Antiproliferative effect of methyl-beta-cyclodextrin in vitro and in human tumour xenografted athymic nude mice.

The anti-tumour activity of methyl-beta-cyclodextrin (MEBCD), a cyclic oligosaccharide known for its interaction with the plasma membrane, was investigated in vitro and in vivo and compared with that of doxorubicin (DOX) in the human tumour models MCF7 breast carcinoma and A2780 ovarian carcinoma. In vitro proliferation was assessed using the MTT assay. In vivo studies were carried out using xenografted Swiss nude mice injected weekly i.p. with MEBCD at 300 or 800 mg kg(-1) or DOX at 2 mg kg(-1), during 2 months. Under these conditions, MEBCD was active against MCF7 and A2780 cell lines and tumour xenografts. For each tumour model, the tumoral volume of the xenografted mice treated with MEBCD was at least twofold reduced compared with the control group. In the MCF7 model, MEBCD (800 mg kg(-1)) was more active than DOX (2 mg kg(-1)). After 56 days of treatment with MEBCD, no toxicologically meaningful differences were observed in macroscopic and microscopic parameters compared with controls. The accumulation of MEBCD in normal and tumour tissues was also assessed using a chromatographic method. Results indicated that after a single injection of MEBCD, tumour, liver and kidneys accumulated the highest concentrations of MEBCD. These results provided a basis for the potential therapeutic application of MEBCD in cancer therapy.

Toxicitv and drug resistance are probably the major mechanisms for failure of therapy in cancer. To oxvercome these problems. there is a need to develop antiproliferatiVe agents actix e on other cellular targets. such as the cell membrane. Cvclodextrins (CDs) are known for modifying, the physicopharmaceutical properties of various drugs and components through inclusion complex formation (-Hiravama and Uekama. 1987: Allegre and Deratani. 1994: Bressolle et al. 1996. The inclusion of the drug may have several advantages. such as an increased aqueous solubilitx and stabilitv or a reduction of unwanted side effects (Szetlji. 1994). However. CDs should not be regarded as simple excipients or solubility enhancers because the formation of inclusion complexes might occur with some biological components such as cholesterol and lipid components of the biological and cellular membrane. leading to an enhanced permeability to various molecules (Cho et al. 1995: Hovgaard and Brondsted. 1995: Krishnamoorthy et al. 1995. Thus. we have previously shown that. at non-cvtotoxic concentrations. methyl-[-Bcyclodextrin (MEBCD) was able to potentiate the in vitro anti-tumoral actix itv of doxorubicin (DOX) in several parental sensitix e cancer cell lines and their multidrug-resistant sublines. but x-e also showed that the action of MEBCD on the cell was independent of that of DOX (Grosse et al. 1997a(Grosse et al. . 1998. Several studies confirm that some CDs have their own cellular activiity in terms of interaction with the plasma membrane. permeabilization or haemolytic activitv (Szejtli et al. 1986: Castelli et al. 1989; Kilsdonk et al. 1995). Only a fexx in vivo studies concerning the effect and the toxicitxof CDs injected directlx in human or animal organisms hax-e been performed. x-hereas there are no reports on the use of CDs in cancer (Bellrinner et al. 1995). The toxicolooical effects appeared to be related to the structure of the CD. Non-substituted CDs were found to be highly toxic for the kidneys (Brexster et al. 1990: Bellringer et al. 1995. whereas the toxicitv of substituted CDs varies with the degree and the nature of the substitution (Frijlink et al. 1990. 1991: Giordano. 1991: Flourie et al. 1993. MEBCD is considered an interesting candidate for experimental cancer treatment because of its relatively low toxicity. contrary to diand tri-methyl-$-CD. and its demonstrated activitx in cancer cell lines. In this report. x e present in X itro growth-inhibitory data obtained for MEBCD in two human carcinoma cell lines (MCF7 and A2780) and comparative data on their in xixo anti-tumour activitx in human xenoggrafted mice. In addition. xxe investigated the murine tumoral and tissular distribution of MEBCD.

MEBCD sensitivity in vitro
The human breast adenocarcinoma cell line MCF7 (Soule et al. 1973: Minnaugh et al. 1991 ) was obtained from the American Type Culture Collection (Rockxille. MD. USA). The human ovarian adenocarcinoma A-780 was a kind gift from Dr Canal (Centre Claudius Regaud, Toulouse. France). Exponentially growing cells were used for experinments and all cells were free of mycoplasma. Cells were maintained as suspension cultures at 37°C in a humidified atmosphere containing 5% carbon dioxide in RPMI-1640 medium supplemented with 10% FCS, antibiotics and glutamnine. The viability of the cells was assessed by their ability to exclude 0.5% tiypan blue dye. Cell density in culture flasks was determined by a Coulter counter (Model ZI, Hialeah, FL, USA). To determine the cytotoxic effect of MIEBCD, preconfluent cells from stock cultures (106 cells ml) were treated as follows: adherent tumour MCF7 and A2780 cells were detached with typsin-EDTA (0.25:0.02% w/v) in PBS, washed twice with PBS and resuspended in complete culture medium to obtain single-cell suspension. Cells were counted and then seeded at a final density of 6 x 103 cells per well in 96-well microtitre plates in a final volume of 100 g1. The cells were then allowed to attach for 24 h at 370C. After reconstitution in purified water, MEBCD was diluted in culture medium and was added in various concentrations to wells (O.1-1O mM), then cells were incubated for 96 h at 37°C (atmosphere containing 5k carbon dioxide). The cytotoxicity of MEBCD was quantified by the MTN assay (Alley et al, 1988;Heo et al, 1990;Colangelo et al, 1992). Metabolic reduction of the tetrazolium salt MTI [3-(4,5dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide] leads to fonnation of MTT-formazan. MUT (50 1 of 1 mg ml-' in sterile PBS) was added to each well and plates were incubated for 4 h at 37°C. Blue formazan crystals formed were dissolved in a mixture of isopropanol and hydrochloric acid 1 M (96:4 vlv). The plates were then gently agitated for 10 min and the absorbance measured at 570 nm on a microculture plate reader (Dynatech MR5000, France). The IC 0 values were defined as the concentration of drug resulting in 50% survival of the eated cells compared with controls and were calculated using a program implemented on EXCEL 5.0 software. For each assay, three different experiments were performed in triplicate.
Detemination of MEBCD LD.0 in mice MEBCD LD 5 was determined using female Swiss mice aged S weeks and weighing 22-28 g. Seven groups of six mice (MEBCD at 100, 200, 500, 1000, 1500, 2000 and 3000mg kg-') were injected i.p. weekly for 4 weeks and were then monitored over a span of 2 months.

MEBCD sensitivity in vivo
MCF7 and A2780 cells diluted in RPMI-1640 medium (107 cells in 250 1) were inoculated subcutaneously into the flank area of female nude congenic athymic mice of Swiss strains homozygous for the nude gene (nu+/nu+). All mice were purchased from Iffa Credo (Lyon, France). They were aged 5 weeks and weighed 20-22 g at the start of experiments. These were conducted in accordance with the protocols published by the European Organization for Research and Treatment of Cancer (EORTC) members (Geran et al, 1972). Mice were kept under sterile conditions and were given sterilized food and water. As the MCF7 human cancer cell line requires exogenous oestrogen for efficient tumorigenicity, the mice intended for MCF7 xenografts were, therefore, aseptically implanted subcutaneously in the intrascapular region with 0.72 mg 60-day release 17.oestradiol pellets (Innovative Reseach of America, Toledo. OH, USA) 2 days before injection of tumour cells. The A2780 tumour cell line is oestrogen independent and does not respond to oestradiol stimulation of proliferation in vivo. The growth of MCF7 and A2780 xenografted tumour was monitored every 7 days by measuring the tumour with calipers in three dimensions following the formula length x width x thickness x m16 as described for murine solid tumours (Tomayko and Reynolds, 1989). Four different groups of six mice were used for MCF7 tumours: normal saline (0.9%) as control group, MEBCD in sterile normal saline (0.9%) at 300 or 800 mg kg-', and DOX at 2 mg kg-' as a treatment reference group. Experiments were repeated twice. For A2780 tumours, only two groups were carried out (control and MEBCD at 800 mg kg-'). Drugs or normal saline (250 p1) were injected i.p., weekly, starting 7 days after inoculation of tumour cells, a time when the solid tumours were just palpable and progressed for a total of eight treatments. Tumour response to MEBCD was assessed by comparing the median tumour volume of each MEBCD-treated group with that of the control group.

MEBCD tumoral and tissular distribution
To determine the relative concentrations of MEBCD in MCF7 and A2780 xenografted mice and to contrast tumour drug levels with host tissues, MEBCD at 800 mg kg-' was administered i.p. to mice. Three hours later, the animals were killed by cervical dislocation and tumours, liver, kidneys, lungs, small intestine and brain were removed, washed in PBS, weighed and frozen for later processing. After thawing, tumours and tissues (0.5 g) were quartered and homogenized in 1 ml of purified water with a Tissue Tearor homogenizer (Biospec Products, Bartlesville, OK, USA). Samples were then treated as described in a previous paper (Grosse et al, 1997b). Briefly, after addition of 20 p1 of potassium hydroxide (5 M), alkaline extraction of MEBCD was carried out using 3 ml of chloroform. The lower organic phase was transferred to a fresh tube and evaporated under nitrogen stream. The residue was then reconstituted into 200 1 of mobile phase consisting of a mixture of water and methanol (98:2 v/v) containing 10-4 M of 1naphthol as a fluorophore and injected onto the analytical column. The detection of MEBCD is based on the enhancement of fluorescence of l-naphthol caused by its complexation in the cavity of the CD. The flow rate was 1 ml min-'. A stainless-steel column (300 x 7.5 mm i.dc) packed with exclusion gel TSK 3000 SW was used and fluorimetric detection was performed at excitation and emission wavelengths of 290 and 360 nm respectively. Experiments were carried out in triplicate.

Statistical analysis
In each treated group. data were analysed by ANOVA one-way tests (compared with control values) and differences between mean values at P < 0.05 were considered to be significant.
Britsh Jourmal of Cancer (1998) 78 (9) Determination of MEBCD LD 50 MEBCD LD;,, was determined using doses ranging from 100 to 3000 mg kg-'. injected i.p. in Swiss mice. Injections were performed ex ery week. during 1 month. Doses higher than 2000 mg kg-I were immediatelv lethal. Over a span of 1 month. the LDc dose was found to be close to 1500 mg kg-' week-'. for a cumulative dose of 6000 mg ka-'. No lethality was found in groups treated with MEBCD at 100. 200. 500. and 1000 mg kg-' week-'. Three. five and six deaths were observed in groups treated with 1500, 2000 and 3000 mg kg-' week-' respectively. Doses lower than 1000 mg kg-' week-' appeared to be non-toxic after four injections of MEBCD and for 2 further months of monitoring.

Antiproliferative effect of MEBCD in human xenografts
The anti-tumour activity of MEBCD has been checked in nude mice xenografted with an oestrogen-dependent human breast carcinoma MCF7 model. Tumour volume of mice injected i.p. weeklx either with normal saline (control) or drugs (MEBCD at 300 or 800 mgy kg-' or DOX at 2 mg kg-' as a reference treatment) were estimated each week during 2 months. Tumour growth curses are presented in Figure 3A. Experiments on the four groups of six mice were carried out in duplicate and show. after 5 weeks of treatment. a clear reduction of the tumour x-olume in mice receix ing MEBCD at 300 or 800 mg kg-1. compared with control. The antiproliferative activity of MEBCD was statistically higher than the DOX one. To confirm these data in a non-oestrogen-dependent tumour model. assays were carred out using ovarian A2780 tumour-xenografted mice treated with either MEBCD at 800 mg kg--' or normal saline as control ( Figure 3B). Results obtained are similar to those in the MCF7 model. Results are expressed as the mean tumour volume reduction ratio observed in the different groups and are presented in Table 1. After each of the eight injections. mice treated with MEBCD and controls were in good form and did not lose any body weight. No lethality was obsersed in these two groups. At the end of the studies. the autopsies of mice treated with MEBCD revealed no macroscopic anomalies compared with the control group. Mice treated for more than 6 weeks with DOX exhibited body weiaht losses exceedinr 25% and cumulative toxicity became lethal after sexen treatments (4 deaths out of 18 mice).

Tumour and normal tissues determination of MEBCD
Tumour and normal tissues determination of MEBCD w as performed in mice undergoing a single dose i.p. of MEBCD at 800 mg kg-'. using a high-performance liquid chromatography (HPLC) method. Assay parameters are defined as follow%s: limit of quantitation of MEBCD was 0.5 Ixm. the method of quantitation w-as based on the MEBCD vs internal standard (daunorubicin) peak area ratios and the retention times w ere 4.8 and 11.1 min for daunorubicin and MEBCD respectix ely. Linearity of the method was statisticallv confirmed o-er a range of concentrations of 1-100 -iM. Results are shown in Figure 4. Intratissular MEBCD concentrations are expressed in nmol g-' of tissue. After 3 hours. significant amounts of MEBCD were found in tumour. kidnevs.
small intestine and liver. Concentrations in lunas and muscle were almost undetectable.

DISCUSSION
In cancer chemotherapy. there is a need to develop new drugs. This mav involve identifying and exploiting novel molecular features of cancer cells. One possible new target may be the cell membrane that is the potential site for some new antineoplastic agents like miltefosine (Arancia andDonelli. 1991: Stekar et al. 1995).
Likewise. methylated CDs are cyclic oligosaccharides that have been shown to interact with lipid components of the biological membranes. modifying their fluidity and their permeability. In previous studies. we  actixity of some antineoplastic agents induce( with non-cytotoxic concentrations of MEBCI cell lines (Grosse et al. 1997a(Grosse et al. . 1998. In this st demonstrated also a marked anti-tumour a human breast and ovarian carcinoma cancer cel tions higher than 1.0 mxi (IC< at about 1.5 m-v curves indicate there was no growth inhibition trations up to 1.0 mms. whereas a 100% inhibi achieved at concentrations higher than 2.0 m activity in vivo. athyrmic nude mice have been oestrogen-dependent human breast carcinoma treated every week with MEBCD. DOX or no weeks of treatment, tumour volume in mice tn at 300 and 800 mg kg' was statistically redu control. After eight injections. this reduction c control tumour volume. Moreover. the most n that the antiproliferative activity of MEBCD O least equal to the DOX one. a reference a 56 Figure 4 Relative concentratons of MEBCD in various tissues and tumour xenografts of mice (n = 3). Analysis conditions were as described in Materials and methods commonlv used in chemotherapy regimens against most carcinomas (Muggia and Green. 1991 : Frijlink et al. 1991: Giordano. 1991). As 0.5 ml of plasma is at MEBCD concenrequired for the HPLC method. MEBCD plasma levels were not ition of growth was determined in mice. In contrast. the tropism of MEBCD for the um. To confirm this xenografted tumour is proven by its tumoral accumulation. which inoculated with the is undoubtedly linked to the tumour growth inhibition observed.

MCF7 model and
This antiproliferative activity should be due to the great affinity ormal saline. After 5 of MEBCD for cell membrane lipid components. particularly eated with MEBCD cholesterol. which play a major role in the structure and the funciced compared with tioning of the cell membrane. Thus. MEBCD is able to include ivertook 50% of the cholesterol in its cavity and then to remove it from the cell -emarkable aspect is membranes (Castelli et al. 1989: Szejtli et al. 1986 800 mg kg'l) was at 1995: Hovgaard and Brondsted. 1995: Kilsdonk et al. 1995: mtineoplastic agent Krishnamoorthy et al. 1995).
British Journal of Cancer (1998) 78(9), 1165-1169 C Cancer Research Campaign 1998 In vitro and in vivo antitumoral effect of methyl-f-cyclodextrin 1169 Plasmatic and tissular pharmacokinetics of MEBCD and DOX. alone or in combination, are actually being performed in rabbit to gather further knowledge on biological comportment of MEBCD. Moreover, assays will be carried out to investigate the antiproliferative activity of MEBCD in cell lines and xenograft tumour models overexpressing the MDR phenotype.
The cytotoxicity against two human carcinoma cell lines, the inhibition of MCF7 and A2780 xenografted tumour growth comparable to that obtained with DOX, and the apparent innocuity of doses injected in xenografted nude mice make of MEBCD. alone or in combination with other antineoplastic agents, a potential candidate for cancer therapy.