vinca alkaloid Navelbine was induced in the J82 human bladder carcinoma cells. The resistance factor of the resistant cell line (J82-NVB) to Navelbine was 17. The resistance phenotype of these cells is not a multidrug-resistance (MDR) phenotype. J82-NVB cells lack overexpression of P-glycoprotein and cross-resistance to MDR drugs like doxorubicin, epipodophyllotoxins or colchicine. Navelbine efflux was similar in sensitive and resistant cells, and resistance could not be explained by a difference of drug accumulation in these two cell lines. The cells were cross-resistant to vinca alkaloids and taxoids whose targets are microtubules. Immunofluorescence study of microtubules showed that depolymerization occured for the same Navelbine concentration in sensitive and resistant cells. This concentration induced growth inhibition in sensitive but not in resistant cells. Moreover, depolymerization induced by Navelbine treatment was reversible, after drug removal, in resistant cells only. This study suggests that J82-NVB cell resistance mechanism involves alterations of microtubule dynamics, allowing recovery of microtubules functions after treatment.

Scanning microspectrofluorometry has been developed to perform the mapping of fluorescence spectra from all locations in a living cell. This new method has been applied to study the molecular environment of rhodamine 123 (R123) in sensitive (K562, CEM) and multidrug-resistant (K562-R, CEM/VLB100) tumor cells. All cells exposed to R123 showed a similar distribution of fluorescence in the perinuclear region. A lower cytoplasmic fluorescence intensity corresponding to a reduced drug accumulation was observed in resistant cells, as expected in the multidrug resistance process. Fluorescence emission spectra of R123 are useful to probe the polarity of the R123 environment. Thus, fluorescence spectra of R123-treated cells have been analyzed as a linear combination of model spectra: R123 in water and R123 in tensio-active Triton X-100. In sensitive cells, emission spectra of R123 underwent a red shift, equivalent to those observed in isolated coupled mitochondria. This suggests the formation of a complex in hydrophobic sites. In contrast, R123 spectra were less shifted in resistant cells, showing two types of both hydrophobic and hydrophilic binding sites. This could be related to an intracellular redistribution of R123 in resistant cells.