Rhodacyanine Dye MKT-077 Inhibits in Vitro Telomerase Assay But Has No Detectable Effects on Telomerase Activity in Vivo

The rhodacyanine dye MKT-077 (formerly known as FJ-776) is a water-soluble delocalized lipophilic cationic dye. It can pass through the hydrophobic barriers of membrane lipid bilayers because of its positive charge and is retained inside mitochondria, which carry a high negative charge. Because cancer cells have a higher mitochondrial membrane potential than normal cells, MKT-077 is preferentially retained in cancer cell mitochondria and is far more toxic to them than to normal cells (1, 2, 3, 4). This selective toxicity for cancer cells has led to the use of MKT-077 in preclinical and clinical cancer therapeutic trials (5, 6). However, the molecular mechanism(s) of MKT-077-induced growth arrest have not been completely understood. Some effects of MKT-077 include reversible impairment of mitochondrial function (1), cross-linking of actin filaments (7), inactivation of telomerase (8), and abrogation of p53-mortalin (hmot-2) interactions leading to reactivation of wild-type p53 function (9). In the present study, we subjected a variety of transformed human cells to MKT-077 and found that most of them were sensitive to this compound, although to a variable extent. MKT-077 inhibited in vitro, but not in vivo, telomerase activity did not cause telomere shortening and was toxic to cells lacking telomerase. Therefore, we suggest that the in vivo toxicity of MKT-077 in cancer cells is not mediated, at least primarily, by its effect on telomerase.

Cells were cultured in DMEM supplemented with 10% fetal bovine serum. MKT-077 [1-ethyl-2-{[3-ethyl-5-(3-methylbenzothiazolin-2-ylidene)]-4-oxothiazolidin-2-ylidenemethyl}] pyridinium chloride or its analogue FJ5002 was added to the culture medium for the time periods indicated in each experiment.

The PCR-based TRAP assay for telomerase activity was used as described (10). Cell lysates were prepared using the 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid detergent lysis method. The protein concentration of the lysates was measured using the Bio-Rad Protein Assay kit. PCR amplification products were separated on a 10% nondenaturing polyacrylamide gel. Quantitative TRAP assays were performed using Telomerase PCR ELISA as described in manufacturer instructions (Roche). All of the lysates were examined in triplicates. Lysis buffer, an ALT cell line, and heat-treated lysate from a telomerase-positive cell line were used as negative controls.

Genomic DNA prepared from 70% confluent cultures was digested with restriction enzymes Hinf I and Rsa I. Digested DNA (0.6 μg, quantitated by fluorometry) was separated by either conventional (1% agarose gel in 0.5 × Tris-borate EDTA buffer) or pulsed-field electrophoresis (CHEF-DRII apparatus; Bio-Rad; with recirculating 0.5 × Tris-borate EDTA buffer at 14°C and a ramped pulse speed of 1–6 s at 200 V for 14 h). The gels were dried, denatured, hybridized to a [γ-³²P]dATP 5′ end-labeled telomeric oligonucleotide probe (TTAGGG)₃, and exposed to Kodak XAR film at room temperature for 18 h as described (11).

Cell lysates (300 μg) were incubated with MKT-077 for 30 min followed by antimortalin antibody (12) for 90 min at 4°C with slow rotation. Mot-immunocomplexes were washed six times with NP40 lysis buffer [20 mm Tris (pH 7.5), 1 mm EDTA, 1 mm EGTA, 0.1 mm phenylmethylsulfonyl fluoride, 150 mm NaCl, and 1% NP40]. Immunocomplexes (1/5 by volume) and the supernatants (1/50 by volume) were analyzed by TRAP assay.

Oncogenic Ras, mot-2, and telomerase are the three known cellular targets by which MKT-077 has been shown to mediate its cellular toxicity in cancer cells. We used MKT-077 to treat a variety of human transformed cell lines (Table 1) that varied in their telomerase, Ras, and p53 status. Each of the cell types, grown at 50–60% confluency, was exposed to MKT-077-supplemented medium, and their cell number and morphology were observed for 1–2 weeks. An approximate EC₅₀ of MKT-077 for each cell type was determined by cell counting, bromodeoxyuridine labeling, and a cell proliferation and viability assay (Cell Proliferation Reagent WST-1; Boehringer Mannheim), typically after 48–60 h of treatment (9). As summarized in Table 1, transformed cells were comparably sensitive to MKT-077 with or without oncogenic Ras, wild-type p53, or telomerase. Furthermore, cells lacking all three of the components (Table 1) were equally sensitive to MKT-077. This suggested that any one of these three cellular targets (oncogenic Ras, telomerase, or mot-2 interactions with wild-type p53) is not essential for the cellular toxicity of MKT-077 and that this may also involve unknown cellular targets/pathways. Therefore, we extended the analysis to assess the effect of MKT-077 on telomerase activity in vitro and in vivo.

An analogue of MKT-077, FJ-5002, has been shown to inhibit telomerase activity in monoblastoid leukemia cells (U937; Ref. 8) suggesting that telomerase inhibition may underlie the selective toxicity of MKT-077 to cancer cells. However, telomerase-negative cells with the ALT mechanism (13) were sensitive to MKT-077 (Table 1). These include SV40-immortalized skin fibroblasts, GM847 (EC50–1.02 μm); 4NQO-transformed liver fibroblasts, SUSM-1 (EC50–1.02 μm); and osteogenic sarcoma cells, Saos-2 (EC50–1.84 μm) and U2OS (EC50–1.02 μm). Thus, we examined the effects of MKT-077 on telomerase activity in vitro and in vivo. Endogenous telomerase activity in HT1080 cells and exogenous activity in GM847/hTERT cells were completely abolished when the lysates were incubated with 0.023 and 0.23 μm MKT-077 (Fig. 1,A). Next, we treated cells with endogenous (HeLa) and exogenous [HCt-6, an immortalized clone of normal neonatal skin fibroblasts (HCA2) stably transfected with hTERT] telomerase activity with different concentrations of MKT-077 and FJ-5002 for 3–5 days. Both of these cell lines were sensitive to MKT-077 (Table 1) and FJ-5002 (data not shown). We examined telomerase activity of these cells treated with various concentrations of MKT-077 and FJ-5002 by quantitative TRAP assay (Roche). Cells treated with all of the concentrations of MKT-077 and FJ-5002 were positive for telomerase activity (Fig. 1,B). Notably, both cell types maintained considerable telomerase activity (Fig. 1,B) at even the highest concentrations of MKT-077 and FJ5002, when they were severely growth arrested, suggesting that telomerase inactivation is not a direct cause of growth arrest. Taken together with the data in Fig. 1 A, these results indicated that although MKT-077 inhibited telomerase activity in vitro, even at low concentrations, it functions differently in a cellular environment and does not directly inactivate telomerase.

MKT-077 may inhibit telomerase activity in vitro by binding (because of its positive charge) G-rich single-stranded DNA (negatively charged), thus blocking its extension by telomerase. A similar mechanism has been proposed for the telomerase inhibiting activity of acridine derivatives (14, 15, 16). We tested this possibility by performing quantitative TRAP assay with MKT-077. A fixed amount of lysate was treated with various concentrations of MKT-077 and subjected to TRAP assay with a constant amount of primers (Fig. 1,C). We found that higher concentrations of MKT-077 inhibited the TRAP assay, whereas lower concentrations were neutral (Fig. 1,C). In agreement with this finding, inhibition of the TRAP assay by MKT-077 was removed when an excess concentration of primer was used (Fig. 1,D). These data suggest that MKT-077 inhibits telomerase activity by binding to and blocking telomerase primers. Accordingly, MKT-077-induced inhibition of the telomerase assay is an indirect effect because of the unavailability to telomerase of primer DNA molecules rather than a direct repression of telomerase function. This inhibition is seen in a cell-free system, where MKT-077 can bind to the telomerase primer to impede its extension by telomerase (Fig. 1,A). However, in a cellular environment, MKT-077 is mainly adsorbed by mitochondria and exists in a different subcellular compartment than the telomeres (1, 2, 3). Therefore, MKT-077 will not primarily target the telomere DNA and will not affect telomerase function per se (Fig. 1 B). In contrast, the mitochondrial pool of mortalin protein (17) may be its major target (9).

GM847 cells lack telomerase and yet have heterogeneous telomeres, ranging from very short to very long, as a result of a mechanism referred to as ALT. These cells were shown to contain ALT-specific nuclear structures, called APBs (18). Growth arrest of GM847 cells induced by confluency or by treatment with a variety of DNA-damaging agents including H₂O₂ cause an increase in APB formation.³ Because these cells were sensitive to MKT-077 (growth arrested at EC₅₀ comparable with telomerase positive cells; Table 1), we examined their APB formation with and without MKT-077 treatment. The number of cells showing APBs was increased from 5–10% in untreated cells to 40–50% in MKT-077 treated cells (Fig. 2,A, left panel). This was also accompanied by an increase in APB size and number per nucleus (Fig. 2 A, right panel). This result is consistent with the observation that MKT-077 causes growth arrest of cells that lack telomerase and with the conclusion that telomerase inactivation is not essential for this process.

We next analyzed telomere length in telomerase-positive and ALT cells after culturing them in the presence of subacute (2.3 nm) concentrations of MKT-077 for 30–35 population doublings. Proliferation in MKT-077-treated cultures was slow, as compared with normal cultures, and became even slower after a few population doublings. However, none of the four cell types tested (HT1080, MCF7, A1698, and GM847) showed any significant decrease in telomere size after 35 population doublings, even though their proliferation was retarded (Fig. 2,B). FJ-5002-treated U937 cells exhibited end to end fusion of their chromosomes as a consequence of telomere shortening (8). Here, however, after MKT-077 treatment, GM847, SW480, and A1698 cells all had no evidence of telomere shortening as assessed both by terminal restriction fragment Southern analysis (Fig. 2B) and telomere fluorescence in situ hybridization (data not shown).

MKT-077 has been shown to bind to mortalin and reactivate wild-type p53 (9). In our growth assays, cells lacking wild-type p53 were also arrested by MKT-077 and showed comparable sensitivity to those cells expressing wild-type p53 (Table 1). The data suggested that wild-type p53 is not mandatory for MKT-077-induced growth arrest. Change in the subcellular distribution of mortalin occurred in NIH 3T3 and MCF7 (wild-type p53) cells when these were arrested by MKT-077 (9). Interestingly, cells that possess mutant p53 and were arrested by MKT-077 also showed change of subcellular distribution of mortalin from nonpancytosolic to the pancytosolic one (data not shown) that is characteristic of normal cells (12, 19, 20). These findings suggest that MKT-077 binding to mot-2 may do more than activate wild-type p53 and that this interaction may mediate the cellular toxicity induced by MKT-077 in cancer cells.

We therefore asked whether the binding of MKT-077 to hmot-2 (9, 21) could explain its potential to block telomere extension by telomerase in vitro. Adding MKT-077 to cell lysates with endogenous (HT1080) and exogenous (GM847/hTERT) telomerase expression led to inhibition of telomere primer extension (Fig. 3). We then immunodepleted mot-2 from cell lysates by three rounds of immunoprecipitation. MKT-077 coprecipitated with mot-2 was apparent from the orange color of the pellets. The pelleted mot immunocomplexes, also containing MKT-077, did not show any telomerase activity (Fig. 3). Interestingly, mot-immunodepleted lysates that were mostly cleared of MKT-077 (as shown by lack of orange color) did display telomere primer extension, showing that the effect of MKT-077 on telomerase activity is reversible. These findings, taken together with the results in Fig. 1, A and B, support the conclusion drawn above that MKT-077 has an indirect effect on telomerase function. Besides the change in cellular distribution of mortalin, its level was seen to increase in cells treated with MKT-077 (9). Additional studies are required to examine if MKT-077-induced mortalin is modified in some way and to elucidate which of its predicted functions (mitochondrial biogenesis, energy generation, chaperoning, and intracellular trafficking) is targeted in MKT-077-induced growth arrest of cancer cells..

Differential cellular distribution patterns have been described for mortalin in normal and immortal/transformed cells (12). Induction of senescence in transformed cells, by introducing a single chromosome (19), chromosome fragments, or genes into the cells (22), or treating with chemicals (20), was accompanied by reversion of mortalin from a nonpancytosolic to pancytosolic subcellular distribution. A similar change in the mortalin staining pattern after treating transformed cells with MKT-077 suggests that their growth arrest may involve induction of a senescence-like phenotype and that this is independent of wild-type p53 or telomerase status.

Mortalin (hmot-2) has a mitochondrial localization and consequently could be a prime target of MKT-077, and its proposed role as a chaperone means that it potentially can interfere in multiple pathways that are critical for the control of cell growth and arrest (23, 24). Elucidation of these additional pathways may facilitate development of effective cancer therapeutics.

We thank Axel Neumann, Thomas Yeager, and Manami Ohtaka for valuable technical help, and Olivia M. Pereira-Smith for the kind gift of human transformed cells.