[Quality control of dose delivered by in vivo dosimetry: can one tolerance be used for all localizations?]. / Contrôle qualité de la dose délivrée par dosimétrie in vivo : un critère de tolérance unique peut-il satisfaire toutes les localisations ?

PURPOSE: In vivo dosimetry measurements are accepted when the difference between measured and calculated dose is under 5%. A statistical analysis has been conducted to determine whether this tolerance matched the clinical practice for the studied localizations: pelvis, thorax, head and neck, breast. MATERIALS AND METHODS: The technical characteristics of the detectors were checked before being used in clinical practice. Then an automatic statistical analysis was implemented using the 2450 in vivo dosimetry measurements obtained during 1 year. MAIN RESULTS: The global average is 1.10%, the standard deviation 2.46% and the percentage of out of level measurements 4.09%. By distinguishing the localizations, the 5% tolerance appeared to be too narrow for the breast localization. DISCUSSION/CONCLUSION: Several investigations were initiated to justify the modification of the tolerance for the breast localization. They highlighted an underestimation of the calculated dose when high beam angles are set: a new correction factor was defined to take account this error. A specific tolerance was also specified for the breast localization.

A series of Fas receptor agonist antibodies that demonstrate an inverse correlation between affinity and potency.

Receptor agonism remains poorly understood at the molecular and mechanistic level. In this study, we identified a fully human anti-Fas antibody that could efficiently trigger apoptosis and therefore function as a potent agonist. Protein engineering and crystallography were used to mechanistically understand the agonistic activity of the antibody. The crystal structure of the complex was determined at 1.9 Å resolution and provided insights into epitope recognition and comparisons with the natural ligand FasL (Fas ligand). When we affinity-matured the agonist antibody, we observed that, surprisingly, the higher-affinity antibodies demonstrated a significant reduction, rather than an increase, in agonist activity at the Fas receptor. We propose and experimentally demonstrate a model to explain this non-intuitive impact of affinity on agonist antibody signalling and explore the implications for the discovery of therapeutic agonists in general.

Adhesion, actin cytoskeleton organisation and the spreading of colon adenocarcinoma cells induced by EGF are mediated by alpha2beta1 integrin low clustering through focal adhesion kinase.

Both epidermal growth factor (EGF) and the extracellular matrix components have been implicated in the pathobiology of adenocarcinomas by somewhat poorly understood mechanisms. We have addressed this problem using an in vitro model comprising the colon adenocarcinoma cell line HT29-D4, wherein the role of EGF and type IV collagen on cell adhesion was examined. We demonstrated that the effect of EGF on HT29-D4 cell adhesion was regulated by type IV collagen in a time- and dose-dependent manner. The incorporation of a panel of monoclonal antibodies to integrins alpha1beta1, alpha2beta1 and alpha3beta1 in adhesion medium revealed that EGF-mediated increase in the cell adhesion was mediated essentially by alpha2beta1, and the use of flow cytometry led us to conclude that this EGF effect was mediated by an increase in alpha2beta1 activation and not by an increase in cell surface expression of integrin. An indirect immunofluorescence technique was employed to demonstrate that focal adhesion kinase (FAK) and alpha2beta1 integrin were present in focal complexes in large EGF-induced lamellipodia whereas actin cytoskeleton was organised in small tips that colocalised with FAK. This pattern was observed at early time points (15 min) with a strong FAK tyrosine phosphorylation and with an increase in mitogen-activated protein kinase activity (5-15 min) as measured by immunoprecipitation and immunoblotting. We conclude that at early time points of cell adhesion and spreading, EGF exerted an inside-out regulation of alpha2beta1 integrin in HT29-D4 cells. This regulation seemed to be mediated by EGF-dependent FAK phosphorylation entailing an increase in integrin activation and their recruitment in numerous focal complexes. Furthermore after activation, FAK induced aggregation of actin-associated proteins (paxillin, vinculin and other tyrosine phosphorylated proteins) in focal complexes, leading to organisation of actin cytoskeleton that is involved in lamellipodia formation. Finally, activated alpha2beta1 integrins intervened in all these processes clustered in small focal complexes but not in focal adhesions.

Repair of oxidized proteins. Identification of a new methionine sulfoxide reductase.

Oxidation of methionine residues to methionine sulfoxide can lead to inactivation of proteins. Methionine sulfoxide reductase (MsrA) has been known for a long time, and its repairing function well characterized. Here we identify a new methionine sulfoxide reductase, which we referred to as MsrB, the gene of which is present in genomes of eubacteria, archaebacteria, and eucaryotes. The msrA and msrB genes exhibit no sequence similarity and, in some genomes, are fused. The Escherichia coli MsrB protein (currently predicted to be encoded by an open reading frame of unknown function named yeaA) was used for genetic, enzymatic, and mass spectrometric investigations. Our in vivo study revealed that msrB is required for cadmium resistance of E. coli, a carcinogenic compound that induces oxidative stress. Our in vitro studies, showed that (i) MsrB and MsrA enzymes reduce free methionine sulfoxide with turn-over rates of 0.6 min(-1) and 20 min(-1), respectively, (ii) MsrA and MsrB act on oxidized calmodulin, each by repairing four to six of the eight methionine sulfoxide residues initially present, and (iii) simultaneous action of both MsrA and MsrB allowed full reduction of oxidized calmodulin. A possibility is that these two ubiquitous methionine sulfoxide reductases exhibit different substrate specificity.

Resistance to Taxol in lung cancer cells associated with increased microtubule dynamics.

Microtubule dynamics are crucial for mitotic spindle assembly and chromosome movement. Suppression of dynamics by Taxol appears responsible for the drug's potent ability to inhibit mitosis and cell proliferation. Although Taxol is an important chemotherapeutic agent, development of resistance limits its efficacy. To examine the role of microtubule dynamics in Taxol resistance, we measured the dynamic instability of individual rhodamine-labeled microtubules in Taxol-sensitive and -resistant living human cancer cells. Taxol-resistant A549-T12 and -T24 cell lines were selected from a human lung carcinoma cell line, A549. They are, respectively, 9- and 17-fold resistant to Taxol and require low concentrations of Taxol for proliferation. We found that microtubule dynamic instability was significantly increased in the Taxol-resistant cells. For example, with A549-T12 cells in the absence of added Taxol, microtubule dynamicity increased 57% as compared with A549 cells. The length and rate of shortening excursions increased 75 and 59%, respectively. These parameters were further increased in A549-T24 cells, with overall dynamicity increasing by 167% compared with parental cells. Thus, the decreased Taxol-sensitivity of these cells can be explained by their increased microtubule dynamics. When grown without Taxol, A549-T12 cells were blocked at the metaphase/anaphase transition and displayed abnormal mitotic spindles with uncongressed chromosomes. In the presence of 2-12 nM Taxol, the cells grew normally, suggesting that mitotic block resulted from excessive microtubule dynamics. These results indicate that microtubule dynamics play an important role in Taxol resistance, and that both excessively rapid dynamics and suppressed dynamics impair mitotic spindle function and inhibit proliferation.

Polyamines secreted by cancer cells possibly account for the impairment of the human erythrocyte sodium pump activity.

Using an original microcalorimetric method, we previously showed that in erythrocytes from cancer patients, the sodium pump activity was decreased and returned to normal in patient in remission. In addition we suggested that a plasma-borne factor probably secreted by cancer cells accounted for this impairment of the sodium transporter. In the present study we sought to identify this factor as well as its mechanism of action. First we determined the effect of culture media from undifferentiated and differentiated colon cancer cell lines (Caco-2 and HT29-D4) on the sodium pump activity of normal human erythrocytes. The inhibitory powers of culture media from undifferentiated cells were higher than those of differentiated cells (38.6 +/- 3.5% vs 6.9 +/- 4.6%, p<0.05 for Caco-2 and 45.8 +/- 6.2% vs 9.0 +/- 5.0%, <0.05 for HT29-D4). The use of alpha difluoro-methylomithine (2 mM) to inhibit ornithine decarboxylase, the rate-limiting enzyme for polyamine biosynthesis, dramatically reduced the sodium pump inhibition induced by the two undifferentiated cell lines (75% for Caco-2 and 89% for HT29-D4). Polyamines secreted by undifferentiated cells and then taken up by human erythrocytes thus appeared as inhibitors of sodium pump of these red blood cells. Putrescine, spermidine, and spermine (the main polyamines) exerted a similar inhibitory effect (33 +/- 2%). Tested in vitro on Na,KATPase, these polyamines (3 mM) were inhibitors (putrescine = 23 +/- 2%; spermidine= 48 +/- 3%; spermine= 55 +/- 2%) when assay condition for the ATPase reaction was suboptimal (Na+ = 10 mM; K+ = 1 mM). The inhibitory effect appeared to be related to their charge and their aliphatic chain length. The effect of spermidine and spermine on the ionic substrates and ATP-Mg showed that molecules decreased the affinity (Km) of the Na,K-ATPase for Na+ (11.24 +/- 0.49 mM for control vs 23.51 +/- 1.53 mM for spermine and 18.86 +/- 0.98 mM for spermidine), indicating that polyamines exerted their inhibitory effect in a competitive manner.

Involvement of nuclear factor kappaB in c-Myc induction by tubulin polymerization inhibitors.

We showed previously that microtubule disassembly by vinblastine induces the proto-oncogene c-myc in epithelial mammary HBL100 cells. In this study, we demonstrate that vinblastine treatment in these cells, in contrast to what was observed with the colon adenocarcinoma cell line HT29-D4, activated the transcription factor NFkappaB, which has been involved in c-myc regulation. The microtubule disassembly also induced IkappaB degradation. Using transient transfection analysis, we show that the trans-activation of c-myc by vinblastine was decreased when NFkappaB binding sites on c-myc promoter were mutated. Additionally, we demonstrate that microtubule dissolution trans-activated a thymidine kinase-CAT construct containing an NFkappaB binding site at -1180 to -1080 bp relative to the P1 promoter. Thus, vinblastine up-regulates the enhancer activity of the NFkappaB binding site. These results suggest that microtubule disassembly induced by vinblastine can trans-activate the c-myc oncogene through NFkappaB. Taking into consideration the paradoxical roles of both c-myc and NFkappaB in proliferation or apoptosis, this data reveals the complex action mechanism of this microtubule interfering agent.

Hyperphosphorylation of tau is mediated by ERK activation during anticancer drug-induced apoptosis in neuroblastoma cells.

Phosphorylated tau protein is the major component of paired helical filaments in Alzheimer disease (AD). We have previously shown that abnormal tau phosphorylation was induced in neuroblastoma SK-N-SH cells by the anticancer drug, paclitaxel, during apoptosis [Guise et al., 1999: Apoptosis 4:47-58]. In the present study, we first demonstrated a shift from fetal tau to hyperphosphorylated tau after incubation with paclitaxel, that showed some similarities with the hyperphosphorylated tau in AD, by using several tau antibodies, N-Term, Tau-1 and AT-8. Tau phosphorylation occurred independently of caspase-3 activation. We next showed that a sustained activation of ERK (extracellular signal-regulated kinase) induced both tau phosphorylation and apoptosis during paclitaxel treatment (1 microM). The inhibition of ERK activation by using the pharmacological MEK1/2 inhibitor, PD98059 (50 microM), or an antisense strategy, reduced tau phosphorylation and neuronal apoptosis (P < 0.001), indicating a link between ERK activation, tau phosphorylation and apoptosis. Doxorubicin (0.2 microM), an anticancer drug whose mechanism of action is independent of microtubules, also induced ERK activation, tau phosphorylation and apoptosis. Moreover, doxorubicin induced some morphological features of neurodegeneration such as loss of neurites and disorganization of the cytoskeleton in apoptotic neuroblastoma cells. Altogether, our results suggest that tau phosphorylation plays a significant role in apoptosis enhancing disruption of microtubules that in turn leads to formation of apoptotic bodies, suggesting that neurodegeneration and apoptosis are related.

Caulerpenyne from Caulerpa taxifolia has an antiproliferative activity on tumor cell line SK-N-SH and modifies the microtubule network.

Caulerpenyne, the major secondary metabolite synthesized by the green marine alga Caulerpa taxifolia, is cytotoxic against several cell lines. To identify possible targets of this toxin, we investigated the effect of caulerpenyne on the neuroblastoma SK-N-SH cell line. Caulerpenyne induced an inhibition of SK-N-SH cell proliferation with an IC50 of 10 +/- 2 microM after 2 hr of incubation. We observed no blockage in G2/M phase and an increase in cell death. On immunofluorescence microscopy, caulerpenyne affected the microtubule network in SK-N-SH cell line; we observed a loss of neurites and a compaction of the microtubule network at the cell periphery. In vitro, after 35 min of incubation, caulerpenyne inhibited the polymerization of pig brain purified tubulin or microtubule proteins, with an IC50 of 21 +/- 2 microM and 51 +/- 6 microM respectively. Analysis by electron microscopy indicated that caulerpenyne induced aggregation of tubulin, which may be responsible for inhibition of microtubule polymerization and bundling of residual microtubules.

Outside-in regulation of integrin clustering processes by ECM components per se and their involvement in actin cytoskeleton organization in a colon adenocarcinoma cell line.

We investigated in a colon adenocarcinoma cell line, the exclusive role of extracellular matrix (ECM) components in the absence of soluble factors regarding the integrin clustering processes, and their implication in cell adhesion, spreading and organization of the actin cytoskeleton. Caco-2 cells were shown to express at the plasma membrane 11 integrins, some of which (e.g. alpha3beta1, alpha5beta1, alpha6beta1/beta4, alpha8beta1 and alpha(v)beta1/beta5/beta6) were identified for the first time in this cell line. Cell adhesion and spreading processes were governed essentially by lamellipodium, the regulation of which was shown to be induced by two types of integrin clustering processes mediated by ECM proteins alone. During these phenomena, alpha2beta1, alpha(v)beta6 and alpha6beta1 integrins, the Caco-2 cell specific receptors of type IV collagen, fibronectin and laminin, respectively, were clustered in small focal complexes (point contacts), whereas alpha(v)beta5, the vitronectin receptor in this cell line, was aggregated in focal adhesions. The two levels of integrin clustering induced only F-actin cortical web formation organized in thin radial and/or circular filaments. We conclude thus that ECM components per se through their action on integrin clustering are involved in cell adhesion, cortical actin cytoskeleton organization and cell spreading.

Caspase-8 activation independent of CD95/CD95-L interaction during paclitaxel-induced apoptosis in human colon cancer cells (HT29-D4).

Antimicrotubule agent-induced apoptosis was examined in the proliferating human colon cancer cell line HT29-D4. G2/M arrest and subsequent apoptosis were dose-dependent, both observed with 100 nM paclitaxel or docetaxel and 10 nM vinorelbine. Bcl-x(L) phosphorylation was observed simultaneously with mitotic block, then caspase-3 cleavage and poly(ADP-ribose) polymerase degradation were detected 48 hr later. By using both enzymatic assay and immunoblot detection of cleaved fragments, we showed that caspase-8, a central component of the CD95-induced apoptotic pathway, was significantly activated during paclitaxel exposure, contemporary to apoptosis occurrence. Caspase-8 activation and apoptosis were independent of CD95 ligation and evidenced only for concentrations inducing Bcl-x(L) phosphorylation and a decrease in mitochondria permeability. Similar results were obtained with docetaxel and vinca alkaloids. Thus, antimitotic drugs may induce apoptosis via caspase-8 activation independently of CD95/CD95-L. Caspase-8 may be a common mediator of anticancer drug-induced apoptosis that could represent a promising target for future therapies.

Paclitaxel induces release of cytochrome c from mitochondria isolated from human neuroblastoma cells'.

Paclitaxel is an antimicrotubule agent that induces mitotic block and apoptosis. We show for the first time that paclitaxel acts directly or mitochondria isolated from human cancer cells. In isolated yeast mito chondria, paclitaxel (15 microM) induced an 18% increase in the respiration rate, with no concomitant release of cytochrome c. In isolated neuroblas toma mitochondria, paclitaxel (10-100 microM) induced a 27-72% release o cytochrome c. Release was prevented by cyclosporin A, suggesting the involvement of the permeability transition pore. Doxorubicin did no induce cytochrome c release, whereas vinorelbine, another antimicrotu bule agent, did. Thus, antimicrotubule agents can directly affect mito chondria to induce apoptosis.

Opposite effects of antimicrotubule agents on c-myc oncogene expression depending on the cell lines used.

We investigated the expression of c-myc in HT29-D4, HBL100 and Caco-2 cells treated with microtubule stabilising (paclitaxel) or depolymerising agents (vinblastine, nocodazole). After induction by epidermal growth factor (EGF), c-myc expression decreased in HT29-D4 cells treated with all the antimicrotubule agents. In HBL100 and Caco-2, when microtubules were stabilised with paclitaxel, c-myc expression also decreased. In contrast, its expression increased after treatment with depolymerising agents. In both cell lines, we also observed that depolymerising agents alone induced c-myc expression whilst paclitaxel had no effect. This mRNA induction was confirmed at the protein level. In HT29-D4, no variation of c-myc expression was observed. Then, we showed that the increase of mRNA level was due to activation of gene transcription. These results indicate that modulation of c-myc expression varied depending on the cell lines used and the type of antimicrotubule agents. This work provides a potential link between the microtubular network and c-myc gene expression.

Differentiation of human colon cancer cells changes the expression of beta-tubulin isotypes and MAPs.

The human colon adenocarcinoma HT29-D4 cell line is an interesting model for studies on epithelial cell differentiation. Undifferentiated cells are malignant proliferating cells, whereas differentiated cells act like epithelial polarized cells. In the present study, we first characterized the action of taxoids on the microtubular network of HT29-D4 cells according to the state of differentiation. Microtubular bundles were found in undifferentiated cells but not in differentiated cells, even with 500-fold higher taxoid concentrations for 96 h. This finding led us to study changes in microtubules according to the polarity status of the cell. E-MAP-115 was expressed only in differentiated cells; expression of beta-tubulin isotypes was altered in them relative to undifferentiated cells. Classes I, II, III, IVa and IVb isotypes were expressed in both phenotypes; however, differentiated epithelial cells displayed a specific increase in class III beta-tubulin. Thus, the increase in expression of this beta-tubulin isotype in differentiated cells is not restricted to neuronal cells. Moreover, these expression changes may reflect a higher stability of microtubular network in differentiated cells, which may explain the lower activity of anti-microtubule agents, independently of the mitotic process. These results indicate that the composition of microtubules should be considered as one of the criteria involved in the response of tumour cells to chemotherapy with anti-microtubule agents.

Tau protein is involved in the apoptotic process induced by anti-microtubule agents on neuroblastoma cells.

Paclitaxel and docetaxel are potent anti-microtubule and antimitotic agents that induce apoptosis in bone marrow-derived cells and epithelial cells. This study examined apoptosis induced by anti-microtubule agents in the neuroblastoma SK-N-SH cell line with a special focus on tau protein which is one of the main Microtubule-Associated- Proteins (MAPs) in neuronal cells. In time, treatment with 1 microM paclitaxel successively induced formation of bundles, then pseudo-asters concomitantly with mitotic block and phosphorylation of bcl-2 (48 h), then phosphorylation of tau and externalization of phosphatidylserine at the early phase of apoptosis (72 h) and finally DNA fragmentation (96 h). Similar results were obtained with 0.5 microM vinorelbine. Paclitaxel induced a lower increase in tau phosphorylation in differentiated SK-N-SH/RA+ cells which are less sensitive to apoptosis. Moreover, doxorubicin whose mechanism of action is independent of microtubules also induced immunostaining of tau at 72 h treatment. In conclusion, our results on neuroblastoma cells show that overexpression of hyperphosphorylated tau is involved in the apoptotic process induced by anti-microtubule agents and may be extended to others cytostatic drugs. Thus, tau protein may play a role in the cellular events observed in neuroblastoma cells undergoing apoptosis.

Convergent effects of growth factors, hormones, and fibronectin are necessary for the enterocyte differentiation of a colon adenocarcinoma cell line (HT29-D4).

The aim of this work was to show in serum-free medium a convergent effect of physiological factors and extracellular matrix proteins on the differentiation process of enterocytes by taking as a model the HT29-D4 clone that has the feature of differentiating when subcultured in fetal bovine serum glucose-free medium. We show that triiodothyronine (T3) as well as insulin promotes limited cell growth and differentiation, whereas fibronectin or bovine serum albumin (BSA) induces cell growth and a low level of differentiation. However, insulin, T3, fibronectin, and BSA together with epidermal growth factor and transferrin promoted satisfactory growth and enterocyte morphology with epithelial electrophysiological properties in HT29-D4 cells. With these factors adequate protein targeting was achieved since cells apically expressed the carcinoembryonic antigen, and basolaterally transferrin and insulin receptors, beta 1 and alpha v beta 6 integrins, talin, vinculin, and focal adhesion kinase (FAK). Talin, vinculin, FAK, and alpha v beta 6 integrin, the fibronectin receptor, were clustered in focal contacts, which agrees with a possible role of fibronectin in final cell growth, the latter process mediating the final phase of differentiation. This level of differentiation can be maintained for a long time. Thus HT29-D4 cells appear to be a suitable model to study the implication of integrins in the differentiation process of human enterocytes.

The effect of combining antitubulin agents on differentiated and undifferentiated human colon cancer cells.

The cytotoxicity of sequential combinations of a taxoid [paclitaxel (TAX) or docetaxel (TXT)] with a vinca alkaloid [vinorelbine (NVB)] was compared in differentiated and undifferentiated HT29-D4 cells. Agents were titrated from low doses inducing no modification of microtubule network to high doses corresponding to the clinically relevant concentrations that block mitosis. For undifferentiated cells, the sequential combination NVB/TAX was more efficient than TAX/NVB (22% cell survival versus 37% for 5 nM TAX and NVB). Surprisingly, we successively obtained synergism for low doses of both compounds [NVB (1-5 nM) and TAX (1-15 nM)], then additivity and finally antagonism when one of the compounds was at the concentration inducing mitotic block. The three patterns of results were also obtained with NVB/TXT combinations. For the synergistic combinations at the lowest concentrations, cytotoxicity occurred by apoptosis following mitosis. For differentiated cells, the most cytotoxic combinations were 1 microM TAX or TXT for 3 days followed by 1 microM NVB for 3 days, and 0.75 nM TAX or TXT for 9 days followed by 1 microM NVB for 3 days, the latter producing synergistic effects. Cytotoxicity occurred by apoptosis for the two states of differentiation. Major differences depending on cell phenotype were demonstrated: low sensitivity of differentiated cells to antitubulin agents and the difference in apoptotic pathways since mitosis is not involved in differentiated cells.

Effect of microtubule disruption on cell adhesion and spreading.

Microtubules have been involved in a variety of cellular processes. In this study, we examined the role of the microtubular system in the adhesion and spreading of the adenocarcinoma cell line HT29-D4. Disruption of microtubules by nocodazole or navelbine resulted in an increase in cell adhesion to purified ECM proteins. This enhanced cell adhesion is mediated by integrins, but is not attributable to quantitative changes in the number of integrin receptors at the cell surface, as determined by flow cytometric analysis. In contrast to attachment, spreading of HT29-D4 cells was reduced by nocodazole treatment in a dose-dependent manner. Thus, microtubule depolymerization appears to increase initial attachment of cells to extracellular matrix, while impeding subsequent cell spreading.

Pretreatment by tubulin agents decreases C-MYC induction in human colon carcinoma cell line HT29-D4.

For HT29-D4 cell line, we confirmed the interaction between c-Myc protein and microtubules by immunoprecipitation. We then studied the effect of antimitotic agents, nocodazole and the taxoids [paclitaxel (taxol) and docetaxel (taxotere)] on c-myc oncogene expression. The expression was analyzed by RT-PCR and Western blot. Taxol (1 microM), taxotere (1 microM) and nocodazole (3 microM) inhibited by 30-50% the c-myc induction produced by growth factors in culture medium. According to the flow cytometry analysis, the inhibition is not linked to the mitotic block. These results observed for both stabilizing and depolymerizing agents suggest that microtubular system is involved in c-myc expression more through its dynamic properties which influence signal transduction and intracellular transports than through its direct interaction with c-Myc protein.

[Modification of the response to paclitaxel of human differentiated colon cancer line after long-term treatment at very low dose]. / Modification de la réponse au paclitaxel d'une lignée cancéreuse humaine de côlon différenciée après traitement continu à très faible dose.

Taxoids (paclitaxel, Taxol and docetaxel, Taxotere), drugs which stabilize microtubules, demonstrate marked activity against ovarian, brain and lung cancer. We investigated, on the human colon adenocarcinoma HT29-D4 cell line, firstly the effects of taxoids in function of the differentiation state and secondly the effects of a low dose of paclitaxel on the differentiation process. The cellular parameters studied were microtubular network, cell cycle and cell tubulin content. Undifferentiated cells treated with paclitaxel or docetaxel (10-100 nM) classically induced bundles in interphasic cells and pseudoasters in mitotic cells, arrest in cell cycle in G2M and increase in tubulin content. Inversely, no modification was observed in differentiated cells after similar taxoid treatment. Long-term low-dose pretreatment of differentiated cells restitutes some sensitivity to taxoids since these cells become responsive to further taxoid treatment, as reflected by modifications of the microtubule network.