L1210 cells selected for resistance to methoxymorpholinyl doxorubicin appear specifically resistant to this class of morpholinyl derivatives.

We investigated the mechanism of resistance in murine L1210 leukaemia cells selected after treatment with FCE 23762 methoxymorpholinyl doxorubicin: (MMRDX), a methoxymorpholinyl derivative of doxorubicin active in vitro and in vivo on multidrug-resistant (mdr) cells, currently undergoing phase I clinical trials. The resistant subline obtained after repeated in vitro treatments, L1210/MMRDX, is resistant in vitro and in vivo to all tested methoxymorpholinyl derivatives and to cyanomorpholinyl doxorubicin, but shows resistance to morpholinyl derivatives only in vivo or following their activation with rat S9-liver fractions in vitro. L1210/MMRDX cells are sensitive to classic mdr- and altered topoisomerase (AT)-mdr-associated drugs. These cells do not appear to overexpress the mdr1 gene, nor do they exhibit impaired intracellular drug accumulation and efflux or altered levels of glutathione and glutathione S-transferase. The extent of DNA single-strand break formation and, after microsomal activation, of DNA interstrand cross-links after treatment with MMRDX was similar in the parent and the resistant subline. The mechanism of resistance in L1210/MMRDX cells remains to be identified but may prove a novel one, highly specific for this class of mdr-active anthracyclines.

Treatment with doxorubicin (DX) or daunorubicin commonly selects resistant tumour cells that express the multidrug resistant (mdr) phenotype, characterised by enhanced drug efflux mediated by a high molecular weight membrane glycoprotein (Endicott & Ling, 1989;Hayes & Wolf, 1990;Roninson, 1992). Among the several classes of anthracyclines synthesised in the past 20 years, the morpholinyl anthracyclines are of particular interest since they appear to be active in vitro and in vivo against mdr tumour cells (Watanabe et al., 1988;Coley et al., 1990;Ripamonti et al., 1992).
Methoxymorpholinyl doxorubicin (MMRDX) is a lipophilic compound, able to reach high intracellular levels in sensitive and mdr tumour cells (Grandi et al., 1 990a;Ripamonti et al., 1992). In addition, compounds of this class appear to be effective against tumour cells expressing the altered topoisomerase-mdr phenotype (AT-mdr) (Grandi et al., 1990b). MMRDX is currently under investigation in phase I clinical trials. It was thus of interest to determine whether resistant tumour cells could be selected for resistance to MMRDX following in vitro exposure to the drug and, if so, to identify the mechanisms involved. Murine leukaemia L1210 cells resistant to MMRDX (L1210/MMRDX) were isolated after repeated in vitro treatments with the drug and characterised for their pattern of cross-resistance in vitro and in vivo to a panel of antineoplastic drugs and to selected anthracyclines bearing either the methoxymorpholinyl or the morpholinyl substitution on the 3' position of the sugar moiety.
A clone resistant to MMRDX was selected and propagated in vitro in the absence of the drug. Resistance was stable for at least 1 year. The doubling time was determined by seeding the cells at the concentrations of 5 x 104 and 105 cells ml-' (I ml per well, 12-well plates; Costar, Cambridge, MA, USA). Every 24 h two replicate samples were harvested and the cell number was determined by a ZBI Coulter counter (Hialeah, FL, USA).
For Northern blot analysis 20 gig of total RNA was fractionated on 1% agarose gel containing 6.7% formaldehyde and transferred to nylon membranes (Gene-Screen Plus, NEN, Boston, MA, USA). The filters were hybridised for 16 h at 42°C in 50% formamide, 10% dextran sulphate, I M sodium chloride, 1% SDS, 100 gIg ml-1' of denatured salmon sperm DNA and 106 c.p.m. ml-' denatured 32P-labelled probe. After hybridisation, the filters were washed sequentially in 2 x SSC at room temperature and in 2 x SSC, 1% SDS, at 65°C. The probes utilised were the 1.3-kb EcoRI/ Sall insert of pcDR.3 (Gros et al., 1986) containing the human mdr gene (Gros et al., 1986) and the 1.8-kb PstI insert of the murine action gene. Both probes were 32P-labelled using the multiprime DNA labelling system and [32P]dCTP (Amersham, Aylesbury, UK).
Glutathione-S-transferase (GST) determination Expontentially growing cells were lysed by sonication in distilled water, the cell lysate was centrifuged (10,000 r.p.m. for 15 min) and the supernatant was used for enzyme assay according to the method of Habig and Jakoby (1981) using CDNB as substrate.
Glutathione (GSH) determination Cells were analysed during the exponential phase of growth. GSH total content was measured as described by Tietze (1969).
Alkaline elution DNA damage was detected by the alkaline elution technique described by Kohn et al. (1981). Briefly, [3H]thymidineprelabelled cells were layered and lysed on polycArbonate filters (0.8 ytm pore size, Nucleopore, Pleasanton, CA, USA) at room temperature with 5 ml of lysis solution containing 0.1 M glycine, 0.025 M EDTA and 2% SDS (pH 10). After proteinase K (Merck, Darmstadt, Germany) digestion, alkaline elution was carried out using a solution containing 0.02 M EDTA, 0.1% SDS and tetrapropylammonium hydroxide (Eastman Kodak, Rochester, NY, USA) to give a pH of 12.2. The pumping rate was 0.04 ml min-' and fractions were collected at 180 min intervals for 15 h. 3H-labelled DNA was quantitated by liquid scintillation P counter.
In experiments carried out to identify DNA damage after microsomal activation of MMRDX, cells were incubated with the drug in the presence of the S9 fraction of rat liver (1 mg of protein per ml), glucose 6-phosphate 1 mg ml-' and NADP 2 mg ml-' in a final volume of 1 ml for 1 h before being processed as for DNA single-stand break assay. Intracellular drug accumulation and retention Intracellular drug content was determined in L 1210 and L1210/MMRDX cells treated with 10 and 100 nM MMRDX and incubated at 37°C for up to 4 h.
For drug efflux determination, cells incubated with MMRDX for 1 h were washed in PBS, then resuspended in drug-free medium and reincubated at 37°C. Drug was extracted from the cells with 0.6 M hydrochloric acid-ethanol (1:1 mixture) and samples were analysed in an HPLC system using a C18 reversed-phase column and a spectrofluorimeter as detector (excitation and emission wavelengths were 479 and 593 nm respectively).
The intracellular accumulation of the drug at various time intervals is reported as ng per 106 cells.
In vitro drug sensitivity Exponentially growing L1210 and L1210/MMRDX cells were exposed to various concentrations of drugs continuously for 48 h. The antiproliferative activity of the drugs was evaluated by counting surviving cells with a Coulter counter and results were expressed as IC50 (dose causing 50% inhibition of cell growth in treated cultures relative to untreated controls).
The cytotoxicity of morpholinyl derivatives on L 1210 and L1210/MMRDX cells with and without microsomal activation was determined on cells incubated at 37°C in aerobic conditions for 1 h in the presence of various concentrations of drugs with or without an incubation mixture consisting of 0.33 mg ml-' protein of S9 fraction of rat liver homogenate, 0.33 mg ml-' NADP and 0.16mg ml-' glucose 6-phosphate (Boehringer Mannheim Italia, Milan, Italy).
S9 was prepared according to the method of Hilton and Sartorelli (1970). The incubation was stopped by washing cells with ice-cold RPMI-1640 medium. The cells were then incubated for 48 h in drug-free medium and cell growth was assessed as described above.
In vivo studies Inbred DBA2 and CD2F1 adult female mice (Charles River, Calco, Italy), 2-3 months old, weighing 20 -24 g, were kept under standard laboratory conditions. The L 1210, obtained from the National Cancer Institute Toxicity was evaluated on tumour-bearing mice on the basis of the gross autopsy findings and weight loss.  (Table I), and maintain resistance after > 100 passages in drug-free medium. L1210/MMRDX cells have the same doubling time (9 h) and in vivo tumorigenicity (8-10 days) as the parent line. The levels of GSH (7.76 and 7.83 fmol per cell) and of glutathione S-transferase (100 ± 14 and 111 ± 8 relative units) are also similar in L1210 and L1210/MMRDX cells.
Results obtained comparing the levels of mdr-J mRNA in L1210, L1210/MMRDX and L1210/DX cells a subline 20fold resistant to DX (not shown) as the positive control clearly indicate that L1210/MMRDX cells do not overexpress mdr-J mRNA gene.

Intracellular accumulation and efflux
The kinetics of accumulation and efflux in L1210 and L1210/ MMRDX cells exposed to 10 and 100 nM MMRDX is presented in Figure 2a  Alkaline elution studies The frequency of single-strand breaks (DNA-SSBs) and formation of DNA interstrand cross-links (DNA-ISCs) in response to MMRDX treatment is reported in Table II. No differences were observed between L1210 and L1210/ MMRDX cells. DNA-SSB levels after exposure to 1 ltg ml-' were similar to those found with DX at the same concentration (not shown). The repair of DNA-SSBs appeared very quick as no breaks were detectable after 1 h drug washout in both L1210 and L1210/MMRDX cells.   MMRDX did not cause DNA-ISCs in either cell line when incubated without rat liver S9 fraction; conversely, after I h incubation in the presence of rat liver S9 fraction the number of DNA-ISCs appeared similar in L1210 and L1210/ MMRDX cells (Table II). In L1210 cells 30% and 55% of DNA-ISCs were repaired after 1 and 4 h respectively; in L1210/MMRDX 20% and 64%.
In both cell lines DNA-ISCs were no longer detectable at 24 h. These data suggest that the mechanism of resistance is not related to differences in DNA damage produced by MMRDX.
Pattern of in vitro and in vivo sensitivity to different anti-tumour compounds The activity of different anti-tumour molecules tested in vitro and in vivo on L1210/MMRDX cells in comparison with L1210 cells is reported in Table I. The subline is resistant to MMRDX and sensitive to all other tested drugs, including mdr-inactive drugs such as DX and vinblastine. Moreover, the anti-tumour activity of L-PAM, cisplatin and mitomycin C is markedly higher in the resistant than in the parent line.
Pattern of in vitro and in vivo sensitivity to morpholinyl anthracyclines It is reported that most morpholinylanthracyclines are activated to highly cytotoxic metabolites when administered in vivo and in vitro in the presence of liver microsomes (Lau et al., 1989;Duran et al., 1991;Lewis et al., 1992). The relative sensitivity of L1210/MMRDX cells to a series of molecules of the same chemical class (Figure 1) was thus assayed with and without rat liver homogenate (S9) ( Table  III). L1210/MMRDX cells are resistant to all tested methoxymorpholinyl anthracyclines with resistance indexes (RI) ranging between 10.5 and 4.2. The RI values are unaltered after treatment with S9. Conversely, L1210/MMRDX cells are sensitive to the analogues bearing the morpholinyl instead of the methoxymorpholinyl group but become resistant after treatment with S9. In particular, the cytotoxicity of MRDX and 4'-epi-MRDX is increased 50-fold after microsomal activation. The only cyanomorpholinyl derivative tested, CN-MRDX, was inactive, independently of the treatment with microsomes, which also did not augment its cytotoxic activity.
Results reported in Table IV indicate that, when administered in vivo to mice bearing ascitic L1210 and L1210/MMRDX cells, all tested compounds have antitumour activity against the sensitive line, and are inactive against the resistant one. These results confirm the data obtained in vitro, which indicate that L1210/MMRDX cells are resistant to methoxymorpholinyl derivatives, but only show resistance to morpholinyl derivatives after activation in the presence of liver microsomes.

Discussion
We describe the isolation and characterisation of a murine L1210 cell line selected for resistance to MMRDX, a new anthracycline derivative undergoing phase 1 clinical studies. The mode of action of MMRDX appears to differ from that of anthracyclines (Ripamonti et al., 1992), since it is active against mdr and AT-mdr cells (Grandi et al., 1990b), and forms DNA-ISCs when tested in the presence of liver microsomes Lau et al., 1991).
Our results using L1210 cells indicate that exposure to MMRDX selects a stable cell population specifically resistant in vitro and in vivo to the selecting agent and compounds of the same chemical class, the morpholinyl anthracyclines.
L1210/MMRDX cells were found to be sensitive to antitumour drugs associated with the classic mdr phenotype and to topoisomerase II inhibitors. No overexpression of the mdrl gene, or any alteration in drug accumulation or efflux was identified in these resistant cells. The levels of GSH and GST were found to be unchanged in L1210/MMRDX cells, suggesting that GSH-mediated detoxification is not the basis of resistance to this compound.
As regards the DNA damage induced by MMRDX treatment, we found that the number of DNA-SSBs and, after microsomal activation, of DNA-ISCs was similar in L1210 and L1210/MMRDX cells.
Although these data suggest that the mechanism of resistance selected by treatment with MMRDX is not related to differences in DNA damage or in repair mechanisms, the drug concentrations required to obtain a number of DNA lesions detectable by alkaline elution are far greater than the minimal cytotoxic concentrations. Therefore, we cannot exclude the possibility that the different drug sensitivity is due to a more efficient repair of a lower number of DNA lesions.
The pattern of cross-resistance to morpholinylanthracyclines is an interesting finding: in fact, all compounds bearing the methoxymorpholinyl group are inactive in vitro and in vivo against L1210/MMRDX cells, as well as cyanomorpholinyl doxorubicin, whereas the morpholinyl derivatives are inactive only in vivo, and in vitro when tested in the presence of rat liver S9 fraction. Such results can be interpreted assuming that: (a) the active metabolite(s) of MMRDX and MRDX have similar modes of action and mechanisms of resistance; (b) the mechanism of resistance to MMRDX and its active metabolite is the same, as demonstrated by the evidence that the RIs before and after metabolic activation of MMRDX are equivalent; (c) CN-MRDX is inactive against L1210/MMRDX cells without requiring metabolic activation. Therefore is seems plausible that MMRDX, as well as its active metabolite(s), MRDX active metabolite(s) and CN-MRDX act by the same mechanism.
The anti-tumour activity of alkylating agents such as L-PAM, BCNU, cisplatin and mitomycin C against L1210/ MMRDX cells suggests that, if MMRDX activity is associated with alkylating species, its mechanism of interaction with macromolecules is different from that of the classical alkylating agents.
In conclusion, L1210/MMRDX is a cell line which may be a useful tool for investigating the mechanism of in vitro and in vivo resistance to MMRDX and its analogues.
The results obtained so far strongly support the view that this drug is profoundly different from previously investigated anthracyclines and should not be considered as one of the various DX analogues, but as a new type of anti-tumour drug.
The contribution of Massimo Broggini, Giorgio Belvedere, Giovanna Tagliabue and Maurizio D'Incalci is partially supported by the CNR project 'Target Project on Biotechnology and Bioinstrumentation' and by the Italian Association for Cancer Research.