Skip to main content
NCBI home page
Journal of Cancer Research and Clinical Oncology logoLink to Journal of Cancer Research and Clinical Oncology
. 2003 Dec 4;130(3):178–186. doi: 10.1007/s00432-003-0516-9

In vivo establishment and characterization of a paclitaxel-resistant human ovarian cancer cell line showing enhanced growth properties and drug-resistance only in vivo

Kaoru Okugawa 1, Hiroaki Kobayashi 1,✉, Toshio Hirakawa 1, Takanori Sonoda 1, Tomonori Ogura 1, Hitoo Nakano 1
PMCID: PMC12161817  PMID: 14655049

Abstract

Purpose

In order to establish a more faithful model of clinically recurrent ovarian cancer after paclitaxel-based chemotherapy, a paclitaxel-resistant human ovarian cancer cell line was established in vivo, and its biological profiles were compared with the conventional in vitro established drug-resistant cell line.

Methods

An in vivo paclitaxel-resistant subline (OM1/Tvivo) was established from the parental human ovarian cancer cell line (OVMG1) by repeated paclitaxel administration into tumor-bearing mice. As a control, the in vivo drug-sensitive subline (OM1/Cvivo) was made in the same manner, without paclitaxel. An in vitro paclitaxel-resistant subline (OM1/Tvitro) was established by exposure to stepwise increased concentrations of the drug in a cell culture medium. Chromosomal analysis, evaluation of growth, invasiveness and metastasis, in vivo and in vitro drug sensitivity, and a pharmacokinetic study were performed.

Results

Both in vivo sublines confirmed their human origin by G-band chromosomal analysis and showed a similar cell growth rate in cell culture. As for in vivo tumor growth, OM1/Tvivo showed enhanced tumor growth property compared with OM1/Cvivo, while OM1/Tvitro lost tumorigenicity. Both OM1/Tvivo and OM1/Cvivo sublines as well as their parental OVMG1 could not form either invasive or metastatic lesions. Compared with the OM1/Cvivo subline, the OM1/Tvivo tumor showed stable drug-resistance and lower drug distribution after paclitaxel administration into mice, whereas cultured OM1/Tvivo cells lost both completely. On the other hand, an unreasonably higher level of drug-resistance and lower drug concentration was detected in vitro only in OM1/Tvitro cells after exposure to the drug in a culture medium.

Conclusions

These results suggest that the in vivo established paclitaxel-resistant cell line, rather than the conventional in vitro established cell line, is a suitable and faithful model for clinically recurrent tumors showing transformed aggressiveness. The in vivo specific drug-resistant mechanism should involve an interaction between the tumor and host stromal tissue rather than only changes in cellular drug sensitivity. The present study is probably the first report of an in vivo established paclitaxel-resistant human ovarian cancer cell line, and the elucidation of such an in vivo drug-resistance mechanism may be clinically important in preventing or overcoming acquired drug-resistant ovarian cancers recurring after paclitaxel-based chemotherapy.

Keywords: Paclitaxel-resistance, Human ovarian cancer, In vivo established drug-resistant cell line

Introduction

Many drug-resistant sublines of human cancer have been established to elucidate the drug-resistant mechanism. In these studies, most of the sublines have been established in vitro in a cell culture system by repeated or prolonged exposure to increasing doses of the antineoplastic agent (Teicher et al. 1986). However, these in vitro selected resistant cell lines usually show strongly suppressed metastatic and/or tumorigenic properties in vivo (Bashir et al. 1994; Ganapathi et al. 1987; Scotlandi et al. 1996). This phenomenon of retarded malignant phenotypes expressed by in vitro established drug-resistant cell lines has been described as ‘reverse transformation’ (Biedler and Spengler 1994), which seems to be incompatible with what is clinically observed as recurrence in cancer patients after chemotherapy since their lesions tend to show increased potentials in metastasis and/or tumor growth (Kamura 1996). Therefore, the drug-resistant cell lines established by ‘classic’ in vitro serial selection procedures do not seem to be suitable models for accurately analyzing the biological behavior of clinically recurrent tumors.

Some of these problems might be overcome by in vivo establishment of drug-resistant cell lines which exhibit enhanced malignant potentials as well as drug-resistance (Antoine et al. 1988). By the administration of cisplatin into tumor-bearing animals, we previously established, in vivo, drug-resistant murine cell lines showing enhanced invasive and metastatic properties, and reported that they were considered to be more faithful and useful models for simulating biological aggressiveness of clinically recurrent cancers rather than conventional ‘classic’ in vitro established drug-resistant cell lines (Mitsumoto et al. 1998).

Paclitaxel, an anti-microtubule agent isolated from Taxus brevifolia, has been shown to demonstrate clinical efficacy in ovarian cancer (Einzig et al. 1992). This agent binds to and stabilizes microtubules, and consequently, induces mitotic arrest and apoptotic cell death (Manfredi et al. 1982). Paclitaxel-based chemotherapy produced higher response rates and better prognosis, and is considered to be the international standard regimen against ovarian cancer (McGuire et al. 1996). However, acquired-resistance to paclitaxel has become a serious clinical issue with its increased use. With the recent demand to analyze the biological behavior of paclitaxel-resistant tumors and uncover the paclitaxel-resistance mechanism to improve the therapeutic efficacy against recurrent ovarian cancer, it has become necessary to establish and analyze paclitaxel-resistant ovarian cancer cell lines which properly and faithfully simulate the clinical situation.

We could not find any studies concerning in vivo established paclitaxel-resistant cell lines, although in vivo established cell lines against other drugs (cisplatin, alkylating agents, and vinca alkaloids, etc.) were reported (Abe et al. 1996a; Abe et al. 1996b; Starling et al. 1990; Teicher et al. 1990). In this study, we established paclitaxel-resistant human ovarian cancer cell line by in vivo drug exposure and analyzed its biological properties in order to understand the clinical behavior of recurrent tumors after paclitaxel chemotherapy.

Materials and methods

Parental cell line and cell culture

A human ovarian cancer cell line of OVMG1 was established in our department from a surgical specimen of serous adenocarcinoma of the ovary. This cell line was established (Kobayashi et al., in preparation) as a nude mouse-transplantable human cancer cell line, by means of co-injection of a primary cell culture with matrigel, which is a reconstituted basement membrane, known to enhance tumorigenicity of various human cancer cells in nude mice (Fridman et al. 1990). The parental cell line and its sublines established in this study were maintained in vitro at 37 °C, in a humidified 5% CO2 atmosphere, in RPMI1640 (NIPRO, Osaka, Japan) containing 10% fetal bovine serum, 100 µg/ml streptomycin, 100 U/ml penicillin G, and 2.5 µg/ml amphotericin B.

Animals

Female athymic BALB/c nu/nu mice between 6–8 weeks old (Charles River Japan, Atsugi, Japan) were used throughout the experiments. The mice were maintained in a laminar-flow cabinet under specific-pathogen-free conditions, while receiving standard feed and water ad libitum. Our experiments were reviewed by the Committee of Ethics in Animal Experiments in the Graduate School of Medical Sciences, Kyushu University, and were carried out under the Guidelines for Animal Experiments in the Graduate School of Medical Sciences, Kyushu University, and the Law (No. 105) and the Notification (No. 6) of the Government.

Paclitaxel

Paclitaxel (Taxol) was kindly provided by Bristol Pharmaceuticals K.K. (Tokyo, Japan). It was dissolved in dimethyl sulfoxide to make a stock solution of 10 mM and stored at −20 °C until used for in vitro culture. As for in vivo experimental use, paclitaxel was dissolved in equal volumes of absolute ethanol and Cremophor EL to adjust the drug concentration to 18 mg/ml, and this solution was diluted 1:8 with sterile physiologic saline just before intravenous (i.v.) administration.

Establishment of paclitaxel-resistant sublines

The paclitaxel-resistant subline was established, in vivo, by the previously reported procedure (Mitsumoto et al. 1998; Teicher et al. 1990). Briefly, a single cell suspension containing 2×106 OVMG1 cells was subcutaneously (s.c.) transplanted into a mouse. When the average diameter of tumors reached 10 mm, 40 mg/kg paclitaxel was injected i.v. through the tail vein. After 24 h, the tumor was excised, minced, and re-transplanted s.c. into another fresh mouse. After six rounds of this procedure, over a 6-month period, the final tumor-derived cells were established and maintained in a cell culture as the in vivo established paclitaxel-resistant subline, designated as OM1/Tvivo. The parental OVMG1 cells were passaged six times in the same manner using a vehicle (i.e., drug-free buffer) administration instead of paclitaxel, and established as the in vivo-passaged control cell lines OM1/Cvivo.

The in vitro established paclitaxel-resistant cell line, designated as OM1/Tvitro, was developed in culture dishes, by exposure to paclitaxel, with stepwise increased concentrations, until it could grow in the medium containing 20 nM paclitaxel. Before experimental use, OM1/Tvitro cells were maintained in a paclitaxel-free culture medium and subcultured at least three times.

Chromosomal analysis of in vivo established sublines

Cytogenetic analysis using G-banding was performed on the 18th-passaged culture cells from both in vivo established sublines of OM1/Tvivo and OM1/Cvivo, to demonstrate human epithelial cancer cells of the ovary.

Evaluation of growth, invasive, and metastatic properties

To evaluate in vitro cell growth property, a single cell suspension containing 5×104 cells of OM1/Tvivo, OM1/Cvivo, OM1/Tvitro or OVMG1 was seeded into each well of 6-well culture plates, in a complete medium. The number of cells was monitored every day by a Coulter counter (Coulter Electronics, England, UK). The cells were counted in triplicate wells and the doubling time of each line was determined from the cell growth curve. For evaluation of in vivo tumor growth property, a single cell suspension containing 2×106 cells of each line was transplanted s.c. into the right flank of each mouse. Tumor growth was followed by measuring the length (L; longest dimension) and the width (W; distance perpendicular to and in the same plane as the length) with a caliper once a week. Tumor volume (mm3) was calculated by LW2/2. Tumor growth rate was compared by the tumor doubling time calculated from the growth curve of the tumor.

As for comparison of invasiveness, 2×106 cells of each cell line were transplanted s.c. into each mouse. When the average diameter of each tumor reached 10 mm, the tumor was excised and histologically evaluated regarding its invasiveness by hematoxylin and eosin staining.

The spontaneous metastatic property was evaluated as we described before (Mitsumoto et al. 1998). Briefly, when the mean thickness of the foot reached 8 mm in each subline group, after the intrafootpad (i.f.p.) inoculation of 106 cells, the tumor-bearing legs were amputated at the middle of the femurs. As for experimental metastasis, 106 cells of each subline were injected i.v. through the tail vein. After i.f.p. or i.v. injection, the inoculated mice were monitored until 6 months later, and finally the lung and other organs were checked for metastasis by autopsy.

Evaluation of in vivo drug sensitivity

When the average diameter of the tumor reached 6 mm after the s.c. inoculation of each cell line by the aforementioned procedure, 40 mg/kg paclitaxel (treated group) or drug-free vehicle (untreated group) was injected i.v. through the tail vein. The tumor volume was monitored as mentioned above, and in vivo drug sensitivity was compared by the tumor growth curve after the treatments, and by the percent ratio of the tumor volume calculated as (mean tumor volume of treated group)/(mean tumor volume of untreated group) × 100.

Evaluation of in vitro drug sensitivity by a WST-1 assay

The sensitivity of the cultured tumor cells to paclitaxel was determined by WST-1 assay (Dojindo Laboratories, Kumamoto, Japan) as previously reported (Ishiyama et al. 1996). A single cell suspension containing 2×104 cells in 100 µl of RPMI1640 medium, without phenol red, was seeded into each well of 96-well culture plates. After overnight preincubation at 37 °C with 5% CO2, 100 µl of the medium containing various concentration of paclitaxel was added. After 72 h incubation, 20 µl WST-1 solution containing 66 µg WST-1 (a monosodium salt of 2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium), and 1.4 µg 1-methoxy-5-methylphenazinium methosulfate was added to each well. After another 4 h incubation, the absorbance (abs.) at 450 nm was measured with a microplate reader (Bio-Rad Laboratories, Tokyo, Japan). The percentage of viable cells was calculated as (abs. in treated well − abs. in cell free well)/(abs. in untreated well − abs. in cell free well) × 100. The mean viability was calculated in triplicate. The concentration of 50% inhibition (IC50) was defined as the paclitaxel concentration required for 50% reduction of viable cells and obtained from the graphical plots for cell viability.

Evaluation of in vitro drug sensitivity by a colony-forming assay

Cells were suspended in 6-well culture plates at various cell concentrations (103, 5×102, 102, and 50 cells/well for untreated controls; 5×103, 103, 5×102, and 102 cells/well for the 5 nM paclitaxel-treated group; 104, 5×103, 103, and 5×102 cells/well for the 50 nM drug-treated group; and 5×104, 104, 5×103, and 103 cells/well for the 500 nM drug-treated group). After overnight incubation, cultures were exposed to various concentrations of paclitaxel (0 nM, 5 nM, 50 nM, and 500 nM) for 24 h at 37 °C, and thereafter the drug was removed and each well was washed with cold phosphate buffered saline (PBS) and re-incubated by a complete medium. Ten days later, the plates were fixed with Carnoy’s fixative (ethanol:chloroform:glacial acetic acid = 6:3:1) and stained with crystal violet. The number of visible colonies, consisting approximately of more than 50 cells, were counted and the plating efficiency (P.E.) was calculated as (number of the colonies)/(number of the seeded cells). The surviving fraction was calculated as (P.E. in treated well)/(P.E. in untreated well). The mean value±S.E. was calculated in triplicate.

Drug concentration in tumor tissues

Tumor cells of OM1/Tvivo and OM1/Cvivo were inoculated s.c. into a mouse as mentioned above. When the average diameter of the tumor reached 6 mm, 40 mg/kg paclitaxel was injected i.v. through the tail vein. Three hours, 15 h, 24 h, and 48 h after the injection, tumors were enucleated and stored at −80 °C until pharmacokinetics evaluation. Each tumor was homogenized and analyzed by solid-phase extraction high performance liquid chromatography (SBS, Sagamihara, Japan). The paclitaxel concentration in tumor tissue was calculated as (total amount of paclitaxel)/(wet tumor weight).

Drug concentration in cultured cells

The subconfluently cultured cells of OM1/Tvivo, OM1/Cvivo, and OM1/Tvitro were exposed to 100 nM paclitaxel for 2 h. After rinsing twice with cold PBS, cells were harvested into tubes and cell pellets were collected after centrifugation (2,700 rpm, 5 min). Paclitaxel concentration in cultured cells was indicated as the total amount of paclitaxel per 106 cultured cells, using the aforementioned method of chromatography.

Statistical analysis

Data were subjected to Student’s t-test, using the statistical software StatView version 4.5 (Abacus Concepts, Berkeley, Calif., USA). The significance level was set at P<0.05.

Results

Chromosomal analysis of in vivo established sublines

As shown in Fig. 1, both in vivo established sublines of OM1/Tvivo and OM1/Cvivo showed similar chromosome abnormalities, suggesting that these two cell lines derived from the common ancestor human cells. In addition, double minutes or homogenously staining regions that often appeared in drug-resistant cancer cells failed to be detected. Their composite karyotype was as follows; OM1/Tvivo: 45~47, XX, +X, inv (1) (p32q42), add (2) (q13), -3, add (3) (q21), del (4) (p12), add (5) (q?31), del (6) (q?25), add (7) (q11.2), add (9) (q?32), add (9) (q?34), -10, add (10) (p11.2), add (10) (q26), +12, add (12) (q24.1), -13, add (13) (q14), -17, -19, +mar1, +mar2 [cp20]; OM1/Cvivo: 46~47, X, -X, inv (1) (p32q42), add (2) (q13), -3, add (3) (q21), del (4) (p12), add (5) (q?11.2), add (7) (q11.2), add (9) (q?32), add (9) (q?34), add (10) (p11.2), add (10) (q26), add (12) (q24.1), add (13) (q14), -17, -19, +mar1, +mar2, +mar3, +mar4 [cp20]. No obvious numerical or structural changes associated with paclitaxel-resistance were found when comparing karyotypes of OM1/Tvivo cells (Fig. 1a) and OM1/Cvivo (Fig. 1b).

Fig. 1a, b.

Fig. 1a, b

Open in a new tab

Composite karyotypes of a OM1/Tvivo and b OM1/Cvivo analyzed by a G-band technique. a: OM1/Tvivo: 45~47, XX, +X, inv (1) (p32q42), add (2) (q13), -3, add (3) (q21), del (4) (p12), add (5) (q?31), del (6) (q?25), add (7) (q11.2), add (9) (q?32), add (9) (q?34), -10, add (10) (p11.2), add (10) (q26), +12, add (12) (q24.1), -13, add (13) (q14), -17, -19, +mar1, +mar2 [cp20]; b: OM1/Cvivo: 46~47, X, -X, inv (1) (p32q42), add (2) (q13), -3, add (3) (q21), del (4) (p12), add (5) (q?11.2), add (7) (q11.2), add (9) (q?32), add (9) (q?34), add (10) (p11.2), add (10) (q26), add (12) (q24.1), add (13) (q14), -17, -19, +mar1, +mar2, +mar3, +mar4 [cp20]

Evaluation of growth, invasive and metastatic properties

The cell doubling time in monolayer cultures of OM1/Tvivo, OM1/Cvivo, OM1/Tvitro, and parental OVMG1 cells was 26.1 h, 23.7 h, 27.8 h, and 26.6 h, respectively (Table 1), which indicates that both drug-resistant sublines of OM1/Tvivo and OM1/Tvitro did not show significant changes in the in vitro cell growth property.

Table 1.

Biological properties of sublines compared with their parental OVMG1 cell line. (ND not determined because of poor tumorigenicity)

Subline Profile of subline In vitro cell growth ratea In vivo tumor growth rateb In vivo invasivenessc In vivo metastasisd
OM1/Tvivo In vivo-established drug resistant Unchanged (26.1 h)a Increased (11.1 days)b Unchanged (none) Unchanged (none)
OM1/Cvivo In vivo-established control Unchanged (23.7 h) Unchanged (16.8 days) Unchanged (none) Unchanged (none)
OM1/Tvitro In vitro-established drug resistant Unchanged (27.8 h) ND ND ND
Open in a new tab

a Cultured cell number was monitored after cell seeding. Parenthesis, cell doubling time

b Tumor volume of s.c. inoculated tumors was monitored. Parenthesis, tumor doubling time

c Invasiveness of s.c. inoculated tumors was evaluated when their mean diameters reached 10 mm

d Spontaneous and experimental metastasis were checked after s.c. and i.v. inoculation, respectively

As shown by tumor growth curves (open symbols in Fig. 2a), the in vivo growth property of OM1/Tvivo was significantly higher than that of OM1/Cvivo. The tumor doubling times of OM1/Tvivo and OM1/Cvivo were 11.1 and 16.8 days, respectively (Table 1). Although the parental OVMG1 cell line showed a similar tumor growth curve to that of OM1/Cvivo (data not shown), in vitro established drug-resistant subline of OM1/Tvitro could not form subcutaneous tumor in the animals.

Fig. 2a, b .

Fig. 2a, b

Open in a new tab

a In vivo tumor growth of OM1/Tvivo untreated group (○), OM1/Tvivo treated group (●), OM1/Cvivo untreated group (△) and OM1/Cvivo treated group (▲). A single cell suspension containing 2×106 cells was transplanted s.c. into mice. Paclitaxel (40 mg/kg) was administrated i.v. on Day 0. Each point of the growth curve shows the mean tumor volume±SE calculated from four to eight tumors as described in Materials and methods. OM1/Tvitro cells showed poor tumorigenicity even without the treatment; b Ratio of tumor volume of OM1/Tvivo (●) and OM1/Cvivo (▲) after paclitaxel. The ratio of tumor volume was calculated by the volume of treated tumor divided by that of the untreated tumor

Because of its poor tumorigenicity, we could not evaluate either the invasive/metastatic properties or the tumor growth curve in the case of the OM1/Tvitro subline. None of other cell lines of OM1/Tvivo, OM1/Cvivo or parental OVMG1 could invade beyond the capsule of subcutaneous tumor, even with 10 mm in tumor diameter (Table 1). Similarly, none of these three cell lines made spontaneous metastatic lesions, even when observed 6 months after amputation of the tumor-bearing leg. Even when trying to create experimental metastases, which is generally considered to produce lung colonies easily by injection of tumor cells into the tail vein, no metastatic colonies were observed even at 6 months after i.v. inoculation of these three lines (Table 1).

In vivo drug sensitivity evaluated by tumor growth retardation after paclitaxel treatment

To evaluate in vivo drug sensitivity of OM1/Tvivo and OM1/Cvivo cells, we monitored the s.c. tumor volume after the i.v. administration of paclitaxel. The tumor growth curves of both sublines, with or without paclitaxel treatment, are shown in Fig. 2a. The tumor doubling time of OM1/Cvivo was prolonged from 16.8 days to 29.3 days by the treatment, while that of OM1/Tvivo remained unchanged (from 11.1 days to 10.6 days). The drug-sensitive OM1/Cvivo tumor showed obvious growth retardation after a 40 mg/kg paclitaxel injection, while no tumor growth inhibition was observed in the case of the drug-resistant OM1/Tvivo tumor.

For easy evaluation of in vivo paclitaxel sensitivity, the ratio of the treated tumor volume divided by the untreated tumor volume was plotted (Fig. 2b). After paclitaxel administration, the size of the OM1/Cvivo tumor kept on decreasing and reduced its volume by 45% of its untreated tumor on Day 35. On the other hand, the OM1/Tvivo tumor diminished its volume by only 20% until Day 7 after paclitaxel treatment, and thereafter increased rapidly to overtake the volume of its untreated tumor before Day 35. The stability of in vivo paclitaxel-resistance in OM1/Tvivo cells was confirmed by the repeated in vivo drug sensitivity test after maintenance in a drug-free culture medium over 2 months (data not shown).

We could not evaluate the in vivo drug sensitivity of OM1/Tvitro subline because of its poor tumorigenicity.

In vitro drug sensitivity evaluated by a WST-1 assay and a colony-forming assay

Using the WST-1 assay, we evaluated the in vitro drug sensitivity shown by cell viability after 72 h exposure to paclitaxel (Fig. 3a). Exposure to 10 nM paclitaxel produced no difference in cell viability among OM1/Tvivo, OM1/Cvivo, and OM1/Tvitro, but in the case of 50 nM or more drug concentrations, both OM1/Tvivo and OM1/Cvivo showed an obvious decrease in cell viability with similar decay curves. On the other hand, OM1/Tvitro showed a mild decrease in cell viability after paclitaxel exposure. The IC50 of paclitaxel for OM1/Tvivo, OM1/Cvivo, and OM1/Tvitro were 81.0 nM, 79.6 nM, and 1,560 nM, respectively. The relative paclitaxel-resistance of OM1/Tvitro compared with OM1/Tvivo and OM1/Cvivo was 19.3-fold and 19.6-fold, respectively.

Fig. 3.

Fig. 3

Open in a new tab

In vitro drug sensitivity of OM1/Tvivo (●), OM1/Cvivo (▲) and OM1/Tvitro (■) evaluated by a WST-1 assay or b colony-forming assay. a: Cells were incubated in 96-well culture plates with various concentrations of paclitaxel for 72 h. WST-1 solution was added and cell viability was determined by the measurement of 450 nm absorbance. Each point shows the mean±S.E. calculated in triplicates; b: Ten days after 24 h-exposure to various concentrations of paclitaxel, the plating efficiency (P.E.) was determined by counting the colony number. Surviving fraction was calculated as (P.E. in treated well)/(P.E. in untreated well)

Using the colony-forming assay, we evaluated the in vitro drug sensitivity indicated by the surviving fraction of colony formation, after 24 h exposure to paclitaxel (Fig. 3b). Both OM1/Tvivo and OM1/Cvivo sublines showed similar curves of impaired colony forming ability after exposure to 5–500 nM paclitaxel. On the other hand, the OM1/Tvitro subline showed a higher drug-resistance, characterized by a much shallower slope in the curve of the surviving fraction versus drug dose, and its colony-forming ability was not impaired, even against 500 nM paclitaxel. Compared with the surviving fractions of both OM1/Tvivo and OM1/Cvivo, OM1/Tvitro showed approximately 7,500-fold higher resistance against 500 nM paclitaxel.

As for the parental OVMG1 cell line, a similar level of drug sensitivity with OM1/Tvivo was observed, in vitro, by both WST-1 and colony-forming assays (data not shown).

Drug concentration in tumor tissues

Figure 4a shows the concentrations of paclitaxel in the tumor tissues of OM1/Tvivo and OM1/Cvivo after i.v. administration of 40 mg/kg paclitaxel. We could not evaluate the drug concentration of OM1/Tvitro-derived tumor because of its severely impaired tumorigenicity. The concentration of paclitaxel in OM1/Tvivo tumors was significantly lower than that in OM1/Cvivo tumors at both 24 h and 48 h after drug injection (P <0.01). At 48 h after paclitaxel administration, the OM1/Tvivo tumor contained approximately one-third of the drug concentration, compared with the case of OM1/Cvivo.

Fig. 4.

Fig. 4

Open in a new tab

Paclitaxel concentration in the a tumor tissues or the b cultured tumor cells. Paclitaxel concentration was measured by solid-phase extraction high-performance liquid chromatography. a: Subcutaneous tumors were enucleated after 3 h, 15 h, 24 h, and 48 h after the i.v. injection of 40 mg/kg paclitaxel. * P<0.01; b: Intracellular concentration of paclitaxel after 2 h exposure of 100 nM paclitaxel. Each point is the mean±S.E. calculated in triplicates. ●: OM1/Tvivo, ▲: OM1/Cvivo, ■: OM1/Tvitro

Drug concentration in cultured cells

Figure 4b shows concentrations of paclitaxel in the cultured cells of OVMG1-derived sublines. No difference was observed in paclitaxel levels between OM1/Tvivo and OM1/Cvivo cells in spite of the evident difference existing in the case of subcutaneous tumors. OM1/Tvitro cells contained approximately 12% and 4% of paclitaxel at 2 h and 4 h after the beginning of paclitaxel exposure, respectively, compared with the OM1/Tvivo and OM1/Cvivo sublines.

Discussion

In this study, we established the paclitaxel-resistant human ovarian cancer cell line (OM1/Tvivo) by in vivo strategy, i.e., selection of the drug-resistant subline after repeated paclitaxel-administrations into tumor-bearing mice. The characterization of the subline was performed under comparison with the in vivo-passaged drug-sensitive control subline (OM1/Cvivo) and the in vitro established paclitaxel-resistant subline (OM1/Tvitro), the latter of which was established by ‘classic’ in vitro serial selection procedure, i.e., selection after exposure to stepwise increased concentrations of paclitaxel in cell culture media. Both in vivo established sublines of OM1/Tvivo and OM1/Cvivo were confirmed as human originated tumors and no drastic changes were observed in chromosomal analysis between them (Fig. 1). Both the histological appearance of nude mouse-transplanted tumors and the morphology of the cultured cells were almost identical among these sublines and their parental OVMG1 (data not shown). Table 1 summarizes the biological phenotypes of each subline comparing their parental line. Although OM1/Tvivo tumor showed higher tumor growth as well as higher in vivo paclitaxel-resistance than the control OM1/Cvivo tumor (Fig. 2), OM1/Tvitro lost its tumorigenicity in exchange for obtaining an unreasonably high level of paclitaxel-resistance, which could be evaluated by in vitro drug sensitivity assays (Table 1 and Fig. 3). Considering the phenomenon of ‘reverse transformation’ (Biedler and Spengler 1994), the term which indicates the strongly suppressed tumorigenic and/or metastatic properties of in vitro selected drug-resistant cell lines, OM1/Tvitro also showed retarded tumorigenicity probably through this phenomenon. ‘Reverse transformation’ is considered to occur due to the absence of host-derived (in vivo) selection pressure during the process of establishing the drug-resistant subline, which results in receiving only drug-derived (in vitro) selection pressure, i.e., drug-exposure in culture dishes with extraordinary high concentration over prolonged periods (Biedler and Spengler 1994). This non-physiologic environment in the ‘classic’ in vitro serial selection procedure probably produced the poor-tumorigenic OM1/Tvitro subline, in spite of no growth retardation and an extremely high paclitaxel-resistance, in the case of in vitro culture condition. In a general clinical situation, tumorigenicity suppression is rarely observed in recurrent lesions after chemotherapy, which indicates that the in vitro established OM1/Tvitro subline is not a suitable model for clinically recurrent and drug-resistant ovarian cancers.

On the other hand, in vivo established OM1/Tvivo showed a higher tumor growth rate than its control (drug-sensitive) subline of OM1/Cvivo, with their tumor doubling times of 11.1 days and 16.8 days, respectively (Table 1 and Fig. 2a). Although OM1/Tvivo acquired an enhanced in vivo growth property—as previous reports describing the in vivo established drug-resistant cell lines often show increased malignant potentials (Antoine et al. 1988)—no invasion or metastasis was observed as in its parental OVMG1 (Table 1). In the case of our previously reported in vivo established cisplatin-resistant murine cancer cell lines, both invasive and metastatic properties were enhanced, compared with the in vivo-passaged control (drug sensitive) cell lines (Mitsumoto et al. 1998). Different types of enhanced malignant properties observed between our previous and present studies are probably due to the different species of inoculated tumors, that is, drug-resistant ‘murine’ cell lines inoculated into ‘syngeneic mice’ in our previous study resulted in enhanced invasive and metastatic properties, while the drug-resistant ‘human’ cell line inoculated as xenograft into ‘nude mice’ in the present study could not acquire the ability to make invasive and/or metastatic lesions. The acquisition of malignant phenotype by the OM1/Tvivo subline, however, could be confirmed by its enhanced tumor growth property. Therefore, the in vivo established OM1/Tvivo subline is considered to be a more suitable and faithful model for clinically recurrent ovarian cancers, after paclitaxel-based chemotherapy, rather than the in vitro established OM1/Tvitro subline which is generated by the ‘classic’ in vitro serial selection procedure.

As for the expression of paclitaxel-resistant profiles of OM1/Tvivo, drug-resistance properties were only manifested in vivo: surprisingly, the OM1/Tvivo cells maintained in culture dishes were no more-resistant than the control OM1/Cvivo cells which were as sensitive to paclitaxel as their parental OVMG1 cells (Fig. 2 and Fig. 3). Reinjection of cultured OM1/Tvivo cells into mice after being maintained in a drug-free culture media for 2 months resulted in the reexpression of their drug-resistance properties in vivo (data not shown), which strongly suggests the existence of a paclitaxel-resistance mechanism operative only in vivo. This is similar to the results of Teicher et al. who established a murine EMT-6 mammary tumor-derived drug-resistant subline in vivo (Teicher et al. 1990), and Starling et al. who established a human lung adenocarcinoma-derived drug-resistant subline in vivo (Starling et al. 1990). As one explanation of why the in vivo drug-resistance of OM1/Tvivo was not recaptured as cellular-resistance in vitro, we found a significant difference in pharmacokinetics after paclitaxel administration, i.e., the paclitaxel concentration in OM1/Tvivo tumors was significantly lower than that in OM1/Cvivo tumors, 24 h and 48 h after paclitaxel injection (Fig. 4a), but no difference in cellular accumulation was observed between these two sublines, in the case of in vitro cell culture (Fig. 4b). These results suggest that the decreased influx and/or the increased efflux of the drug into the OM1/Tvivo cells (but not the OM1/Cvivo cells) occurred only in vivo. Paclitaxel-resistance has been reported to be associated with overexpression of P-glycoprotein (Han et al. 2000), depletion of cellular glutathione (Liebmann et al. 1993), acetylation of alpha-tubulin (Ohta et al. 1994), increase in alpha-tubulin (Han et al. 2000), mutation in beta-tubulin (Monzo et al. 1999), altered expression of specific beta-tubulin isotypes (Kavallaris et al. 1997), and so on. Among these candidates for explaining the in vivo-derived paclitaxel-resistance mechanism, we now focus on the P-glycoprotein, a well-known drug-efflux pump, and examine whether its overexpression is specific or not to OM1/Tvivo tumor.

The emergence of a drug-resistant subpopulation is explained by the theories reported by Coldman and Goldie (Coldman and Goldie 1985), which mathematically described the genetic instability of cancers and advantageous survival observed in genetic variants that are resistant to a particular drug. In addition to their explanation, we believe that the drug-resistance mechanism observed in our in vivo established subline involves an interaction between the tumor and the host stromal tissues rather than only changes in cellular sensitivity of the tumor to paclitaxel, i.e., a host-environment is probably necessary for the expression of a drug-resistant profile of OM1/Tvivo. It has been reported that co-culture with host stromal cells (Zhu et al. 1999), culture on extracellular matrix (Zhu et al. 1999; Fridman et al. 1990), or three-dimensional cell culture (Frankel et al. 1997) could enhance the anticancer drug-resistance through additional in vivo-like cell culture conditions of tumor cell-stromal cell interaction, tumor cell-extracellular matrix interaction, and tumor cell-tumor cell interaction, respectively. By especially making multicellular aggregates (i.e., tumor spheroids) in the three-dimensional cell culture system, we previously disclosed a drug-resistance mechanism operative ‘only in vivo’ reported in Teicher’s EMT-6 sublines, and succeeded in recapturing the drug-resistance even in a cell culture system (Kobayashi et al. 1993). Considering these results, there is a possibility that the above-mentioned factors reveal not only paclitaxel-resistance but also enhanced tumor growth of OM1/Tvivo in host animals. Therefore, experiments using various in vivo-like culture conditions are under way to resurrect the drug-resistance of OM1/Tvivo in the cell culture system.

Unfortunately, there are few reports describing the malignant progression and/or drug-resistance mechanism of established human drug-resistant sublines by in vivo selection procedures. Although a study of drug-resistance appearing mainly in vivo (tissue-level-resistance) is more difficult than the study of the drug-resistance mechanism operating in cultured cells (cellular-level-resistance), an elucidation of such in vivo drug-resistance mechanisms may be important, clinically, to prevent or overcome acquired drug-resistance in ovarian cancers.

To our knowledge, this is the first report to describe the biological characterization of an in vivo established paclitaxel-resistant human ovarian cancer cell line. Further research with regard to this cell line should be useful for understanding the clinical behavior of drug-resistant ovarian cancer recurring after paclitaxel-based chemotherapy.

Acknowledgment

We thank Ms. Linda Saza for help with manuscript preparation and Ms. N. Hirakawa, Ms. M. Ogawa, Ms. Y. Nakajo, and Ms. E. Hori for technical assistance. We sincerely thank Dr. N. Wake for critical comments on our results. This work was supported in part by Grants-in-Aid for Scientific Research (C) from the Ministry of Education, Culture, Sports, Science, and Technology, Japan (No. 12671612, No. 12671613 and No. 15591757).

References

  1. Abe Y, Ohnishi Y, Yoshimura M, Ota E, Ozeki Y, Oshika Y, Tokunaga T, Yamazaki H, Ueyema Y, Ogata T, Tamaoki N, Nakamura M (1996a) P-glycoprotein-mediated acquired multidrug-resistance of human lung cancer cells in vivo. Br J Cancer 74:1929–1934 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Abe Y, Yamazaki H, Oshika Y, Suto R, Tsugu A, Ota E, Satoh H, Ohnishi Y, Yanagawa T, Ueyama Y, Tamaoki N, Nakamura M (1996b) Advantage of in vivo chemosensitivity assay to detect vincristine-resistance in a human epidermoid carcinoma xenograft. Anticancer Res 16:729–734 [PubMed] [Google Scholar]
  3. Antoine E, Pauwels C, Verrelle P, Lascaux V, Poupon MF (1988) In vivo emergence of a highly metastatic tumour cell line from a rat rhabdomyosarcoma after treatment with an alkylating agent. Br J Cancer 57:469–474 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bashir I, Sikora K, Abel P, Foster CS (1994) Establishment and in vivo characterization of multidrug-resistant dunning R3327 rat prostate-carcinoma cell-lines. Int J Cancer 57: 719–726 [DOI] [PubMed] [Google Scholar]
  5. Biedler JL, Spengler BA (1994) Reverse transformation of multidrug-resistant cells. Cancer Metast Rev 13: 191–207 [DOI] [PubMed] [Google Scholar]
  6. Coldman AJ, Goldie JH (1985) Role of mathematical modeling in protocol formulation in cancer chemotherapy. Cancer Treat Rep 69:1041–1048 [PubMed] [Google Scholar]
  7. Einzig AI, Wiervik PH, Sasloff J, Runowicz CD, Goldberg GL (1992) Phase II study and long-term follow-up of patients treated with paclitaxel for advanced ovarian adenocarcinoma. J Clin Oncol 10:1748–1753 [DOI] [PubMed] [Google Scholar]
  8. Frankel A, Buckman R, Kerbel RS (1997) Abrogation of taxol-induced G2-M arrest and apoptosis in human ovarian cancer cells grown as multicellular tumor spheroids. Cancer Res 57:2388–2393 [PubMed] [Google Scholar]
  9. Fridman R, Giaccone G, Kanemoto T, Martin GR, Gazdar AF, Mulshine JL (1990) Reconstituted basement membrane (matrigel) and laminin can enhance the tumorigenicity and the drug-resistance of small cell lung cancer cell lines. Proc Natl Acad Sci USA 87:6698–6702 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ganapathi R, Grabowski D, Schmidt H, Bell D, Melia M (1987) Characterization in vitro and in vivo of progressively adriamycin-resistant B16-BL6 mouse melanoma cells. Cancer Res 47: 3464–3468 [PubMed] [Google Scholar]
  11. Han EK-H, Gehrke L, Tahir SK, Credo RB, Cherian SP, Sham H, Rosenberg SH, Ng SC (2000) Modulation of drug-resistance by alpha-tubulin in paclitaxel-resistant human lung cancer. Eur J Cancer 36:1565–1571 [DOI] [PubMed] [Google Scholar]
  12. Ishiyama M, Tominaga H, Shiga M, Sasamoto K, Ohkura Y, Ueno K (1996) A combined assay of cell viability and in vitro cytotoxicity with a highly water-soluble tetrazolium salt, neutral red and crystal violet. Biol Pharm Bull 19:1518–1520 [DOI] [PubMed] [Google Scholar]
  13. Kamura T (1996) Alteration of metastatic potential of ovarian cancer in clinical course. Acta Obst Gynaec Jpn 48: 607–617 [PubMed] [Google Scholar]
  14. Kavallaris M, Kuo DY, Burkhart CA, Regl DL, Norris MD, Haber M, Horwitz SB (1997) Taxol-resistant epithelial ovarian tumors are associated with altered expression of specific ß-tubulin isotypes. J Clin Invest 100:1282–1293 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kobayashi H, Man S, Graham CH, Kapitain SJ, Teicher BA, Kerbel RS (1993) Acquired multicellular-mediated-resistance to alkylating agents in cancer. Proc Natl Acad Sci USA 90:3294–3298 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Liebmann JE, Hahn SM, Cook JA, Lipschultz C, Mitchell JB, Kaufman DC (1993) Glutathione depletion by L-buthionine sulfoximine antagonizes taxol cytotoxicity. Cancer Res 53:2066–2070 [PubMed] [Google Scholar]
  17. Manfredi JJ, Parness J, Horwitz SB (1982) Taxol binds to cellular microtubles. J Cell Biol 94:688–696 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. McGuire WP, Hoskins WJ, Brady MF, Kucera PR, Partridge EE, Look KY, Clarke-Pearson DL, Davidson M (1996) Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and IV ovarian cancer. N Engl J Med 334:1-6 [DOI] [PubMed] [Google Scholar]
  19. Mitsumoto M, Kamura T, Kobayashi H, Sonoda T, Kaku T, Nakano H (1998) Emergence of higher levels of invasive and metastatic properties in the drug-resistant cancer cell lines after the repeated administration of cisplatin in tumor-bearing mice. J Cancer Res Clin Oncol 124:607–614 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Monzo M, Rosell R, Sanchez JJ, Lee JS, O’Brate A, Gonzalez-Larriba JL, Alberola V, Lorenzo JC, Nunez L, Ro JY, Martin C (1999) Paclitaxel-resistance in non-small-cell lung cancer associated with beta-tubulin gene mutations. J Clin Oncol 17:1786–1793 [DOI] [PubMed] [Google Scholar]
  21. Ohta S, Nishio K, Kubota N, Ohmori T, Funayama Y, Ohira T, Nakajima H, Adachi M, Saijo N (1994) Characterization of a taxol-resistant human small-cell lung cancer cell line. Jpn J Cancer Res 85:290–297 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Scotlandi K, Serra M, Nicoletti G, Vaccari M, Manara MC, Nini G, Landuzzi L, Colacci A, Bacci G, Bertoni F, Picci P, Campanacci M, Baldini N (1996) Multidrug-resistance and malignancy in human osteosarcoma. Cancer Res 56: 2434–2439 [PubMed] [Google Scholar]
  23. Starling JJ, Maciak RS, Hinson NA, Hoskins J, Laguzza BC, Gadski RA, Strnad J, Rittmann-Grauer L, DeHerdt SV, Bumol TF, Moore RE (1990) In vivo selection of human tumor cells-resistant to monoclonal antibody-Vinca alkaloid immunoconjugates. Cancer Res 50:7634–7640 [PubMed] [Google Scholar]
  24. Teicher BA, Cucchi CA, Lee JB, Flatow JL, Rosowsky A, Frei E III (1986) Alkylating agents: in vitro studies of cross-resistance patterns in human cell lines. Cancer Res 46:4379–4383 [PubMed] [Google Scholar]
  25. Teicher BA, Herman TS, Holden SA, Wang Y, Pfeffer MR, Crawford JW, Frei E III (1990) Tumor-resistance to alkylating agents conferred by mechanisms operative only in vivo. Science 247:1457–1461 [DOI] [PubMed] [Google Scholar]
  26. Zhu B, Block NL, Lokeshwar BL (1999) Interaction between stromal cells and tumor cells induces chemoresistance and matrix metalloproteinase secretion. Ann NY Acad Sci 878:642–646 [DOI] [PubMed] [Google Scholar]

Articles from Journal of Cancer Research and Clinical Oncology are provided here courtesy of Springer

RESOURCES