• vault;
  • MVP;
  • LRP;
  • U-937;
  • multidrug resistance


  1. Top of page
  2. Abstract

U-937 human leukemia cells were selected for resistance to doxorubicin in the presence or absence of a specific drug modulator that inhibits the activity of P-glycoprotein (Pgp), encoded by the multidrug-resistance gene (MDR1). Parental cells expressed low basal levels of the multidrug-resistance-associated gene (MRP1) and major vault protein (MVP) mRNAs and no MDR1 mRNA. Two doxorubicin-resistant cell lines were selected. Both drug-resistant cell lines upregulated the MVP mRNA level 1.5-fold within 1 cell passage. The MVP mRNA level continued to increase over time as the doxorubicin selection pressure was increased. MVP protein levels generally paralleled the mRNA levels. The 2 high molecular weight vault protein mRNAs were always expressed at constitutive levels. Fully formed vault particles consisting of the MVP, the 2 high molecular weight proteins and the vault RNA assembled and accumulated to increased levels in drug-selected cells. MVP induction is therefore the rate-limiting step for vault particle formation in U-937 cells. By passage 25 and thereafter, the selected cells were resistant to doxorubicin, etoposide, mitoxantrone and 5-fluorouracil by a pathway that was independent of MDR1, MRP1, MRP2 and breast cancer resistance protein. In summary, U-937 doxorubicin-selected cells are programmed to rapidly upregulate MVP mRNA levels, to accumulate vault particles and to become multidrug resistant. © 2002 Wiley-Liss, Inc.

Vaults are ribonucleoprotein particles that are conserved throughout evolution in diverse phylogeny including mammals, avians, amphibians and the slime mold.1 They were first observed in preparations of clatharin-coated vesicles and were named based on their structural similarity to arched cathedral ceilings.2 Vault particles have a mass of approximately 13 MDa and are composed of multiple copies of 3 proteins and a unique, untranslated RNA. The major vault protein (MVP) constitutes 70% of the total mass of the particle. The remaining mass comprises vault RNA and 2 high molecular weight proteins, vault poly(ADP-ribose) polymerase3 and telomerase-associated protein 1.4 The MVP has also been referred to as the lung resistance-associated protein (LRP).5

Several groups have documented that the cellular level of MVP is an excellent predictor of multidrug resistance (MDR) in cancer cell lines and in clinical tumors.6 However, it is not known whether vault particles act alone or in combination with other drug-resistance factors to confer a multidrug-resistance phenotype, or whether vaults are purely a marker for the phenotype. Previous studies have noted that MVP is upregulated early during drug selection.7–10 Many cancer cell lines upregulate not only MVP but also MRP1, MDR1 and other drug transporters. In revertant MDR cancer cell lines, the number of vault particles decreases.11

Previously, human U-937 myeloid leukemia cells were selected for drug resistance in the presence of increasing concentrations of doxorubicin (Dox).12 By Northern blot analysis, parental U-937 cells expressed a low level of MRP1 and expressed no MDR1. In the presence of Dox, the level of MRP1 first increased followed by a later increase in MDR1 expression. None of the sublines showed gene amplification. Verapamil was able to significantly modulate IC50 values thus implicating these ATP binding cassette (ABC) drug transporters in the drug-resistance pathway. Vault levels were not studied. We have studied these cells in greater detail, especially with regard to the time course of vault induction and how this relates to the time course for the induction of multidrug resistance.

As an extension of our previous studies with U-937 cells, we now show that the MVP mRNA is induced within 1 cell passage in the presence of a low concentration of Dox. In marked contrast, the MRP1 mRNA is induced in only 1 drug-selected line but only after 35 cell passages in the presence of a higher concentration of drug. Therefore, an early induction of MVP mRNA can be separated in time from a much later induction of MRP1 mRNA. MDR1 is not expressed and MRP2 and breast cancer resistance protein (BCRP) are not induced. U-937 cells are programmed to rapidly upregulate MVP mRNA levels in response to Dox selection. This is followed by the accumulation of vault particles and the acquisition of a multidrug-resistant phenotype.


  1. Top of page
  2. Abstract


[α-3H]rUTP (35 Ci/mmol) and [α-33P]dCTP (2,000 Ci/mmol) were purchased from NEN Life Science Products (Boston, MA). Dox, etoposide, mitoxantrone and 5-fluorouracil were purchased from the Sigma Chemical Co. (St. Louis, MO). Eli Lilly and Company licensed the Pgp inhibitor, LY335979 ((2R)-anti-5- {3-[4-(10,11-difluoromethanodibenzo-suber-5-yl]-2-hydroxypropoxy}quinoline trihydrochloride), molecular weight 637 Daltons, from the Syntex Company (RS-33295-198).13

Cell culture

The human myeloid leukemia cell line U-937,14 free of K562 contamination, was obtained from the American Type Culture Collection (Rockville, MD). Reports before 1994 may have been impacted by possible K562 contamination of U-937 cell lines.15 Cells were grown in T25 flasks in suspension culture in a 5% CO2 atmosphere in RPMI 1640 medium containing 10% fetal bovine serum (Life Technologies, Grand Island, NY) and 0.05% DMSO with or without 5 μM of the Pgp inhibitor, LY335979. Two different Dox-selected, drug-resistant sublines of U-937 were derived by growing U-937 cells in medium containing 5.0 ng/ml (10 nM) of Dox in the presence and absence of 5 μM of LY335979. Subsequently, U-937 sublines capable of growing in 10, 25 and 50 ng/ml of Dox were selected. Cells were passaged approximately once a week as a 1-to-10 split when cell density was ∼1 × 106/ml. All cell lines grew at the same rate.

Drug sensitivity assays

Sensitivity of the U-937 sublines to Dox, etoposide, mitoxantrone and 5-fluorouracil were determined after 96 hr of drug exposure using an MTT assay performed in quadruplicate for each of 3 separate experiments.16 The IC50 values were calculated using JMP software. Fold resistance values were expressed as the ratio of the IC50 of a drug-resistant cell line divided by the IC50 of the parental U-937 cell line.

Quantitative RT-PCR

Total cell RNA was isolated from 1.0 × 107 cells of each cell line using TRIzol Reagent (Life Technologies). The final RNA pellet was dissolved in 50 μl of nuclease-free water, aliquoted at 125 ng/μl, and kept at −80°C. The nucleotide numbering system for all cDNAs is based on the GenBank accession number, with base 1 being the 5′ most base of the 5′ untranslated region. The quantitative RT-PCR protocol used an rTth polymerase because this enzyme is stable at high temperatures and can therefore reverse transcribe RNA at high temperatures (EZ rTth RNA PCR Kit, PE Applied Biosystems). The system was modified to allow both a hot start of the reverse transcription portion of the reaction as well as the analysis of multiple samples simultaneously. RNA standards were used to allow for accurate quantitation. The exact protocol is described in detail elsewhere.17 Basically, reverse transcriptase reactions of both sample total RNAs and RNA standards were run at 60°C for 1 hr followed by a single DNA denaturation step at 94°C for 40 sec. PCR amplifications were performed at 94°C for 20 sec, 60°C for 30 sec, and 68°C for 60 sec for various numbers of cycles depending on the mRNA analyzed. RT-PCR products were electrophoresed on 4–20% polyacrylamide gels (Novex, San Diego, CA). Table I provides the primer sequences used and specific details unique to individual PCR reactions. A control reaction run without reverse transcriptase was included in every experiment to ensure that the generated PCR product was not a result of DNA contamination. RNA integrity was verified by RT-PCR amplification of glyceraldehyde-3-phosphate dehydrogenase mRNA transcripts.

Table I. Primer Sequences and PCR Reaction Conditions
Gene (Genbank1 accession)PrimerPrimer sequences 5′ to 3′Manganese concentration (μM)No. of PCR cycles
  • 1

    The nucleotide numbering system for all primers is based on the GenBank accession number, with base 1 being the 5′-most base of the 5′- untranslated region.


Subcellular fractionations

Cells (1 × 108) were resuspended in 1.0 ml cold buffer A (50 mM Tris-HCl pH 7.4, 1.5 mM MgCl2, 75 mM NaCl) containing 1% Triton X-100, 1 mM dithiothreitol, and the following protease inhibitors: 1 mM phenylmethylsulfonyl fluoride, 2 μg/ml aprotinin, 0.5 mM benzamidine, 2 μg/ml chymostatin, 5 μM leupeptin and 5 μM pepstatin. All subsequent steps were performed at 4°C. Cells were vortexed, incubated on ice for 10 min and centrifuged at 20,000g for 20 min. This supernatant, designated the S20 fraction, was used for analysis, or was further centrifuged at 100,000g for 1 hr. This supernatant was designated the S100 fraction. The pellet (designated P100) was resuspended by Dounce homogenization in buffer A containing 1% Triton X-100, 1 mM dithiothreitol and protease inhibitors in the original volume. Protein concentrations were determined in triplicate using the Bio-Rad protein assay reagent (Bio-Rad).

Sucrose gradient fractionation of vaults

P100 fractions (equivalent to 2 × 108 cells) were applied to a 20/30/40/45/50/60% sucrose step gradient in buffer A (containing 1.5 ml in each layer) and centrifuged at 28,000 rpm in a Beckman SW-41 rotor for 16 hr. Under these conditions, intact vaults localize to the 40/45% layers.11 Gradient fractions for each layer were collected and diluted 5-fold with buffer A and centrifuged at 100,000g for 3 hr. Pellets were resuspended in buffer A and analyzed for both protein and RNA content. Protein samples were solubilized in SDS sample buffer and resolved by 6% SDS-PAGE, then transferred to Hybond-C (Amersham Corp.) by electroblotting. Western blot analyses were performed with affinity purified polyclonal antibodies against MVP, VPARP and TEP1.3, 4 All antibodies were made in rabbits. The VPARP and TEP1 antibodies were made against human proteins while the MVP antibody was made to a rat protein. The MVP antibody cross-reacts with the human protein. Reactive bands were visualized by enhanced chemiluminescence (Amersham Corp.). Quantification of reactive bands was carried out using a Molecular Dynamics Personal Densitometer using Imagequant software (Molecular Dynamics). RNA from cellular fractions was purified by phenol/chloroform extraction and ethanol precipitation. RNA samples were resolved using 8 M urea, 10% polyacrylamide gels, then electroblotted to a Zeta GT membrane (Bio-Rad). The membrane was hybridized with a randomly primed human vRNA probe.11

Immunoprecipitation analysis of vaults and vault components

MVP, VPARP or TEP1 were immunoprecipitated from P100 fractions (500 μg of total protein) using 25 μl of the cognate affinity purified antibodies, in a final volume of 200 μl of buffer A containing 1% Triton X-100, 1 mM dithiothreitol and protease inhibitors. Incubations were overnight at 4°C. Protein-A Sepharose (Pharmacia) was then added and incubated for 2 hr at 4°C. Precipitates were pelleted and washed 3 times with buffer A containing 1% Triton X-100 and 1 mM dithiothreitol and washed twice with PBS. Samples were resuspended in 50 μl of SDS sample buffer and resolved by SDS-PAGE. Immunoprecipitated proteins were detected by Western blot analysis as detailed above.

Statistical methods

In calculating whether mRNA levels significantly differed, both the error in the RNA standard curve and in the data were taken into account. When inductions were small (generally less than 2-fold) more points were required on the RNA standard curve to achieve significance. When inductions were large (generally greater than 2-fold) fewer points were required on the RNA standard curve. Samples of unknown concentrations were analyzed in triplicate. The concentration corresponding to the mean OD value for each treated sample was obtained from the RNA standard curve. All standard curves were expressed using a linear model on the logarithmic scale of both the concentrations and the OD values. The standard error associated with each sample was derived by pooling together the 2 sources of variability: the variability in the fitting of the standard curve and the variability between the replicate samples. The first source of variability was derived by using a formula provided in Applied Linear Statistical Models.18 The second source of variability was derived using the sample standard deviation of the replicate OD measurements.

Two 6-point RNA standard curves at concentrations of 10, 5, 1.25, 0.625, 0.313 and 0.165 attomoles were obtained, one for the analysis of MVP mRNA levels at passages 1, 3, 5, 7 and 10, and the other for the analysis of passages 20, 25, 30, 35 and 40. Two experiments were performed because of the large number of samples to be analyzed. MVP mRNA levels in the 4 treated samples (U-937, U-937/LY335979, U-937/Dox and U-937/LY335979/Dox) were measured. MRP1 mRNA inductions were larger and therefore simpler to analyze. Two 3-point RNA standard curves at concentrations of 1, 0.1 and 0.01 attomoles were obtained, one for the analysis of MRP1 at passages 1, 3, 5, 7 and 10, and the other for the analysis of passages 20, 25, 30, 35 and 40.

Two-way analysis of variance (ANOVA) was performed on the IC50 data comparing cell passage numbers (10, 25 and 35), cell treatments (none, Pgp inhibitor, Dox and Dox plus inhibitor) and the Passage × Treatment interaction effect. Pairwise comparisons were made by splicing the Passage × Treatment interaction. Comparisons done this way are generally far superior to Student's t-tests and provide a much greater statistical power for detecting differences. The results are summarized in Table II.

Table II. Drug Sensitivity of Passage 10, 25 and 35 U-937 Cells Grown in the Presence of Doxorubicin or Doxorubicin/LY335979
DrugsIC50 (μg/ml)
  • The numbers in parentheses are the fold resistances of doxorubicin or doxorubicin/LY335979 selected cell lines compared with parent cells.

  • 1

    p < 0.05, statistically significant compared to parent cell line U937 by 2-way analysis of variance.

  • 2

    p < 0.01, highly statistically significant compared to parent cell line U937 by 2-way analysis of variance.

Pasage 10 U-937 cells
Doxorubicin0.012 ± 0.0060.008 ± 0.0080.016 ± 0.0090.017 ± 0.011
Etopside0.189 ± 0.0640.176 ± 0.0460.337 ± 0.2190.248 ± 0.164
Mitoxantrone0.001 ± 0.0010.001 ± 0.0010.001 ± 0.0010.001 ± 0.001
Pasage 25 U-937 cells
Doxorubicin0.015 ± 0.0040.018 ± 0.0090.059 ± 0.012 (3.9)20.076 ± 0.012 (5.1)2
Etopside0.202 ± 0.0370.200 ± 0.0571.961 ± 0.422 (9.7)22.243 ± 0.0.231 (11.1)2
Mitoxantrone0.002 ± 0.0020.003 ± 0.0030.011 ± 0.007 (5.5)10.009 ± 0.007 (4.5)1
Pasage 35 U-937 cells
Doxorubicin0.015 ± 0.0060.013 ± 0.0040.150 ± 0.021 (10.0)20.112 ± 0.030 (7.5)2
Etopside0.212 ± 0.0490.225 ± 0.0624.780 ± 1.271 (22.54)23.964 ± 0.959 (18.70)2
Mitoxantrone0.003 ± 0.0020.005 ± 0.0040.024 ± 0.018 (7.67)10.009 ± 0.005 (3.33)1
5-Fluorouracil140 ± 42157 ± 64319 ± 108 (2.3)1748 ± 237 (4.8)1


  1. Top of page
  2. Abstract

MVP mRNA is induced early by Dox selection—mRNAs for the high molecular weight vault proteins are not

To determine at which passage the MVP mRNA is induced in response to Dox, a dose was selected that was minimally toxic. Doxorubicin at 5 ng/ml resulted in 5–10% cell death. After the adaptation of cells to growth in the initial 5 ng/ml of drug, the dose was gradually increased to 10, 25 and finally 50 ng/ml over 40 cell passages (Fig. 1a). U-937 cells were grown in the presence of 5 μM Pgp inhibitor LY335979, Dox alone, with both inhibitor and Dox, or no drug. All 4 cell lines grew at the same rate. A qRT-PCR assay was employed that incorporated RNA mimics to allow accurate quantitation of mRNA levels. Results were expressed as attomoles/μg total RNA and as number of mRNA molecules/cell.17

thumbnail image

Figure 1. qRT-PCR determinations of mRNA levels for the major vault protein (MVP) and the 2 high molecular weight vault proteins during doxorubicin (Dox) selection of U-937 cells grown over 40 cell passages. U-937 cells were grown either alone, with the Pgp inhibitor LY335979, with Dox alone or with the inhibitor and increasing concentrations of Dox as indicated over 1–40 cell passages. A qRT-PCR assay was used to calculate MVP mRNA levels (a), vault poly(ADP-ribose)polymerase (VPARP) mRNA levels (b) and telomerase-associated protein 1 (TEP1) mRNA levels (c). mRNA levels are expressed as ± SE as based on 2 sources of error, that from the RNA standard curve and that from the sample measurements.

Download figure to PowerPoint

One cell passage in the presence of a low dose of Dox was sufficient to induce the synthesis of the MVP mRNA (Fig. 1a). This finding could be demonstrated only with a sensitive qRT-PCR assay and was statistically significant relative to parental cells (p < 0.05). Other assays, such as Western blots, lack the sensitivity to accurately quantitate a 1.5-fold induction. In subsequent passages, the level of MVP mRNA continued to increase, reaching a 2.3-fold induction in Dox-selected cells and a 4.1-fold induction in Dox/LY335979 cells by passage 35 (Fig. 1a). In these experiments, both time in culture and the level of exposure to Dox were increasing. In vehicle-treated cells or modulator-treated cells, the MVP mRNA level remained unchanged over time at all passages (Fig. 1a). It is therefore in response to Dox that the MVP mRNA was initially induced and continued to increase in amount as the drug selective pressure was progressively increased.

In addition to the 100-kDa MVP, vault particles contain 2 high molecular weight proteins and a vault RNA.3, 4 qRT-PCR assays were developed for the 2 high molecular weight protein mRNAs. Over the entire course of the Dox selection, the levels of these mRNAs remained constant (Fig. 1b,c). The levels of both mRNAs, expressed on a per-cell basis, were biologically significant, considered to be higher than 5 mRNA copies per cell. Although the high molecular weight proteins could be detected by immunoprecipitation followed by Western blot analyses (data not shown), protein levels were so low that it was not possible to accurately quantitate changes during drug selection. If the protein levels corresponded to each of the mRNA levels, then it is likely that the high molecular weight proteins were present in excess in the cell supernatant and were not limiting for the formation of vault particles. The vault RNA has already been shown to be present in great excess in the cell supernatant in HeLa cells.11

The MVP protein level was, for the most part, constant in parental cells at all passages (Fig. 2). The minor variability observed in a representative analysis (e.g., a higher level of MVP protein in passage 30 versus passage 35 U-937 cells, Fig. 2) was likely due to the difficulty in obtaining accurate quantitative values based on immunoprecipitation and Western blot analyses. In general, the MVP protein level steadily increased during Dox selection except for 1 time point (passage 35 U-937/Dox cells showed a decrease in MVP protein versus passage 25 and 30 cells, Fig. 2). Again, this is 1 representative analysis and a similar but different variability was seen in other analyses (data not shown). The relative protein quantitations are not as accurate as the mRNA quantitations, which are based on a sensitive and accurate qRT-PCR analysis. The trend was clear, however; as the MVP mRNA level increased during Dox selection the MVP protein level had a corresponding increase.

thumbnail image

Figure 2. Western blot analyses of major vault protein (MVP) levels after doxorubicin (Dox) selection of U-937 cells over 35 cell passages. U-937 cells were grown alone, with LY335979, with Dox alone or with Dox and LY335979. MVP was immunoprecipitated from a P100 fraction that contained intact vault particles. Samples were resolved by SDS-PAGE and proteins detected by Western blot analyses using affinity-purified antibodies. Hybridized bands were detected by chemiluminescence.

Download figure to PowerPoint

Fully formed vault particles were induced by Dox selection

To demonstrate that the Dox-induced increase in MVP mRNA and protein resulted in the assembly of intact vault particles, vaults from the 4 cell lines were partially purified on sucrose gradients. The P100 fractions (vault-enriched fraction) from passage 35 cells were fractionated over a sucrose step gradient (Fig. 3). All of the vault protein components and the vault RNA were demonstrated to cofractionate, confirming that fully assembled vaults are found in each of the 4 cell lines. Because of the complexity of the vault isolation scheme, obtaining accurate quantitative comparisons among the 4 cell lines was difficult. However, it was clear that more vault particles were present in the U-937/Dox/LY335979 cells than in the other 3 lines. It was more difficult to demonstrate an increase in vault number in the U-937/Dox cells after sucrose fractionation, although the MVP and VPARP levels were increased relative to the 2 control cell lines. The demonstration of increased vault particles in these cells relies on our accurate qRT-PCR studies and on the relatively accurate Western blot studies shown in Figure 2. These experiments demonstrated that fully formed vault particles were present in both parental and drug-selected cells.

thumbnail image

Figure 3. Western blot analyses of sucrose gradient fractionated vault particles from passage 35 cells. P100 fractions from U-937 cells grown alone, grown with LY335979, with doxorubicin (Dox) alone or with LY335979 plus Dox were applied to a 20/30/40/45/50/60% sucrose gradient. Gradient fractions were collected, proteins were resolved by 6% SDS-PAGE and transferred to a Hybond C membrane. Western blot analyses were performed with affinity purified polyclonal antibodies for major vault protein (MVP), vault poly(ADP-ribose)polymerase (VPARP) and telomerase-associated protein 1 (TEP1) and reactive bands were visualized by enhanced chemiluminescence. RNA from gradient fractions was purified and resolved on 8 M urea, 10% polyacrylamide gels and was electroblotted to a Zeta GT membrane, which was hybridized with a randomly primed human vault-associated RNA probe.

Download figure to PowerPoint

Heat shock of cells did not induce the MVP mRNA

To demonstrate that the Dox induction of MVP mRNA by the first cell passage was not a result of general cellular stress, U-937 cells were subjected to a heat shock. As shown in Figure 4, heat shock of parental U-937 cells induced the HSP70 mRNA, demonstrating that a heat shock occurred, but the MVP mRNA was not induced. In the present study MVP was likely to be a limiting factor for vault formation, therefore a failure to increase levels of the MVP mRNA infers that vault particle levels were not increased in heat-shocked U-937 cells.

thumbnail image

Figure 4. Heat shock of passage 35 U-937 cells. U-937 parental cells were grown at 37°C until heat shocked by incubation at 42°C for 1.5 hr. Cells were allowed to recover at 37°C for 8 hr, at which time total RNA was isolated and the HSP70 and major vault protein (MVP) mRNAs were quantitated by qRT-PCR. PCR products were resolved on agarose gels and visualized by 33P autoradiography. AHS, after heat shock; BHS, before heat shock.

Download figure to PowerPoint

MDR1 and MRP1 mRNA levels in Dox-selected cells

A previous analysis of U-937 cells selected in the presence of Dox reported first an induction of MRP1 mRNA followed by a later induction of MDR1 mRNA.12 Vault particles were not studied. In our study, the Pgp inhibitor LY335979 was added to 1 U-937 cell line to prevent the emergence of cells that would express MDR1. No expression of MDR1 mRNA was observed. It is unclear why the second cell line selected with Dox alone also did not express the MDR1 mRNA. Because a sensitive qRT-PCR assay was used for our study, it is clear that the MDR1 gene was completely silent in both drug-selected cell lines. MDR1 clearly played no role in any of our results (qRT-PCR data for MDR1 not shown).

The 4 U-937 cell lines initially expressed a low basal level of MRP1 mRNA. At cell passage 35, the MRP1 mRNA was abruptly induced 3.5-fold only in the Dox/LY335979-selected cell line (Fig. 5a). The MRP1 protein showed exactly the same pattern, a low basal level of expression that was abruptly increased at passage 35 (Fig. 5b). MRP1 was not induced in cells grown only in the presence of LY335979 (Fig. 5a). MRP1 was thus induced by Dox selection, but only at very late times after drug exposure and in only 1 of the 2 Dox-selected cell lines.

thumbnail image

Figure 5. Multidrug-resistance-associated (MRP1) mRNA and protein levels observed in response to doxorubicin (Dox) selection of U-937 cells grown over 40 cell passages. (a) U-937 cells were grown either alone, with the Pgp inhibitor LY335979, with Dox alone or with the modulator and increasing concentrations of Dox over 1–40 cell passages. A qRT-PCR assay was used to measure MRP1 mRNA levels. (b) Total cell-extract-derived proteins were resolved by SDS-PAGE and MRP1 was detected by Western blot analysis using an affinity-purified antibody, QCRL-1, obtained from the Signet Laboratories. Hybridized bands were detected by chemiluminescence.

Download figure to PowerPoint

Cytotoxicity assays

In our study, vault particles were induced early in drug selection. MRP1 was induced only at late passages in only 1 drug-selected cell line grown in the presence of LY335979. Therefore, cells in which vault particles but not MDR1 or MRP1 were induced could be analyzed to determine whether they exhibit an increased drug-resistance phenotype. An MTT dye reduction assay was used to measure cell viability in the presence of Dox, etoposide and mitoxantrone at cell passages 10, 25 and 35 and for 5-fluorouracil at passage 35 only (Table II). Both passage 10 Dox-selected cell lines exhibited no resistance to the 3 drugs tested relative to parental cells grown either in the presence or absence of the inhibitor LY335979. Although these cell lines did not exhibit a statistically significant drug-resistant phenotype, there was an upward trend in the data for both Dox and etoposide. Both passage 25 drug-selected cell lines upregulated vaults but not MDR1 or MRP1, and both lines expressed a drug-resistant phenotype that was approximately equal: ∼5-fold resistant to Dox and mitoxantrone, and ∼10-fold resistant to etoposide. Passage 35 cells were more resistant than passage 25 cells to Dox and etoposide, whereas resistance to mitoxantrone remained unchanged. Because Dox, etoposide and mitoxantrone can all inhibit topoisomerase activity we wished to test another drug that clearly did not inhibit topoisomerase. Both drug-selected passage 35 cell lines were resistant to 5-fluorouracil (Table II). It is therefore clear that the drug-resistance profile we observed was not solely a result of topoisomerase alterations. Both drug-selected cell lines were initially selected in duplicate cultures that gave the same phenotype and all cell lines grew at the same rate. The drug selection protocol used was therefore highly reproducible and did not alter the growth characteristics of the cells.


  1. Top of page
  2. Abstract

It is not clear whether vaults directly participate in MDR or whether they are solely a marker for the phenotype. Two recent studies suggested a direct role and 1 study questioned the role of vaults in MDR. In the first study, a colon cancer cell line was incubated with sodium butyrate to induce differentiation.19 Transfection of 2 different anti-MVP ribozymes into these butyrate-treated cells made them more sensitive to a number of therapeutic drugs. While MVP, and presumably vault particles, were induced, MRP1 and MDR1 were also induced and complicated interpretation of the results. A second publication reached similar conclusions and reported that KB-3-1 cells treated with benzo[a]pyrene became drug resistant, increased the level of MVP and excluded Dox from nuclei.20

In contrast, 1 study reported that transfection of the MVP gene into A2780 cells did not confer a drug-resistance phenotype.5 That report did not demonstrate that the other vault components were present and able to interact with the introduced MVP to form intact vault particles. Recently, another laboratory studied the same MVP-transfected A2780 cell line and found that all vault components were present, the number of assembled vaults increased, but the cell line was not more drug resistant than the parental line.21 That study did not rule out a role of vaults in MDR; however, it did suggest that the mechanism for vault-induced MDR is complex. These results led to our current belief that other factors in addition to vaults are involved in a possible vault-mediated MDR.

Based on our hypothesis, U-937 cells were selected to become drug resistant in the presence of Dox plus or minus a P-glycoprotein inhibitor, LY335979,22 added to prevent the emergence of cells that might express MDR1. Cell lines could then be studied for vault induction and the development of multidrug resistance in the absence of any possible MDR1 contribution. A highly selective drug that could block the emergence of MRP1-expressing cells was not available. Others have tried a similar scheme to eliminate MDR1-expressing cells during a drug selection protocol but were forced to use less specific drug-resistance modulators. For example, one study selected MCF-7 cells with Dox and the first generation Pgp modulator, verapamil.23 The Dox-resistant MCF-7AdrVp cells did not overexpress Pgp but instead overexpressed BCRP, another member of the superfamily of ABC transporters that effluxes anthracyclines, mitoxantrone24 and topotecan.25 In another study, researchers selected Dox-resistant cells of MES-SA in the presence of a second generation Pgp inhibitor, PSC-833.26 Both verapamil and PSC-833 can modulate other proteins in addition to Pgp, for example, PSC-833 modulates MRP2 and a bile acid transporter.27 To circumvent the problem of multiple target inactivation, in the present study the highly potent and selective Pgp inhibitor LY335979 was used. The affinity of Pgp for LY335979 is 59 nM and the presence of 0.1 μM modulator fully sensitizes Pgp-expressing drug-resistant cells.13 Moreover, 5 μM LY335979 is unable to modulate MRP1, MRP2 or BCRP.28 Thus, LY335979 should block the expression of Pgp without altering other resistance pathways such as MRP1, MRP2 or BCRP. Unexpectedly, neither the U-937/Dox/LY335979 cell line nor the U-937/Dox cell line expressed Pgp. Apparently, the transcription of the MDR1 gene is difficult to activate in the U-937 cell and a Pgp inhibitor is not required to prevent the emergence of cells that express MDR1. However, the 2 drug-resistant cell lines provided independent confirmation of our results. Vault particle induction and the development of drug resistance were thus able to be studied in the absence of MDR1 expression.

A low basal level of MRP1 mRNA was expressed in U-937 cells. During Dox selection, both drug-selected cell lines expressed this constitutive level of mRNA until passage 35, at which time MRP1 mRNA was dramatically induced only in the U-937/Dox/LY335979 cells. It is not known why MRP1 was induced in only 1 of the 2 drug-selected cell lines.

In marked contrast to the late induction of MRP1 by passage 35 in 1 drug-selected cell line, we observed a rapid induction of MVP mRNA within 1 passage in the presence of Dox. This rapid induction of the MVP mRNA was not simply a response to general cellular stress because heat shock of cells does not induce the MVP mRNA. Another report showed that ultraviolet light treatment of cells also did not induce vault particle formation.3 The vault particle induction that we observed was due to Dox selection.

The mRNAs for the 2 high molecular weight vault proteins are expressed at constitutive levels that do not change in response to Dox selection. These vault proteins could be detected, but not accurately quantitated, because of the small amounts of protein present. Recent studies demonstrated that the 240-kDa high molecular weight vault protein is a shared common subunit with telomerase, TEP1,4 whereas the 193-kDa high molecular weight vault protein is a novel poly(ADP-ribose) polymerase (VPARP).3 Because the 2 high molecular weight proteins are present both free in the supernatant and associated with vaults, they may have multiple roles in the cell. Their levels may be regulated in a different manner than the MVP, which exists solely within vault particles and not free in the cytoplasm. Previous analyses using HeLa cells showed that the vault RNA is also present in excess in cell supernatants.11

The induced MVP that we measured was incorporated into assembled vault particles. Only 1 previous study has reported on the analysis of fully formed vault particles in cancer cells.11 In our study, vault particles from passage 35 cells from all 4 cell lines were analyzed and shown to assemble fully. Earlier passage cells have not yet been examined. The U-937/Dox/LY335979 cells clearly have an increased number of vault particles relative to parental cells. It was more difficult to demonstrate an increase in vault number in U-937/Dox cells. As emphasized previously, it is difficult to quantitate the vault proteins after a sucrose gradient purification step followed by Western analysis. For this reason we observed an increase in the levels of the MVP and VPARP proteins but not the TEP1 protein or the vault RNA in U-937/Dox cells. The MVP protein induction was verified by our qRT-PCR studies of MVP mRNA levels. We have no explanation for why Dox/LY335979 cells had a higher level of vaults than Dox-selected cells. Clearly LY335979 alone does not induce the formation of vault particles.

Cells grown without drug selection (plus or minus the drug-resistance modulator, LY335979) never became drug resistant nor did they upregulate the level of vault particles or MVP mRNA. Both selected cell lines became cross-resistant not only to the selective drug, Dox, but also to etoposide and mitoxantrone. These 3 drugs can inhibit topoisomerase as part of their pleiotropic action on cells. The 2-passage 35 Dox-selected cell lines were shown to have cross-resistance to 5-fluorouracil. This ensures that not all the resistance observed could be attributed to alterations in topoisomerase. In addition, the levels of drug and the distribution of drug within both parental and drug-resistant cells were measured and no differences were detected (data not shown). The acquisition of a MDR phenotype was therefore not accompanied by an obvious redistribution of drug within cells. As expected, there was not an increased efflux of drug from cells as is observed when ABC drug transporters are overexpressed. This finding lends support to our belief that the drug resistance we observed was not conferred by a member of the ABC transporter superfamily.

In both passage 25 and passage 35 drug-selected cell lines the level of resistance for U-937/Dox cells was statistically the same as for U-937/Dox/LY335979 cells, although the drug plus inhibitor-treated cells clearly had an increased level of vaults. Thus there is not a direct correlation between the level of drug resistance and the level of vault particles. We would attribute this finding to the necessity to induce other currently unknown proteins along with vaults to confer a MDR phenotype. It is predicted that the levels of such proteins must vary between these 2 cell lines.

In summary, we have reported a correlation between the induction of MVP mRNA, the accumulation of vault particles and the emergence of drug resistance. The increase in the level of MVP mRNA began within 1 cell passage in the presence of drug. This increase was accompanied by a trend toward an increased drug resistance by passage 10 and a statistically significant increase in drug resistance by passage 25 in 2 distinct drug-selected cell lines. The resistance observed was independent of contributions from MDR1, MRP1, MRP2 or BCRP. These observations are consistent with our model for a possible vault-mediated MDR.


  1. Top of page
  2. Abstract
  • 1
    Kedersha NL, Miguel NL, Bittner MC, et al. 1990. Vaults II. Ribonucleoprotein structures are highly conserved among higher and lower eukaryotes. J Cell Biol 1990;110: 895901.
  • 2
    Rome LH, Kedersha NL, Chugani DC. Unlocking vaults: organelles in search of a function. Trends Cell Biol 1991;1: 4750.
  • 3
    Kickhoefer VA, Siva AC, Kedersha NL, et al. 1999. The 193-kD vault protein, VPARP, is a novel poly(ADP-ribose) polymerase. J Cell Biol 1999;146: 91728.
  • 4
    Kickhoefer VA, Stephen AG, Harrington L, et al. 1999. Vaults and telomerase share a common subunit, TEP1. J Biol Chem 1999;274: 327127.
  • 5
    Scheffer GL, Wijngaard PLJ, Flens MJ, et al. The drug resistance-related protein LRP is the human major vault protein. Nat Med 1995;1: 57882.
  • 6
    Izquierdo MA, Scheffer GL, Schroeijers AB, et al. 1998. Vault-related resistance to anticancer drugs determined by the expression of the major vault protein LRP. Cytotechnology 1998;27: 13748.
  • 7
    Moran E, Cleary I, Larkin AM, et al. Co-expression of MDR-associated markers, including P-170, MRP and LRP and cytoskeletal proteins, in three resistant variants of the human ovarian carcinoma cell line, OAW42. Eur J Cancer 1997;33: 65260.
  • 8
    Verovski VN, Van den Berge DL, Delvaeye MM, et al. Low-level doxorubicin resistance in P-glycoprotein-negative human pancreatic tumour PSN1/ADR cells implicates a brefeldin A-sensitive mechanism of drug extrusion. Br J Cancer 1996;73: 596602.
  • 9
    Versantvoort CHM, Withoff S, Broxterman HJ, et al. Resistance-associated factors in human small-cell lung-carcinoma GLC4 sub-lines with increasing adriamycin resistance. Int J Cancer 1995;61: 37580.
  • 10
    Wyler B, Shao Y, Schneider E, et al. Intermittent exposure to doxorubicin in vitro selects for multifactorial non-P-glycoprotein-associated multidrug resistance in RPMI 8226 human myeloma cells. Br J Haematol 1997;97: 6575.
  • 11
    Kickhoefer VA, Rajavel KS, Scheffer GL, et al. 1998. Vaults are up-regulated in multidrug-resistant cancer cell lines. J Biol Chem 1998;273: 89714.
  • 12
    Slapak CA, Mizunama N, Kufe DW. Expression of the multidrug resistance associated protein and P-glycoprotein in doxorubicin-selected human myeloid leukemia cells. Blood 1994;84: 311321.
  • 13
    Dantzig AH, Shepard RL, Cao J, et al. Reversal of P-glycoprotein-mediated multidrug resistance by a potent cyclopropyldibenzosuberane modulator, LY335979. Cancer Res 1996;56: 41719.
  • 14
    Sundstrom C, Nilsson K. Establishment and characterization of a human histiocytic lymphoma cell line (U-937). Int J Cancer 1976;17: 56577.
  • 15
    Reid YA, McGuire L, O'Neill K, et al. Cell line cross-contamination of U-937 (correction of U-937). J Leukoc Biol 1995;57: 804.
  • 16
    Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 1983;16: 5563.
  • 17
    Tanzer LR, Hu Y, Cripe L, et al. A hot-start reverse transcription-polymerase chain reaction protocol that initiates multiple analyses simultaneously. Anal Biochem 1999;273: 30710.
  • 18
    Neter J, Wasserman W, Kutner M. Applied linear statistical models: regression, analysis of variance, and experimental designs, 3rd ed. Homewood, IL: Irwin, 1990: 175.
  • 19
    Kitazono M, Sumizawa T, Takebayashi Y, et al. Multidrug resistance and the lung resistance-related protein in human colon carcinoma SW-620 cells. J Natl Cancer Inst 1999;91: 164753.
  • 20
    Cheng HS, Lam W, Lee ASK, et al. Low-level doxorubicin resistance in benzo[a]pyrene-treated KB-3-1 cells is associated with increased LRP expression and altered subcellular drug distribution. Toxicol Appl Pharmacol 2000;164: 13442.
  • 21
    Siva AC, Raval-Fernandes S, Stephen AG, et al. Upregulation of vaults may be necessary but not sufficient for multidrug resistance. Int J Cancer 2001;92: 195202.
  • 22
    Starling JJ, Shepard RL, Cao J, et al. Pharmacological characterization of LY335979: a potent cyclopropyldibenzosuberane modulator of P-glycoprotein. Adv Enzyme Regul 1997;37: 33547.
  • 23
    Chen Y-N, Mickley LA, Schwartz AM, et al. Characterization of adriamycin-resistant human breast cancer cells which display overexpression of a novel resistance-related membrane protein. J Biol Chem 1990;265: 1007380.
  • 24
    Doyle LA, Yang WD, Abruzzo LV, et al. A multidrug resistance transporter from human MCF-7 breast cancer cells. Proc Natl Acad Sci USA 1998;95: 1566570.
  • 25
    Maliepaard M, van Gastelen MA, de Jong LA, et al. Overexpression of the BCRP/MXR/ABCP gene in a topotecan-selected ovarian tumor cell line. Cancer Res 1999;59: 455963.
  • 26
    Beketic-Oreskovic L, Duran GE, Chen G, et al. Decreased mutation rate for cellular resistance to doxorubicin and suppression of mdr1 gene activation by the cyclosporin PSC 833. J Natl Cancer Inst 1995;87: 1593602.
  • 27
    Bohme M, Buchler M, Muller M, et al. Differential inhibition by cyclosporins of primary-active ATP-dependent transporters in the hepatocyte canalicular membrane. FEBS Lett 1993;333: 1936.
  • 28
    Shepard RL, Law KL, Starling JJ, et al. Modulation of navelbine and mitoxantrone resistance by the P-glycoprotein modulator, LY335979. Presented at the 91st Annual Meeting of the AACR; April 1–5, 2000; San Francisco, CA, USA.