The microtubule poison vinorelbine kills cells independently of mitotic arrest and targets cells lacking the APC tumour suppressor more effectively

Abstract

Colorectal cancers commonly carry truncation mutations in the adenomatous polyposis coli (APC) gene. The APC protein contributes to the stabilization of microtubules. Consistently, microtubules in cells lacking APC depolymerize more readily in response to microtubule-destabilizing drugs. This raises the possibility that such agents are suitable for treatment of APC-deficient cancers. However, APC-deficient cells have a compromised spindle assembly checkpoint, which renders them less sensitive to killing by microtubule poisons whose toxicity relies on the induction of prolonged mitotic arrest. Here, we describe the novel discovery that the clinically used microtubule-depolymerizing drug vinorelbine (Navelbine) kills APC-deficient cells in culture and in intestinal tissue more effectively than it kills wild-type cells. This is due to the ability of vinorelbine to kill cells in interphase independently of mitotic arrest. Consistent with a role for p53 in cell death in interphase, depletion of p53 renders cells less sensitive to vinorelbine, but only in the presence of wild-type APC. The pro-apoptotic protein BIM (also known as BCL2L11) is recruited to mitochondria in response to vinorelbine, where it can inhibit the anti-apoptotic protein BCL2, suggesting that BIM mediates vinorelbine-induced cell death. This recruitment of BIM is enhanced in cells lacking APC. Consistently, BIM depletion dampens the selective effect of vinorelbine on these cells. Our findings reveal that vinorelbine is a potential therapeutic agent for colorectal cancer, but they also illustrate the importance of the APC tumour suppressor status when predicting therapeutic efficacy.

Fig. 1.

APC deficiency increases vinorelbine-induced cell death. (A) Control (APC +) or APC-depleted (APC -) U2OS cells were exposed to the indicated concentrations of vinorelbine for 4 hours, fixed with PFA, stained for aCasp3 and analyzed using flow cytometry. Cells from ten independent samples were harvested for control cells and six independent samples for cells treated with vinorelbine. For each sample 10,000 cells were analyzed in each case. Results are means±s.e.m. UT, untreated. (B) APC depletion in control (UT) and vinorelbine-treated (Vino) U2OS cells. Immunoblots show successful depletion of APC with GAPDH as the loading control. Cells were treated with scrambled (APC +) or APC-targeting (APC -) siRNA and total lysates prepared from untreated control (UT) or vinorelbine-treated (100 μg/ml, 4 hours) cells. (C) Vinorelbine induces cell death during interphase, whereas nocodazole does not. U2OS cells were arrested in S phase using thymidine as described previously (Dikovskaya et al., 2007). Cells were then treated with either 50 μg/ml vinorelbine or 0.2 μg/ml nocodazole for the indicated times and the number of aCasp3-positive cells was measured by flow cytometry. Vinorelbine induced cell death in G1-arrested cells, whereas nocodazole did not. (D) U2OS cells do not enter G2 or mitosis 4 hours after release from thymidine. Control cells and thymidine-arrested cells were collected 0 hours (Start) and 4 hours after (End) thymidine release and fixed with ice-cold ethanol, and stained with PI for measuring of DNA content using flow cytometry. Cells did not enter G2 during this time window as indicated by the lack of a 4N DNA peak. The same profiles were obtained in cells treated with vinorelbine or nocodazole after thymidine arrest.

Fig. 2.

Cells enter apoptosis directly during interphase when treated with vinorelbine. U2OS cells lacking APC and/or p53, as indicated, and expressing fluorescent lamin B1 to mark the nuclear envelope, were treated with either 50 μg/ml vinorelbine or 1.25 μg/ml nocodazole and followed for 24 hours by fluorescence microscopy in the presence of Nucview. Time-lapse movies were recorded starting 2 hours after vinorelbine or nocodazole treatment. (A) Cell cycle distribution after 2 hours of drug treatment. The cell cycle stage of 200 randomly selected cells was scored 2 hours after adding vinorelbine or nocodazole. Mitotic cells were identified by visualiwing vesiculariwation of fluorescent lamin B1 as a marker for nuclear envelope breakdown. The highest standard error is shown as error bars. (B) The cell cycle stage immediately before cells entered apoptosis was recorded in the same cells as in A. Cell death was detected using NucView and mitotic cells were scored by visualiwing vesiculariwation of fluorescent lamin B1 (nuclear envelope breakdown) as a marker. Vinorelbine treatment induced cell death significantly more frequently during interphase than nocodazole treatment. (P=0.0001). Nocodazole preferentially induced cell death in mitotic cells (P=0.002). The probability of entering cell death was not influenced by the initial cell cycle stage in vinorelbine-treated cells (P=0.26).

Fig. 3.

APC depletion induces the accumulation of BIM in mitochondria, which is enhanced by vinorelbine. (A) APC and p53 were depleted by using RNAi. Cells were treated with 10 μg/ml vinorelbine for 30 minutes, harvested and mitochondria extracted. Equal amounts of protein were loaded and mitochondrial enrichment was confirmed using HSP60 as a mitochondrial marker. APC and p53 protein was measured in the same samples. (B) Mitochondrial fractions were collected and immunoblotted with the antibodies indicated in A. The relative intensity of bands was measured using an Odyssey LiCor imaging system. HSP60 acted as a mitochondrial marker. BIM levels are shown relative to those in asynchronous, control cells (APC- and p53-positive). At least four independent experiments were performed and the mean of the relative amount of BIM (±s.e.m.).

Fig. 4.

Increased sensitivity to vinorelbine in APC-deficient cells is not compromised by loss of p53 but is diminished in the absence of BIM. (A,B) Total cell lysates were immunoblotted to confirm RNAi-mediated depletion of APC, p53 and BIM in control or vinorelbine-treated cells (100 μg/ml, 4 hours) as indicated. Loading of equal amounts of protein was confirmed by measuring the amount of actin. (C) BIM-, APC- and/or p53-deficient U2OS cells, as indicated, were treated for 4 hours with 100 μg/ml vinorelbine, fixed with PFA, stained for aCasp3 and analyzed by using flow cytometry. Vinorelbine treatment was performed in three independent experiments in duplicates and for each sample 10,000 cells were analyzed. Results are means±s.e.m. Depleting APC increased the sensitivity to vinorelbine significantly and BIM depletion reduced this effect.

Fig. 5.

Vinorelbine kills cells during interphase in situ. Wild-type mice were injected with 10 mg per kg of body weight vinorelbine or taxol and tissue was harvested at the indicated times and processed for H&E staining (Radulescu et al., 2010). Mitotic (A) and apoptotic figures (B) were counted in 25 crypts in at least three mice. Their number per crypt is plotted for each time point. Vinorelbine treatment resulted in the appearance of apoptotic cells within 30 minutes, whereas taxol-treated tissue had an accumulation of mitotic cells before the appearance of apoptotic cells. In addition, significantly elevated apoptosis was still detectable in vinorelbine-treated tissue at 24 hours, but this had dropped almost to background in taxol-treated tissue. (C–E) Vinorelbine induces cell death preferentially in APC-deficient adenoma and mitotic arrest in normal tissue. Wild-type or APC^Min/+ mice were treated with 15 mg per kg of body weight vinorelbine for 6 hours, and tissue was harvested and processed for H&E staining. (C) Example images of a wild-type crypt on the left and an adenoma on the right, with insets showing selected apoptotic cells (enlarged 1.8-fold). Scale bars: 20 μm. Apoptotic (D) and mitotic (E) cells were counted in 5 tumours for each type from three mice (APC^Min/+) or 25 crypts each in three mice (wild type). (F) Vinorelbine reduces adenoma size in APC-deficient adenoma. APC^Min/+ mice (45–50 days old) were injected once every 2 weeks with 10 mg per kg of body weight vinorelbine or vehicle. Intestines were harvested at a 90-day timepoint and tumour burden was scored on wholemount tissue by measuring tumour number and size.