Abstract
Peripheral neuropathy (PN) is a major dose-limiting side-effect of microtubule-targeting agents (MTAs), considered to be induced by inhibition of axonal microtubules. Therefore, it was thought that a useful method for predicting the frequencies of severe sensory-PN (FPN) would be to evaluate the neurite-disrupting effects of MTAs. Using neurite outgrowth from neuron-like cell lines, we comprehensively evaluated the neurite-disrupting effects of several anti-cancer drugs including MTAs, and the reversibility of the effects of MTAs. MTAs that induce PN showed neurite-disrupting effects more strongly than MTAs and anticancer drugs that do not induce PN, but the effects were not related to the FPN. On the other hand, MTAs with high FPN exhibited lower reversibility than those with low FPN. These findings suggest that neurite-disrupting effects are associated with the incidence of PN, and the reversibility of the effects is associated with FPN.
- Peripheral neuropathy
- microtubule-targeting agents
- neurite-distupting effects
- in vitro model
- PC-12 cells
- SH-SY5Y cells
Microtubules are tubular polymers composed of α/β-tubulin heterodimers. The continuous equilibrium of microtubule assembly and disassembly makes the microtubules dynamic structures that maintain cell shape, polarity, and motility; provide a scaffold for cellular trafficking of proteins and organelles; and play an integral role in mitosis. Because microtubules and their dynamics are required for mitotic spindle formation and chromosome separation during mitosis, they are thought to be an important target for a chemically diverse group of anticancer drugs that induce mitotic arrest and cell death in vitro.
Microtubule-targeting agents (MTAs) have been clinically used since the 1960s for the effective treatment of leukemia, lymphoma, and various solid tumors (1). MTAs are classified roughly into the Vinca alkaloids (vincristine, vindesine, vinorelbine) that de-stabilize or depolymerize microtubules, and the taxanes (paclitaxel, docetaxel) that polymerize microtubules, thus causing mitotic arrest. Meanwhile, MTAs often induce neurotoxicities by having effects on neurons. Although it is still an open question how neurotoxicities are clinically affected by MTAs, physiological and histopathological studies in pre-clinical research showed this is due, at least in part, to the inhibition of axonal transport in neurons (2). Because of the excessive length of peripheral axons, MTAs are especially disruptive to axons forming peripheral nerves and, therefore, induce peripheral neuropathy (PN). Generally, MTAs induce sensory PN more frequently than motor PN (3). Severe sensory-PN (grade 3/4 according to the National Cancer Institute Common Toxicity Criteria scale) can be a dose-limiting factor of MTAs and significantly reduce QOL.
In Table I, we summarize the frequencies of severe sensory-PN (FPNs) of MTAs in clinical trials. Vincristine and paclitaxel showed especially high FPNs (4-9). FPNs of docetaxel, vindesine, and vinorelbine are lower than those of vincristine and paclitaxel (10-16). Indibulin, now in a phase II clinical trial, was reported not to induce severe PN (17). These lines of clinical evidence indicate that FPNs are characteristic to each MTA, so it is desirable to discover MTAs with relatively low FPNs. To date, FPNs have been evaluated mainly using in vivo models in non-clinical studies (18); therefore, the establishment of an in vitro model in screening for drug discovery is anticipated.
MTAs are considered to induce PN by interacting with axonal microtubules and then disrupting neurites. On the basis of this principle, in terms of in vitro methods for predicting FPNs, in order to evaluate the neurite-disrupting effects of MTAs, it was thought to be useful to use neurite outgrowth evaluation from the PC-12 rat pheochromocytoma cell line differentiated into neuron-like cells or the SH-SY5Y human neuroblastoma cell line (19, 20). However, there have been no reports regarding comprehensive evaluations of various MTAs using these methods, so it was unclear how accurately the neurite-disrupting effects reflect FPN. Meanwhile, it was reported that the FPNs of MTAs became lower in proportion to the dosing interval in clinical trials (3). We, therefore, hypothesized that neurites disrupted by MTAs re-extended reversibly between dosages, which contributed to the recovery from PN. In the present study, we first evaluated the neurite-disrupting effects of MTAs using known methods and then examined how well the effects correlated with FPN. Furthermore, we established a novel in vitro model for evaluating the reversibility of the neurite-disrupting effects and examined the correlation between this reversibility and FPN.
Frequencies of peripheral neuropathy of microtubule-targeting agents in clinical trials.
Materials and Methods
Materials. Paclitaxel, vincristine, vindesine, and vinorelbine were purchased from Sigma Aldrich (St. Louis, MO, USA). Docetaxel was purchased from Fluka (Buchs, Switzerland). Indibulin was purchased from Tocris Bioscience (Bristol, UK). Doxorubicin was purchased from Wako Pure Chemical Industries, Ltd. (Tokyo, Japan). SN-38 and BI2536 were chemically synthesized by Yakult Honsha (Tokyo, Japan). These drugs were dissolved in dimethylsulfoxide (DMSO). The final concentration of DMSO in all treatments was adjusted to 0.1%.
Cells and cell cultures. The PC-12 rat pheochromocytoma cell line was obtained from Riken Cell Bank (Tsukuba, Japan) and cultured in Dulbecco's modified Eagle's medium (DMEM; Life Technologies, Grand Island, NY, USA) with 10% (v/v) fetal bovine serum (FBS) (Sigma-Aldrich) and 5% (v/v) horse serum (HS; MP Biomedicals, Aurora, OH, USA) at 37°C in 5% CO2. HCT116 human colon cancer and SH-SY5Y human neuroblastoma cell line were both obtained from American Type Culture Collection and grown in RPMI1640 medium (Life Technologies) with 10% FBS at 37°C in 5% CO2.
Induction of neurite outgrowth and drug treatment. PC-12 cells were seeded in 96-well plates coated with collagen (2,000 cells/well, Thermo Fisher Scientific, Nepean, ON, Canada). After 24 h, the medium was changed to low-serum (1% HS) medium, and PC-12 cells were treated with 50 ng/mL nerve growth factor (NGF; Merck Millipore, Nottingham, UK) for 72 h to induce neuronal differentiation and neurite outgrowth. Differentiated cells were then treated for 24 h with test drugs at several concentrations in the presence of NGF.
SH-SY5Y cells were seeded in 96-well plates for 24 h, followed by culture for 72 h with 1 μM all-trans retinoic acid (ATRA) (Wako Chem. Co.) to induce neurite outgrowth. Then, the SH-SY5Y cells were treated for 24 h with test drugs at several concentrations in the presence of ATRA.
Assay for neurite-disrupting effects and reversibility of these effects. For evaluation of the neurite-disrupting effects, 300 cells in triplicate wells were randomly chosen and the proportions of neurite-forming cells (neurite length >2 × cell body length) were determined. Using the same cells, cell viability was assessed with WST-8 (Kishida Chemicals, Osaka, Japan) or CellTiter 96 Aqueous One Solution (Promega, Madison, WI, USA). The proportions of neurite-forming cells were corrected using the following equation: (Correction value for the proportion of neurite-forming cells)=(proportion of neurite-forming cells scored by the evaluation of neurite-disrupting effects)/(cell viability).
Thereafter, the concentrations of each test drug causing 50% inhibition (IC50) of neurites were calculated from the correction values.
For evaluating the reversibility of the neurite-disrupting effects, after drug treatment, PC-12 and SH-SY5Y cells were cultured without test drugs for another 48 h in DMEM containing NGF (PC-12 cells) or RPMI1640 medium containing ATRA (SH-SY5Y cells). After the culture, the proportions of neurite-forming cells were scored by the method mentioned above.
Cytotoxicity assay for cell survival. HCT116 cells were seeded in 96-well plates and test drugs were added to the cultures at several concentrations. After 96 h, cell viabilities were measured with WST-8 and the concentrations of test drugs equivalent to the IC50 of cell viabilities were calculated.
Correlations between the proportion of neurite-reextending cells after drug treatment and FPN were determined by calculating r2 as a measure of the goodness of fit for their linear regression.
Microscopic images of NGF-differentiated PC-12 cells treated with test drugs. PC-12 cells were seeded in collagen-coated 96-well plates and treated with NGF for 72 h. The differentiated cells were cultured for another 24 h with DMSO (control), 100 nM vincristine, 100 nM paclitaxel, 100 nM docetaxel, 100 nM vindesine, 100 nM vinorelbine, 100 nM indibulin, 10 μM SN-38, 10 μM doxorubicin, or 10 μM BI2536 in the presence of NGF, and were then observed under a microscope.
Results
Neurite-disrupting effects. We evaluated the neurite-disrupting effects of MTAs (vincristine, paclitaxel, docetaxel, vindesine, vinorelbine and indibulin) using NGF-differentiated PC-12 rat pheochromocytoma cells. As a negative control, we additionally evaluated the following anticancer drugs, not affecting microtubule dynamics directly: SN-38 (topoisomerase I inhibitor), doxorubicin (topoisomerase II inhibitor), and BI2536 (polo-like kinase I inhibitor). Microscopic images of the PC-12 cells after drug treatment are shown in Figure 1. Radio et al. reported that neurites are disrupted as a result of not only microtubule destruction induced by MTAs, but also cytotoxicity induced by some drugs (21). In order to exclude such non-specific effects, we corrected the proportions of neurite-forming cells using cell viability after drug treatment (Figure 2). Vincristine, paclitaxel, docetaxel, vindesine, vinorelbine, and indibulin almost completely disrupted neurites at 10 to 1,000 nM. On the other hand, SN-38, doxorubicin, and BI2536 tended to decrease the proportions of neurite-forming cells concentration-dependently, but did not disrupt neurites completely even at 10 μM, the highest concentration tested in this experiment. When discussing the safety margin of drugs, it is generally necessary to evaluate their active and toxic concentrations. Therefore, we measured the anti-proliferative effects of each test drug on HCT116 cells as an index of their anti-cancer activities (Table II). From the IC50 for (a) neurites of PC-12 cells and (b) cell proliferation of HCT116 cells, values of a/b (mentioned as index values of neuropathy: IVNs) were calculated. Drugs with low IVN are considered to show both anticancer activities and neurite-disrupting effects, and therefore to have a narrow safety margin. Vincristine, paclitaxel, docetaxel, vindesine, and vinorelbine all showed similar IVNs regardless of their different FPNs. Indibulin, a MTA not inducing severe PN, showed clearly higher IVN than the other MTAs. SN-38, doxorubicin, and BI2536 showed even higher IVNs than indibulin, although these could not be calculated because of the lack of their neurite-disrupting effects. We similarly evaluated neurite-disrupting effects using SH-SY5Y human neuroblastoma cells using the IC50 for neurites, and the IVNs of each test drug evaluated are shown in Table III. Index values of neuropathy of vincristine, paclitaxel, docetaxel, vindesine, and vinorelbine were lower than those of doxorubicin and BI2536, which agreed with the results with PC-12 cells. On the other hand, among MTAs, vincristine, paclitaxel, and vinorelbine showed relatively high IVNs.
Effects of test drugs on neurites and cell viability of differentiated PC-12 cells. Differentiated PC-12 cells were evaluated for both neurites and cell viability following 24 h of exposure to DMSO (control), vincristine, paclitaxel, docetaxel, vindesine, vinorelbine, indibulin, SN-38, doxorubicin, or BI2536 at several concentrations. Proportion of neurite-forming cells were corrected as described in Materials and Methods. Data are expressed as percentages of controls, and are presented as means±SD from triplicate wells.
Neurite-reextending PC-12 cells after treatment of MTAs. Differentiated PC-12 cells were treated with DMSO (control), 10 nM vincristine, 1000 nM paclitaxel, 10 nM docetaxel, 10 nM vindesine, or 10 nM vinorelbine for 24 h followed by withdrawal of the MTAs from the culture medium for 48 h, and then the rates of neurite-forming cells were evaluated. Data are expressed as percentages of the control, and are presented as means±SD from triplicate wells (*p<0.05, **p<0.01). Significantly different from 10 nM vincristine-treated group.
Neurite-reextending SH-SY5Y cells after treatment with MTAs. SH-SY5Y cells were cultured with ATRA for 72 h following exposure to DMSO (control), 1 μM vincristine, 10 μM paclitaxel, 1 μM docetaxel, 0.1 μM vindesine, or 0.1 μM vinorelbine for 24 h. Then, the MTAs were removed from the culture medium for 48 h and the proportions of neurite-forming cells were evaluated. Data are expressed as percentages of the control, and are presented as means±SD from triplicate wells (**p<0.01). Significantly different from 1 μM vincristine-treated group.
Reversibility of neurite-disrupting effects. In order to evaluate the reversibility of the neurite-disrupting effects, we measured the proportions of neurite-reextending PC-12 cells after treatments of some MTAs, followed by their withdrawal (Figure 3). The MTAs were applied at the lowest concentration at which they disrupted the neurites of more than 95% of the cells, so as to reduce the non-specific effects caused by cytotoxicity. Vincristine and paclitaxel showed similar reversibility. Docetaxel, vindesine, and vinorelbine showed about 2-fold higher reversibility than vincristine. We similarly evaluated the reversibility of some MTAs using SH-SY5Y cells (Figure 4). Vincristine and paclitaxel showed similar reversibility, whereas the reversibility of docetaxel, vindesine, and vinorelbine was approximately 2- to 2.5-fold higher than that of vincristine. These results showed that the reversibility of the neurite-disrupting effects was similar using either PC-12 or SH-SY5Y cells. By surveying FPN and the reversibility of the neurite-disrupting effects across MTAs, we uncovered negative correlations between them in both PC-12 and SH-SY5Y cells (Figure 5; correlation coefficients r2=0.79 and 0.93, p=0.093 and 0.044, respectively).
Test drugs evaluated for effects on neurite of PC-12 cells and proliferation of HCT116 cells.
Test compounds evaluated for effects on neurites on SH-SY5Y cells.
Discussion
In the present study, with the aim of establishing an in vitro model for predicting the incidence and severity of MTA-induced PN, we evaluated neurite-disrupting effects and the reversibility of the effects induced by clinically used MTAs, and then examined the correlations between FPN and neurite-disrupting effects, or reversibility.
Correlations between reversibility of neurite-disrupting effects and FPN across MTAs in PC-12 and SH-SY5Y cells. The figure shows bivariate correlations between FPN and the proportion of neurite-reextending cells after drug withdrawal across MTAs, in PC-12 (filled square and solid line) and SH-SY5Y cells (open square and broken line). The correlation coefficient (r2) and significance (p) are indicated.
Using NGF-differentiated PC-12 rat pheochromocytoma cells, we measured the neurite-disrupting effects of various MTAs (vincristine and paclitaxel, which show relatively high FPNs (4-9); docetaxel, vindesine, and vinorelbine, which show relatively low FPNs (10-16); and indibulin, which does not induce severe PN (17)) and anticancer drugs not directly affecting microtubule dynamics (SN-38, doxorubicin, and BI2536) by calculating IVN as an index of the safety margin. The neurite-disrupting effect of indibulin was markedly lower than those of the other MTAs, which could induce PN. In addition, SN-38, doxorubicin, and BI2536 did not show neurite-disrupting effects at all. These results indicated that the neurite-disrupting effects against PC-12 cells reflect the incidence of PN. However, the neurite-disrupting effects were considered not to be associated with FPN, since the MTAs that could induce PN showed equivalent neurite-disrupting effects to each other. We similarly evaluated the neurite-disrupting effects using SH-SY5Y human neuroblastoma cells. The neurite-disrupting effects of vincristine, paclitaxel, docetaxel, vindesine, and vinorelbine were not correlated with their FPNs, but were clearly lower than those of doxorubicin and BI2536. Therefore, it was suggested that the neurite-disrupting effects observed in not only rat but also human neuroblastoma cell lines reflect the incidence of PN, but not FPN
We established a novel in vitro model to evaluate the reversibility of the neurite-disrupting effects of MTAs using PC-12 or SH-SY5Y cells, and examined the correlations between the reversibility measured with these models and FPN. As a result of evaluations with both PC-12 and SH-SY5Y cells, there were strong negative correlations between the reversibility of the neurite-disrupting effects and the FPN of each MTA. These results indicated that the FPNs of MTAs were predictable due to the reversibility of their neurite-disrupting effects in vitro.
In summary, it was indicated that neurite-disrupting effects are indexes of the incidence of PN, but not FPN. Meanwhile, as an in vitro model to evaluate the reversibility of these effects, we established that PC-12 and SH-SY5Y cells are useful for predicting FPN. Evaluations of neurite outgrowth have recently been automated with a high-content screening system (21). By combining the in vitro model established in this study with a high-content screening system, it is expected that not only the incidences of PN but also FPN of MTAs can be analyzed in a high-throughput manner.
- Received September 17, 2015.
- Revision received October 13, 2015.
- Accepted October 30, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.