Elevated Levels of Microtubule Destabilizing Factors in a Taxol-resistant/dependent A549 Cell Line with an α-Tubulin Mutation

Introduction

The development of clinical drug resistance poses a major obstacle for the survival of cancer patients. Despite promising initial responses to chemotherapy, many patients experience recurrence of the primary tumor and/or metastases. The mechanisms of resistance to Taxol and other microtubule-stabilizing agents that have been characterized previously in human cell lines include expression of the drug efflux pump, P-glycoprotein (1) , and mutations in the cellular target of Taxol, β-tubulin (2, 3, 4) . In addition, Taxol-resistant cell lines have demonstrated alterations in their β-tubulin isotype levels (5, 6, 7, 8, 9) . The role of these various forms of Taxol resistance, particularly their contribution to resistance in the cancer patient, is not clear.

Prior studies have examined Taxol resistance and dependence of the human non-small cell lung carcinoma cell lines, A549-T12 and A549-T24 (8 , 10 , 11) . The A549-T12 cell line is 9-fold resistant to Taxol and does not express P-glycoprotein. The A549-T24 cell line is 17-fold resistant to Taxol and does express low levels of P-glycoprotein. The accumulation of [³H]Taxol has been measured in A549-T24 cells, and no significant differences have been observed between the parental and Taxol-resistant cell lines, suggesting that P-glycoprotein does not play a major role in Taxol resistance in these cells. 5 The Taxol-resistant cell lines have demonstrated differences in mRNA expression levels of the β-tubulin isotypes. In addition, an increase of β-tubulin class III protein expression was measured in the A549-T24 cell line (8) . These results suggested that increased expression of specific β-tubulin isotypes might be associated with Taxol resistance. Furthermore, a second study demonstrated that decreased expression of β-tubulin class III by antisense treatment in A549-T24 cells resulted in a modest increase in sensitivity to Taxol (12) . The altered expression of β-tubulin isotypes in these cells may influence microtubule dynamics as has been shown to be true under in vitro conditions (13) , especially with regard to β-tubulin class III (14, 15, 16) .

In addition to the above changes, the A549-T12 and A549-T24 cell lines are dependent on Taxol for normal growth. When Taxol is removed, the cells displayed a diminished microtubule cytoskeleton and a block in the mitotic phase of the cell cycle that was reversed by the addition of 2 nm Taxol (10) . In the absence of Taxol, analysis of the microtubule dynamics in the A549 Taxol-resistant cell lines demonstrated an increase in dynamicity when compared with the A549 Taxol-sensitive cells (11) . This may be related to alterations in the expression of microtubule regulatory proteins such as stathmin and MAP4, 6 and of tubulin isotypes and/or mutations. In this paper the mutational status of α- and β-tubulin was evaluated, together with the expression of tubulin/microtubule-interacting proteins that are involved in microtubule stability.

Materials and Methods

Materials.

Taxol was obtained from the Drug Development Branch of the National Cancer Institute (Bethesda, MD), and dissolved in sterile DMSO and stored at −20°C. All of the other chemicals were obtained from Sigma (St. Louis, MO) except where noted.

Cell Culture.

A549, A549-T12 (grown in 12 nM Taxol) and A549-T24 (grown in 24 nM Taxol) were maintained as described previously (8) . The minimum requirement of drug for normal growth of A549-T12 was 2 nM, and for A549-T24, 4 nM.

Tubulin Polymerization Assay.

The ratio of soluble and polymerized tubulin in A549 cell lines was determined using an assay described previously (17) . Before the experiment, A549-T12 and A549-T24 cells were maintained for 1 week in 2 and 4 nm Taxol, respectively. The polymerized (P) and soluble (S) tubulin fractions were subjected to 12% SDS-PAGE, transferred to nitrocellulose, and probed with a monoclonal antibody to α-tubulin (1:1000; Sigma; clone DM1A) and a goat antimouse HRP-linked IgG secondary antibody (1:2000; Transduction Laboratories, San Jose, CA). All of the blots were visualized using enhanced chemiluminescence and quantitated by densitometry (Amersham Biosciences, Piscataway, NJ).

Sequencing of β-Tubulin Class I Isotype.

For RT-PCR and sequencing of the βI isotype in A549 cell lines, four overlapping sets of primers were designed based on the published sequence (GenBank accession no. AF070561) and previously published primer sequences (2) . Total RNA was isolated using Total RNA isolation reagent (ABgene, Rochester, NY) and contaminating DNA removed with RNase-free DNase I treatment (Boehringer Mannheim, Indianapolis, IN) for 30 min at 37°C. Total RNA (1 μg) was reverse transcribed and the cDNA used for RT-PCR. PCR-amplified products were purified using the QIAquick gel extraction kit (Qiagen, Valencia, CA). The amount of cDNA was quantified, and ∼200 ng of cDNA, with 6 pmol of either primer, forward or reverse, was sequenced using an automated DNA sequencing system (ABI Prism).

Sequencing of α-Tubulin Kα1 Isotype.

For RT-PCR and sequencing of the Kα1 isotype in A549 cell lines, four overlapping sets of primers were designed using the Oligo program (Molecular Biology Insights, Inc., Cascade, CO), based on the published sequence (GenBank accession no. K00558). In addition, a second forward primer was designed to confirm the presence of the Kα1 tubulin mutation (Table 1) ⇓ . The method described above for β-tubulin was performed.

High-Resolution IEF and Two-Dimensional Gel Analysis.

Cells were grown to subconfluency in eight 100-mm dishes, washed once with PBS, and then scraped using 1 ml of PBS. The cells were pelleted, yielding 0.3–0.4-ml packed cells. The cell pellet was resuspended in 1.5 volume of MME buffer {0.1 m 2-[N-morpholino] ethane sulfonic acid (pH 6.9), 1 mm MgCl₂, and 1 mm EGTA} and frozen in liquid nitrogen. At the time of use, cell suspensions were rapidly thawed and kept on ice. One-tenth volume of a 10-fold stock solution of a protease inhibitor mixture (tablets; Boehringer Mannheim) in MME buffer and 1 mm DTT were added. Cells were disrupted by sonication 7 × 30 s each with a 30-s rest interval on ice between sonications. The lysates were centrifuged at 120,000 × g (Beckman TLA100.3 rotor) for 1 h at 4°C. The supernatant (cytosolic extract) was transferred to a new tube, and the pellet was discarded.

Taxol-stabilized microtubule pellets were isolated after the method of Vallee (18) . In brief, the supernatant was incubated for 20 min at 37°C in the presence of 10 μm Taxol and 1 mm GTP. Reaction mixtures were transferred to new tubes containing 0.1 ml of a 5% sucrose MME cushion containing 10 μm Taxol and 1 mm GTP. Samples were centrifuged at 80,000 × g for 30 min at 37°C. The microtubule pellets were washed and resuspended in MME buffer containing 0.35 m NaCl and 10 μm Taxol. Microtubules were centrifuged at 80,000 × g for 30 min at 37°C, and the pellets were frozen on dry ice and kept at −70°C until use.

Microtubule pellets (containing approximately 100–200 μg protein) were resuspended in 350 μl of solubilization buffer {7 m urea, 2 m thiourea, 4% 3-[(3-cholamidopropyl) dimethylammonio]-1-propane sulfonate, 0.5% Triton X-100, 0.5% ampholyte-containing buffer (pH 4.5–5.5), 20 mm DTT, and bromphenol blue} and loaded onto IPG 18 cm (pH 4.5–5.5) strips (Amersham Biosciences), and run in an IPGphor IEF system. IPG strips were stored at −20°C, stained with acid violet 17 (19) , or equilibrated for 2D or electrotransfer onto nitrocellulose. Before running the 2D, the strips were incubated in equilibration buffer [50 mm Tris-HCl (pH 8.8), 6 m urea, 30% glycerol, 2% SDS, and bromphenol blue] for at least 30 min. The proteins on the IPG strips were resolved on 10% SDS-polyacrylamide gels and transferred to nitrocellulose. The blots were probed with α-tubulin (1:1000; Sigma; clone DM1A). For the electrotransfer method, strips were incubated twice in equilibration buffer [125 mm Tris-HCl (pH 6.8), 5% β-mercaptoethanol, and 1% SDS] for 20 min, then incubated in strip transfer buffer [25 mm Tris-HCl (pH 8.3) and 192 mm glycine] for 20 min. The strips were electrotransferred (∼20 h) onto nitrocellulose and probed with a rabbit anti-Kα1-tubulin polyclonal antibody (SRa1) or a rabbit anti-α6-tubulin polyclonal antibody (SRa6). The antibodies SRa1 and SRa6 were prepared using chemically synthesized α-tubulin COOH terminus peptides, GVDSVEGEGEE and GADSADGEDEG, respectively. A Cys residue was added at the NH₂ terminus of these peptides for their conjugation to maleimide-activated keyhole limpet hemocyanin. Rabbit bleeds were analyzed in our laboratory by ELISA, using the same Kα1- and α6-tubulin COOH-termini peptides conjugated to maleimide-activated BSA.

Stathmin and MAP4 Protein Analysis.

Cell lysates were prepared for determination of total stathmin as described above except that PEM buffer [100 mm PIPES (pH 6.8), 2 mm EGTA, and 2 mm MgCl₂] was used. The protein content was quantified using the Bio-Rad protein assay (Hercules, CA) and the samples subjected to 15% SDS-PAGE. Before the gels were transferred, the blots were probed with antistathmin and antiactin (polyclonal stathmin 1:2500; Calbiochem, San Diego, CA; monoclonal actin 1:1000; Sigma; clone AC-40) and either a donkey antirabbit HRP-linked IgG secondary antibody (for stathmin, 1:2000; Amersham Biosciences) or a goat antimouse HRP-linked IgG secondary antibody (for actin, 1:2000; Transduction Laboratories). The blots were scanned and then analyzed with ImageQuant software (Amersham Biosciences) to quantitate the protein levels.

Phosphorylated forms of stathmin were analyzed using a modified method for enrichment of stathmin (20) . The modifications were as follows: cell pellets (after treatment with or without Taxol for 16.5 h) were resuspended in lysis buffer [10 mm Tris-HCl (pH 7.4), 150 mm NaCl, 1 mm EDTA, 1 mm phenylmethylsulfonyl fluoride, 10 μg/ml leupeptin, 25 μg/ml aprotinin, 10 μg/ml pepstatin A, 10 mm NaF, and 0.5 μm okadaic acid] and sonicated twice for 30 s at 4°C. The lysates then were boiled for 10 min and centrifuged at 20,000 × g for 10 min. The supernatant was treated with DEAE cellulose for 30 min at 4°C, leaving the partially purified stathmin in the supernatant. This supernatant was diluted with loading buffer [40 mm Tris-glycine (pH 8.8) and 10% sucrose]. The samples were run at 40 V on a 3–9% nondenaturing polyacrylamide gel [upper buffer: 50 mm Tris-glycine (pH 8.9); lower buffer: 63 mm Tris and 50 mm HCl (pH 7.5)], transferred to nitrocellulose, and probed with antistathmin (polyclonal stathmin 1:500; generously provided by Dr. Ulrich Schubart, Albert Einstein College of Medicine, Bronx, NY) and a donkey antirabbit HRP-linked IgG secondary antibody (1:2000; Amersham Biosciences).

MAP4 was analyzed in cell lysates isolated by the procedure described above (IEF/2D gels). The samples were run on both 9% SDS-polyacrylamide gels and 5% polyacrylamide/2 m urea gels, and transferred to polyvinylidene fluoride membranes. The blots were probed with a polyclonal antibody to MAP4 (1:1000; generously provided by Dr. Jeannette C. Bulinski, Columbia University, New York, NY) and a donkey antirabbit HRP-linked IgG secondary antibody (1:2000; Amersham Biosciences). The blots were scanned to quantitate the protein levels and then analyzed with Quantity One software (Bio-Rad).

Modeling of the α-Tubulin Mutation.

Molecular modeling studies were performed using the Insight II software (Accelrys, San Diego, CA). The α-tubulin structure was taken from Nogales et al. (PDB code: 1TUB; Ref. 21 ).

Results

Tubulin Mutation Analysis of the Taxol-resistant Cell Lines Revealed an α-Tubulin Mutation.

β-Tubulin gene expression in the Taxol-resistant cell lines was analyzed by RT-PCR amplification and subsequent sequencing of β-tubulin class I, the predominant β-tubulin isotype in the A549 cell lines (22) . A comparison of the β-tubulin in A549 parental and Taxol-resistant cell lines did not reveal any differences in their nucleotide sequence compared with the published sequence (data not shown). To rule out the possibility of a mutation in α-tubulin, primers were designed for the Kα1 isotype (Table 1) ⇓ , the ubiquitously expressed α-tubulin isotype in A549 cell lines (22) . Sequencing of A549-T12 cells revealed a heterozygous point mutation at nucleotide 1204 (AGC→AGC/A) that was not present in A549 parental cells (Fig. 1A) ⇓ . The A549-T24 cell line displayed the same mutation (data not shown).

The Kα1-tubulin protein sequence obtained from sequencing the A549 parental cell line is shown in Fig. 1B ⇓ . If the α-tubulin mutation were expressed at the protein level, then an amino acid substitution (Ser to Arg) would occur at position 379. This Ser residue, indicated in yellow, is located at the end of β-sheet 10.

The Mutant Kα1 Tubulin Is Expressed in the Taxol-resistant Cell Lines.

Because the point mutation observed in the Taxol-resistant cell lines was heterozygous, it was necessary to demonstrate the presence of the mutant tubulin at the protein level. The use of high-resolution IPG strips, with a pH range of 4.5–5.5, allowed the determination of the α-tubulin isotypes expressed in the A549 cell lines (Fig. 2A) ⇓ . The A549 parental and Taxol-resistant cell lines exhibited two α-tubulin bands, corresponding to Kα1- and α6-tubulin, based on the predicted pI for Kα1- (4.94) and α6-tubulin (4.96; Fig. 2A ⇓ , strips 1 and 2). Moreover, the SRa6 antibody directed against the COOH terminus of α∗ (22) labeled a single band at the exact position of α6 on the IPG strip blot (Fig. 2A ⇓ , strips 3 and 5). The sequence of human α6-tubulin was entered recently in the National Center for Biotechnology Information protein database (accession no. Q9BQE3), and the COOH terminus sequence of α6-tubulin is identical to the COOH terminus of α∗-tubulin; therefore, we suggest that α∗-tubulin is human α6-tubulin. The third α-tubulin band observed in the A549-T12 Taxol-resistant cell line matches the pI value of 4.98 for the mutant Kα1-tubulin (Ser379Arg; Fig. 2A ⇓ , strip 2). The A549-T24 cell line displayed the same three α-tubulin bands (data not shown). All three of the α-tubulins were labeled with the SRa1 antibody directed against the Kα1-tubulin COOH terminus on the IPG strip blot (Fig. 2A ⇓ , strips 4 and 6). The 2D blots of the A549 cell lines demonstrated spots at the expected molecular weight of M_r 50,000 (Fig. 2B) ⇓ . An additional spot (Fig. 2 ⇓ A, strip 6, arrowhead; pI = 4.91), which corresponds to monoglutamylated Kα1-tubulin (22) , was detected on the electrotransferred strips. Quantitation of the different α-tubulin species (Fig. 2C) ⇓ on acid violet 17 stained strips did not demonstrate any significant differences in expression of α6-tubulin in A549 and A549-T12 cell lines (∼ 30% of total α-tubulin). The decrease in the Kα1-tubulin level in A549-T12 (∼ 30%) corresponded to the expression of the mutant Kα1-tubulin.

Normal in Vitro Tubulin Polymerization in the Presence of Taxol Is Observed in the A549 Cell Lines.

The ability of tubulin from Taxol-resistant cell lines to polymerize in the presence of drug was assayed. This method had been used to study two Taxol-resistant cell lines with mutant β-tubulins that exhibited impaired tubulin polymerization in the presence of increasing concentrations of Taxol (2) . In the A549-T12 and A549-T24 cells, there was no significant change in the ability of tubulin to assemble in the presence of Taxol (data not shown).

Stathmin and MAP4 Protein Expression Is Altered in the Taxol-resistant Cell Lines.

The protein levels of stathmin, a tubulin-interacting protein, were determined in the A549 cell lines and an increase of ∼2-fold was found in the Taxol-resistant cell lines compared with the A549 parental cells (Fig. 3A) ⇓ . In addition to total stathmin levels, its phosphorylation status was analyzed. Stathmin is a phosphoprotein of which the microtubule-destabilizing activity is controlled by phosphorylation (23 , 24) . When stathmin is hyperphosphorylated, the protein cannot bind to tubulin and perform its normal destabilizing function. A549 cells exposed to increasing concentrations of Taxol exhibited a shift in expression from the unphosphorylated form of stathmin (P0) to the fully phosphorylated form (P4), thus resulting in inactivation of stathmin (Fig. 3B) ⇓ . In contrast, A549-T12 or A549-T24 cells, in the presence of increasing concentrations of Taxol, demonstrated little change in the phosphorylation status of stathmin.

The protein expression of MAP4, a MAP that stabilizes microtubules, was determined in the A549 cell lines (Fig. 4) ⇓ . When MAP4 is phosphorylated, the protein does not bind to microtubules nor promote stabilization of the microtubule network (25 , 26) . In the Taxol-resistant cell lines, changes were observed in MAP4 compared with the A549 parental cells. In 9% SDS gels, three bands were detected, the lowest molecular weight corresponding to nonphosphorylated MAP4 and the two upshifted bands corresponding to phosphoisoforms (Fig. 4A) ⇓ . Increased resolution of MAP4 isoforms on 5% urea gels confirmed that the lowest band represented nonphosphorylated MAP4 with a molecular weight of M_r ∼200,000 (Fig. 4B) ⇓ . The unmodified form of MAP4 was decreased in the Taxol-resistant cell lines, as revealed by the change in percentage intensity of band 1 as compared with the A549 parental cells. The data demonstrate that there is a progressive shift to expression of the hyperphosphorylated forms of MAP4 as the level of resistance increases.

Discussion

Both of the A549 Taxol-resistant cell lines, A549-T12 and A549-T24, are dependent on low concentrations of Taxol for normal growth (8 , 10) . In an attempt to uncover the molecular basis of this combined phenotype, the dynamicity of individual microtubules was measured in the absence of Taxol. A549-T24 cells displayed the most dramatic increase in microtubule dynamicity, and A549-T12 cells exhibited increased microtubule dynamicity but to a lesser degree, when both were compared with the A549 parental cells (11) . These studies emphasized the importance of a narrow range of microtubule dynamics for normal cell proliferation and suggested that the possible mechanism for Taxol resistance/dependence involved an alteration in microtubule dynamicity. Such alterations could include mutations in α-tubulin or β-tubulin affecting lateral/longitudinal interactions in the microtubule lattice and/or the binding of regulatory proteins that could result in more dynamic microtubules. These microtubules would require Taxol to restore their normal dynamicity and cellular function, thus helping to explain the drug dependency of the cells.

In the present study, we did not find a mutation in the major β-tubulin isotype, βI-tubulin, which is expressed in the A549 cell lines. Furthermore, tubulin from the Taxol-resistant cell lines exhibited normal Taxol-dependent polymerization, indicating that the binding of Taxol to the microtubule was not significantly compromised. One of the possible mechanisms of resistance to Taxol that has not been explored in great detail is a mutation in α-tubulin. A heterozygous mutation in the major α-tubulin isotype, Kα1, was identified in these cell lines. The protein expression of this mutation, which is located at the end of β-sheet 10, was demonstrated (Fig. 1B ⇓ and Fig. 5 ⇓ ). Although electrophoretic mobility shifts in α-tubulin have been reported in Taxol-resistant cells (27 , 28) , these α-tubulin isoforms were not additionally characterized. In plants, studies have demonstrated α-tubulin mutations that conferred resistance to herbicides (29, 30, 31) , and a recent abstract has suggested that a mutation in α-tubulin is a factor in resistance to a hemiasterlin analogue, a microtubule depolymerizing agent (32) .

Alanine-scanning mutagenesis of yeast α-tubulin revealed regions that affected the sensitivity of yeast to benomyl, an antifungal agent that like Taxol has a binding site on β-tubulin (33) . The study demonstrated that mutations in α-tubulin could modulate sensitivity to a drug that interacted with β-tubulin. Mutagenesis at positions α373 and 374 resulted in mutant yeast supersensitive to benomyl, whereas mutations at positions α387 to 391 yielded benomyl-resistant yeast. In the A549-T12 cells, the mutation identified at Ser 379 corresponds to yeast α-tubulin Ser 380, and is positioned between the sensitive and resistant loci in yeast. Richards et al. (33) speculated that mutations in yeast α-tubulin could be altering the binding of regulatory proteins, thereby affecting microtubule stability and benomyl sensitivity.

In the α-tubulin structure, the mutation at α379 in the Taxol-resistant cells is situated between helix 10, which has been shown to interact with stathmin (34 , 35) , and helix 11, which is part of the putative domain interacting with MAPs (Fig. 1B ⇓ and Fig. 5 ⇓ ; Ref. 21 ). Therefore, a mutation at α379 may affect the binding of stathmin and/or MAP4 to tubulin/microtubules, and, thus, alter the stability of the microtubule network. For this reason, we examined the expression of these two well-characterized microtubule assembly regulatory proteins, stathmin, a microtubule destabilizer, and MAP4, a microtubule stabilizer. Stathmin is up-regulated in breast carcinoma cells from patients with more aggressive disease (36) , similar to the increase we observed in the Taxol-resistant cell lines. Therefore, one hypothesis is that the α-tubulin mutation may result in altered binding of stathmin to α-tubulin, and combined with the increased protein expression of stathmin, could lead to hypostable microtubules because of the role of stathmin in sequestering tubulin dimers (37) . The Taxol-resistant cells would be less affected by the hyperstabilizing activity of Taxol because of the combined effects of the α-tubulin mutation and the stathmin changes.

Stathmin is a phosphoprotein of which the activity is controlled by distinct sites of phosphorylation (23 , 24) . In A549 parental cells, stathmin is phosphorylated in the presence of increasing concentrations of Taxol, whereas the phosphorylation status of stathmin in A549-T12 and A549-T24 cells is relatively unaffected by exposure to Taxol. Two pairs of serine residues are differentially phosphorylated in stathmin by distinct kinases during the cell cycle. Serine 25 and 38 appear to be targeted by cdc2 kinase, and their phosphorylation seems to be a prerequisite for the phosphorylation of Ser 16 and 63 by a still-unknown phosphorylation process (38) . Phosphorylation of Ser 16 and 63 inhibit the microtubule destabilizing activity of stathmin. The monophosphorylated form (P1) of stathmin has been reported to be present in interphase HeLa cells and to be mainly composed of stathmin with phosphorylated Ser 38 (38) . We observed the P1 form of stathmin in A549-T12 and A549-T24 cells under normal cell culture conditions, i.e., in the presence of 12 nm and 24 nm Taxol, respectively (Fig. 3B) ⇓ . In this environment, A549-T12 and A549-T24 cells showed no difference in their DNA content (i.e., normal cell cycle profile) when compared with untreated A549 cells as analyzed by flow cytometry, 7 indicating that most of the resistant cells were in interphase. This suggests that the P1 form of stathmin is present throughout the cell cycle and results from phosphorylation of stathmin by cdc2 associated with cyclins A and/or E during interphase. When A549 cells were treated with increasing concentrations of Taxol, they accumulated in the mitotic phase with a concomitant increase in highly phosphorylated forms of stathmin (P3 and P4). In both Taxol-resistant cell lines, 50% of the cells were blocked in mitosis when treated with 100 nm Taxol for 48 h. 7 Although P3 and P4 inactive forms of stathmin were present, they represented only minor species compared with P0 and P1. These results suggested that the kinases and/or phosphatases, such as PP2A, involved in stathmin phosphorylation or dephosphorylation may be altered in Taxol-resistant cell lines compared with Taxol-sensitive A549 cells. Consequently, the 2-fold increase in stathmin expression that was observed in A549-resistant cell lines (Fig. 3A) ⇓ represented an increase in the active microtubule-destabilizing form of stathmin.

MAP4 is the predominant non-neuronal MAP (39) , and the microtubule-stabilizing function of MAP4 also is regulated by phosphorylation (25) . The decrease in expression of the microtubule-stabilizing form of MAP4, in addition to the increased levels of active stathmin, may cause additional destabilization of the microtubule network in the Taxol-resistant cell lines by affecting microtubule dynamics, as suggested by Chang et al. (26) . The appearance of phosphorylated forms of MAP4 in A549-T12 and A549-T24 cell lines are independent of mitotic arrest as noted above. MAP4 can be phosphorylated by cdc2 during interphase at Ser 696 and at Ser 787 during mitosis (40) . Other kinases phosphorylate MAP4 at different sites that also are involved in the binding of MAP4 to microtubules (41 , 42) . The continuous presence of Taxol in the cell culture medium of the A549-resistant cell lines may activate pathways involving these kinases independently of mitotic arrest. The increase in phosphorylated MAP4 species in A549-T12 and A549-T24 cells correlates with an increase in dynamicity of microtubules and resistance to Taxol (11) . Interestingly, Ookata et al. (43) described an interaction of cdc2 kinase with MAP4 that could target cdc2 to microtubules, and Kuntziger et al. (44) described a microtubule-dependent phosphorylation of stathmin at Ser 16. These findings may provide an integrative mechanism to explain the differences in phosphorylation status between stathmin and MAP4 that were observed in the Taxol-resistant A549 cell lines. A decrease in affinity of MAP4 for microtubules could decrease the cdc2-dependent inactivation of stathmin. Additional studies have demonstrated the regulation of Tau phosphorylation by PP2A and thereby binding of Tau to microtubules (45 , 46) . It is possible that MAP4 phosphorylation also may be regulated by PP2A.

Yet another factor to be considered is the role of p53 in stathmin and MAP4 expression. Both stathmin and MAP4 appear to be negatively regulated by wild-type p53, and when induced by DNA damaging agents, the mRNA and protein levels of stathmin (47 , 48) and MAP4 (49) are decreased. Furthermore, the activation of p53 also leads to differential sensitivity to microtubule-binding agents, such as sensitivity to vinblastine and resistance to Taxol (50) . The reverse effects by Taxol and vinblastine are observed when p53 is mutated (51) . A recent study involving breast cancer patients determined the activation of p53, as well as the expression of MAP4 after DNA damage and treatment with antimitotic agents (52) . The data demonstrated that in most patients with active wild-type p53, MAP4 expression was repressed. The sequential administration of a DNA-damaging agent and then of an antimicrotubule drug, such as vinorelbine, resulted in enhanced sensitivity to the treatment as had been shown in previous in vitro studies (50) .

This report documents, in mammalian cells, the importance of investigating mutations in α-tubulin for modulating sensitivity of microtubules to Taxol. We postulate that this occurs through a combination of events, one of which, the binding of regulatory proteins to tubulin/microtubules, could be modulated by a tubulin mutation. In addition, alterations in the phosphorylation of MAP4 and stathmin also are likely to affect the dynamics of microtubules. All of these findings point to the extreme complexity of Taxol-dependent/resistant cells and, in turn, highlight additional targets that need to be examined in patients treated with antimicrotubule agents.