Professor Marta Izquierdo, Molecular Biology Department, Centro de Biologia Molecular Severo Ochoa (CSIC), Autonomous University of Madrid, Nicolás Cabrera, 1 UAM Cantoblanco, 28049 Madrid, Spain. E-mail: firstname.lastname@example.org
Gain-of-function mutations of kit tyrosine kinase receptor are associated with mastocytosis. Two subclones of the HMC1 mast leukaemia cell line were used; both express an identical KIT allele-specific regulatory type mutation (V560G), but differ in that one also expresses an enzymatic site type mutation (D816V) that confers on them resistance to imatinib mesylate tyrosine kinase inhibitor. In both cell lines, proliferation was suppressed and apoptosis induced by the combination of KIT gene silencing and α-tocopherol succinate (α-TOS), a derivate of α-tocopherol, also known as vitamin E. Furthermore, HMC1 cells with decreased kit levels by KIT silencing, failed to form tumours when xenotransplanted into immunocompromised mice and the animals were treated systemically with α-TOS. Targeting kit in the presence of α-TOS represents a new approach against proliferation of human mast leukaemia cell lines.
The KIT protooncogene encodes the kit protein, the tyrosine kinase receptor for stem cell factor (SCFR) (Longley et al, 2001). KIT is essential for normal development of mast cells in humans and other mammals. In cancer cells, activation of KIT is a consequence of autocrine/paracrine production of its ligand, SCF, or gain-of-function mutations either at the juxtamembrane or at the kinase domain of the protein (Longley et al, 2001). Autocrine production of KIT ligand, SCF, has been found in many tumours, such as neuroblastomas (Vitali et al, 2003), gynaecological tumours (Inoue, 1994), small cell lung cancer (Krystal et al, 1996), colon cancer and breast cancer (Hines et al, 1995). Gain-of-function mutations in KIT have been found in mastocytosis (Valent et al, 2005), gastrointestinal tumours (Hirota et al, 1998) and myeloproliferative disorders, among others (Ashman et al, 1999).
Mastocytosis is a disorder characterized by uncontrolled mast cell proliferation and migration to different organs that can lead to a systemic disease with, in the last instance, increased mast cell burdens in both bone marrow and blood (Quintas-Cardama et al, 2006). Two types of KIT-activating mutations have been described in mastocytosis (Longley et al, 2001; Orfao et al, 2007). The first type, exemplified by codon 816 mutations, causes residue substitutions in the activation loop at the entrance to the enzymatic pocket of the protein. The second type, exemplified by mutations in the intracellular juxtamembrane domain of the protein, affects the regulation of the activity of the enzyme by substitution of glycine for valine at position 560 (Val560Gly). Although this latter mutation is less frequent in human mastocytosis, it is found in the majority of canine mastocytomas (Ma et al, 1999) and in human gastrointestinal stromal tumours (GISTs) (Hirota et al, 1998).
Tyrosine kinases play an important role in the modulation of growth factor signalling. Cancer therapies based on tyrosine kinase inhibitors, such as imatinib mesylate (Capdeville et al, 2002), dasatinib (Olivieri & Manzione, 2007) or erlotinib (Perez-Soler, 2007), have enabled the target-specific treatment of selected malignancies in recent years. The clinically approved imatinib mesylate (also known as Gleevec or STI-571), is a specific tyrosine kinase inhibitor of kit, PDGFR and Bcl-Abl, and has proven quite effective in the treatment of GISTs, certain types of mastocytosis with KIT mutations at the regulatory domain of the protein (Val560Gly) (Hartmann et al, 2005), and other tumours which express kit but lack activating mutations (Johnson et al, 2006). However, the majority of mast cell neoplasms and other myeloid disorders are characterized by mutations at the catalytic site of KIT (Asp816Val) (Garcia-Montero et al, 2006). These are resistant to the drug because the substitution mutation promotes the formation of a loop next to the juxtamembrane domain where STI-571 binds (Zermati et al, 2003; Mol et al, 2004) interfering with the normal action of the compound. Furthermore, many tumours treated with STI-571 acquire drug resistance in a time dependent manner due, in most cases, to the development of new point mutations in KIT (Haller et al, 2007).
RNA interference (RNAi) has become a powerful tool to inactivate specific gene targets by degradation of the corresponding mRNA molecules (Kim & Rossi, 2007). RNAi-based approaches represent reasonable alternatives for cancer therapy when treatment options are limited. The ability of RNAi to silence genes in culture cells has been adapted to different expression vectors, such as retroviruses and lentiviruses, to facilitate the delivery of these constructs into mammalian cells (Brummelkamp et al, 2002).
The strategy we present here is based on KIT gene silencing by RNAi in combination with α-tocopherol succinate (α-TOS), a derivate of vitamin E (also known as α-tocopherol) (Neuzil et al, 2007). The combined treatment induced apoptosis in human mast leukaemia cells and abolished xenograph formation in nude mice.
Human mast cell leukaemia cell line HMC1·1, which has the mutation V560G in the juxtamembrane domain of KIT, and HMC1·2 cell line, which is a subclone of the HMC1·1 cell line having an additional D816V kinase domain mutation at the same allele as the original V560G mutation (Butterfield et al, 1988), were kindly provided by Dr L. Escribano (Hospital Virgen del Valle, Toledo, Spain) and Dr J.H Butterfield (Mayo Clinic, Rochester, USA) respectively. Cells were grown in Iscove′s Modified Dulbecco′s medium (IMDM) supplemented with 10% fetal bovine serum and penicillium 100 μg/ml. Phoenix amphotropic cell line and human primary epithelial fibroblasts were cultured in DMEM supplemented with 10% fetal bovine serum.
Drugs and reagents
STI-571 (Imatinib mesylate) was kindly provided by Novartis (Barcelona, Spain). It was dissolved in phosphate-buffered saline (PBS) at 1 mmol/l. Stocks were frozen at −20°C. α-TOS, cisplatin and doxorubicin were purchased from Sigma Aldrich (Manchester, UK). The general pan caspase inhibitor z-VAD-fmk and the caspase 8 inhibitor, z-IETD-fmk, were purchased from TOCRIS Biosciences (Bristol, UK). They were dissolved in dimethyl sulphoxide (DMSO) at different concentrations and kept at 4°C.
The primary antibodies used for Western blot analysis were: rabbit polyclonal anti-kit (Santa Cruz, Inc., Santa Cruz, CA, USA; sc-68), rabbit polyclonal anti-PARP1 (Sigma Aldrich); rabbit polyclonal anti-caspase 8 (Santa Cruz, sc-7890); rabbit polyclonal anti-DR5 (Cell Signalling Technology, Beverly, MA, USA); mouse monoclonal anti-Bax (BD Pharmingen, Hamburg, Germany); mouse monoclonal anti-p53 (BD Pharmingen); rabbit polyclonal anti-p21 (Santa Cruz, sc-397-G) and rabbit polyclonal anti-actin (Sigma Aldrich). Anti-rabbit immunoglobulin G, anti-mouse immunoglobulin G and anti-goat immunoglobulin G horseradish peroxidase-conjugated secondary antibodies were purchased from DAKO pharmaceuticals (Copenhagen, Denmark).
Retroviral-based KIT shRNAs
We followed previously described methods (Brummelkamp et al, 2002), to generate interferent RNAi against KIT wild type and mutant (Val560Gly, Asp816Val) mRNAs present in HMC1·1 and HMC1·2 cells. The synthetic DNA sequences corresponding to the shRNAs for KIT were selected from GenBank accession number NM_00222. 21.
Retroviral infection of human leukaemia mast cells
Plasmid-retroviral vectors carrying the DNA to be transcribed to Tot RNAi, Mut RNAi and control RNAi were transfected into Phoenix amphotrophic packaging cell line using lipofectamine and Plus Reagent following the protocols provided by the manufacturers. After a 48-h transfection, viral supernatants were collected and used to infect HMC1·1 and HMC1·2 cell lines in the presence of polybrene (8 μg/ml). Cells were infected overnight and allowed to recover for 24 h with fresh medium. Twenty-four hours after retroviral infection, puromycin was added to cell cultures.
Western blot analysis
Cells were harvested in TNSEV lysis buffer (50 mmol/l Tris–HCl, pH 7·5, 2 mmol/l EDTA, 100 mmol/l NaCl, 25 mmol/l NaF, 1 mmol/l sodium orthovanadate, 1% Nonidet P-40) or AB lysis buffer [50 mmol/l Tris pH 7·6; 400 mmol/l NaCl; 1 mmol/l EDTA; 1 mmol/l EGTA; 1% sodium doecyl sulphate (SDS)] supplemented with freshly added 20 μg/ml aprotinin, 20 μg/ml leupeptin and 1 mmol/l polyvinylidene difluoride. After incubation on ice for 30 min, cell debris was removed by centrifugation at 25 200 gfor 15 min. Supernatants were transferred to a clean tube and protein concentration was measured using the Bio-Rad protein assay dye reagent, following the manufacturer’s instructions (Bio-Rad, Hercules, CA, USA). Samples (60 μg of protein) were resolved by 8%, 10% or 12% SDS polyacrylamide gel electrophoresis (SDS–PAGE) and subjected to standard immunoblotting procedures. (Uriarte et al, 2005). All Western blot analysis were performed in triplicate.
Cell survival assay
Cells were plated in triplicate into 24-well plates and then treated with STI571 (0–10 μmol/l), cisplatin, doxorubicin, RNAi against KIT and/or α-TOS. At each time point, cell survival rates were analysed following addition of 3, 4-(4,5-dimethylthiazol-2-yl)-2·5-diphenyl-tetrazolium bromide (MTT) reagent to the cultured cells as previously described (Jaattela et al, 1998).
To determine the proportion of apoptotic HMC1·1 and HMC1·2 cells, after incubation in the presence of RNAi against KIT and/or α-TOS, cells were pelleted, washed in PBS, and resuspended in apoptosis binding buffer (10 mmol/l HEPES pH7·4, 140 mmol/l NaCl and 2·5 mmol/l CaCl2). Then they were incubated with annexin-V phycoerythrin (PE) and 7-Amino-actinomycin D (7-AAD) for 15 min in the dark at room temperature. 7-AAD was added to enable the identification of cells that had lost their membrane integrity. Cells stained positive for both annexin-V and 7-AAD would suggest a late apoptotic or necrotic type of cell death, while only annexin-V positive stain cells would imply an early apoptotic stage. After incubation, cells were washed and then resuspended in binding buffer at a concentration of 1 × 106/ml. Samples were measured by flow cytometry. Data were analysed with CellQuest software (BD Biosciences, Hamburg, Germany). All experiments were performed in triplicate.
Cell cycle analysis
Cells treated in triplicate with either shRNA against KIT and/or α-TOS were centrifuged, resuspended in cold PBS and fixed with cold ethanol. After fixation for 24 h, cells were washed and resuspended in PBS 1% bovine serum albumin and incubated with propidium iodide (50 μg/ml) and RNase at 250 μg/ml for 1 h. After that, the subG0/G1 peak of the cell cycle, corresponding to DNA fragmentation (only manifest in apoptotic cells), was measured by flow cytometry. Data were analysed as described above.
Measurement of the mitochondrial transmembrane potential (Δψm)
Cells were seeded in triplicate at a concentration of 5 × 105 cells/ml, and treated with or without α-TOS at different times. Cells were incubated at each time point with 10 nmol/l 3,3′-dihexyloxacarbocyanine DiOC6(3) (Molecular Probes Inc., Eugene, OR, USA) for 20 min. Loss of mitochondrial membrane potential was determined by flow cytometry as previously described. (Rosato et al, 2003).
Eight week-old female athymic nude mice were inoculated with 1·5 × 106 cells, carrying retroviral-based shRNA against KIT, at the right flank. Nine days later, mice were injected intraperitoneally with 50 μl of α-TOS (30 mg/ml) or the vehicle DMSO every two days for 10 weeks. Tumour volumes were estimated by measuring the maximal height, length and width at the times indicated [π/6 (length × width × height)]. Survival of female athymic nude mice was recorded until the tumour reached 10% of the mouse’s weight,when the mice were euthanized.
Values are expressed as the mean ± SD (standard deviation) comparisons were made with the t student; the threshold of significance was fixed at *P < 0·05 and **P < 0·01.
Long-term effects of RNAi against KIT in human mastocytosis
The proto-oncogene KIT is essential for the proliferation and survival of mast cells. We have previously described the construction of two different retrovirus transcribing RNAi against KIT mRNA (Ruano & Izquierdo, 2009). One of the shRNAs was directed specifically to the Val560Gly mutation (Mut), while the other targeted a wild type region of KIT and inhibited the mRNAs of both alleles of the gene (Tot). Retroviral shRNA efficiently decreased kit expression in both HMC1·1 and HMC1·2 cells (Fig 1A, B respectively). The inhibition of kit using the specific shRNA for the mutation 560 (Mut) was about half that observed using the shRNA against the wild type sequence (Tot). In both cell types, depletion of kit significantly increased the apoptotic cell population from day 2 after retroviral infection, to acquire values of 58% (Mut RNAi) and 40% (Tot RNAi) obtained from absolute cell counts at day 5 for HMC1·1. Similarly, in HMC1·2, apoptotic values of 42% and 35% were observed. Apoptosis progression levelled off at day 7 and thereafter (Fig 1C, D). These results confirmed the antiproliferative effect of RNAi against KIT in human mast leukaemia cell lines (Ruano & Izquierdo, 2009) but also showed the limitations of the approach as a large percentage of cells remained alive.
Anti-proliferative effects of α-tocopherol succinate on human mastocytosis
Previous studies showed that α-tocopherol (α-TOH), caused a 50% decrease in HMC1 cell viability 24 h after treatment, followed by a recovery period of 48 h (Kempna et al, 2004). The present study used a derivate of α-TOH, called α-tocopherol succinate (α-TOS), known for its antitumoural properties (Neuzil et al, 2002). This drug promotes a variety of cellular events depending on the concentration used. HMC1·1 and HMC1·2 cells were more sensitive to α-TOS treatment than other tumour cell lines, such as pancreatic adenocarcinoma (Panc-1 cells), cervical cancer (HeLa cells), and colon adenocarcinoma (HT29 cells) (Fig 2A). Next we examined the viability of cells at different α-TOS concentrations. After treating mast cells with 15 μmol/l α-TOS for 5 days, the viability of HMC1·1 cells decreased to 45–50%, without affecting the viability of non-transformed primary human fibroblast cell cultures (Fig 2B). Higher concentrations of α-TOS, such as 25 μmol/l, affected the growth of non-transformed primary fibroblasts to some extent, promoting their cell cycle arrest during G2/M transition (data not shown). We thus established 15 μmol/l as the suitable concentration of α-TOS for the treatment of HMC1·1 cells.
KIT knockdown and α-TOS combined therapy promotes apoptosis in HMC1·1 cells harbouring the V560G mutation
We studied the combined effect of kit knockdown and α-TOS (15 μmol/l) therapy on HMC1·1 cells. Two days after puromycin selection, cells infected with the retroviruses carrying the shRNAs against KIT were treated with 15 μmol/l α-TOS for 5 days (finishing 7 days after the retroviral infection). We then analysed cell death by annexin-V PE and 7-AAD staining followed by flow cytometry analysis. Cell death was highest in the HMC1·1 subclone treated with shRNA against mutated KIT and 15 μmol/l α-TOS, reaching 66·42% (Fig 3). Cells treated with the shRNA against mutated KIT showed apoptosis values of 28·56%, while only 24·24% of cells were apoptotic when treated with α-TOS. Control cells showed apoptosis of 16·36%. The combined therapy was also quite lethal when using the shRNA against the wild type sequence of KIT and α-TOS, reaching almost 50%. These results showed that the combination of RNAi against KIT and α-TOS substantially increase human HMC1·1 mast leukaemia cell death.
The role of caspases in the cell death induced by the combined therapy was next investigated by measuring the cell-cycle distribution of HMC1·1 treated cells in the absence and presence of the general pan caspase inhibitor, z-VAD-fmk (Figs 4A and S1). The accumulation of HMC1·1 cells in the sub-G0/G1 cell-cycle phase, indicating apoptotic cells after the combined treatment, was rescued to control levels (cells without α-TOS treatment) by the addition of the pan caspase inhibitor z-VAD-fmk at 25 μmol/l. We also measured the direct target of caspases 3 and 7, PARP-1 (Simbulan-Rosenthal et al, 1999). The combination of KIT gene silencing and α-TOS inactivated PARP-1 by its cleavage into a 86 KDa protein (Fig 4B).
Tomasetti et al (2006) described that α-TOS sensitized cells to death receptor-mediated apoptosis. We then studied the putative role of death receptor 5 (DR5) and caspase 8, a downstream component of surface death receptors, in the cell death mechanism. Western blot analysis showed that DR5 was upregulated after cell treatment with α-TOS while the inactive form of caspase 8 was downregulated with the combined therapy in HMC1·1 cells (Fig 4B). Furthermore, the presence of the caspase 8 inhibitor, z-IETD-fmk, at 25 μmol/l, blocked the apoptosis generated by the combined system, demonstrating that caspase 8 also plays an important role in the mechanism of cell death (Fig 4C). We further explored the levels of the proapoptotic protein Bax, whose overexpression normally sensitizes cells to death (Xu et al, 2002). Bax protein levels were almost duplicated as a consequence of the combined treatment (Fig 4D).
Apoptosis can be executed in a p53 dependent or independent manner (Clarke et al, 1993). CDKN1A gene, which is regulated by p53 and encodes a cyclin-dependent kinase inhibitor, may be critical for p53-induced cell cycle arrest. p53 and CDKN1A protein levels were measured in cells treated with RNAi against KIT in the absence and in the presence of α-TOS. The combined treatment increased p53 levels as well as its target, CDKN1A (Fig 4D), suggesting that the accumulation of HMC1·1 cells in the sub-G0/G1 cell-cycle phase described earlier, might be caused by p53 upregulation.
Modifications in mitochondrial transmembrane potential (Δψm) have been correlated with the induction of cell death. Thus, we measured mitochondrial membrane potential loss (Δψm) by flow cytometry DiOC6 uptake (Fig 4E, F). After a short period (72 h) of treatment, there was no synergism between kit knockdown and α-TOS treatment, and the 40% reduction on Δψm observed in the combined therapy was mainly due to kit inhibition. Nevertheless, after longer periods of treatment, i.e., 5 days or more, there was a strong synergism in the shRNA and α-TOS combined treatment, leading to a 45–60% reduction of their Δψm.
The results suggest that depleting kit in the presence of α-TOS triggers a death process initiated by elevation of p53 protein probably in response to α-TOS. This is followed by increments in DR5 and Bax, which would in turn promote apoptotic death mediated by caspase 8 accompanied by severe depolarization of the mitochondrial membrane and PARP cleavage in the nucleus.
KIT knockdown and α-TOS combined treatment inhibits the tumourigenic potential of HMC1·1 cells harbouring V560G mutation
To study the tumourigenic potential of cells depleted of kit in the presence of α-TOS, HMC1·1 cancer cells were infected in vitro with the retroviral-based shRNA against total KIT, mutant KIT or the control vector with no foreign DNA insert. Following puromycin selection for 2 days after retroviral infection, 1·5 × 106 cells were inoculated at the right flanks of nude mice (n = 8). Nine days later, the animals were injected intraperitoneally three times per week with 50 μl α-TOS (30 mg/ml) or the vehicle, DMSO, as previously described (Yu et al, 1999; Weber et al, 2002). Mice injected with cells depleted of kit expression and treated with α-TOS drastically reduced tumour cell development even after 10 weeks (Fig 5A, B). In contrast, animals bearing control cells and untreated or treated only with α-TOS, displayed large tumours (Fig 5C). The same result was obtained in animals treated only with Tot shRNA (Fig 5D) or Mut shRNA (Fig 5E) against KIT.
KIT knockdown and α-TOS combined therapy promotes apoptosis in cells harbouring both the V560G/ D816V KIT mutations (HMC1·2 cells)
We then evaluated whether the combined therapy produced the same apoptotic effect in HMC1·2 cells as the one observed previously in HMC1·1 cells. Two days after puromycin selection of cells infected with the retroviruses carrying the shRNAs against KIT (Tot and Mut), we exposed them to 15 μmol/l of α-TOS for 5 days. As shown in Fig 6A, flow cytometry analysis of death showed a high percentage of apoptosis in HMC1·2 cells treated with the Tot RNAi or the Mut RNAi against KIT in combination with α-TOS (59% and 90% respectively) in comparison with cells treated only with the corresponding RNAi (13% for cells treated with Tot RNAi and 17% for cells treated with Mut RNAi). Cells treated exclusively with α-TOS showed a small percentage of apoptosis (20%) while this was 13% in control cells.
Finally we compared the efficacy of the combined therapy with the tyrosine kinase inhibitor, imatinib mesylate. Our data supported previous reports of HMC1·2 cell resistance to that inhibitor (Fig 6B) (Akin et al, 2003). However, the combined therapy was successful in reducing the survival of both HMC1·1 and HMC1·2 cells (Fig 6C) indicating its potential for the treatment of imatinib mesylate-resistant HMC1·2 cells.
Currently there are no curative therapy options for patients with aggressive forms of systemic mastocytosis (Escribano et al, 2002; Orfao et al, 2007). Imatinib mesylate, a tyrosine kinase inhibitor of Bcr-Abl, PDGFR and kit, was approved for the treatment of acute myeloid leukaemia and KIT-related malignancies. Nevertheless, the drug failed to inhibit tumour growth of mastocytosis harbouring the D816V gain-of-function mutation (Akin et al, 2003).
RNA interference has been used with promising results in the treatment of diseases caused by the expression of dominant, gain-of-function-mutations type of genes (Martinez et al, 2002; Duursma & Agami, 2003; Raoul et al, 2005). We have previously observed that the inhibition of KIT mRNAs by interferent RNA (shRNA), in both HMC1·1 and HMC1·2 cells, leads to a decrease in their survival by promoting mast cell leukaemia apoptosis 72 h after treatment, followed by a recovery period of 5 days (Ruano & Izquierdo, 2009). We have here confirmed the antiproliferative effect of that approach. As the combined RNAi + α-TOS treatment achieves higher percentages of cell death, is reasonable to think that maybe some, but certainly not all, of the cells resistant to RNAi would be also resistant to the combined approach. α-TOS is an antitumoural compound that enhances the antiproliferative effects of different chemotherapeutic agents, such as cisplatin and tamoxifen, in melanoma cells, and doxorubicin in prostate cancer and leukaemia cells (Prasad et al, 1994). The inhibition of either mutant or wild type mRNAs of KIT by shRNA in combination with α-TOS could be considered a new strategy against mastocytosis because it extends the transient therapeutic effect of kit knockdown on the survival of human mast cells. Moreover, α-TOS treatment promotes the highest apoptosis when combined with shRNA against the mutant mRNAs of KIT.
The combination of kit RNAi with low concentrations of α-TOS (15 μmol/l) sensitizes cells to a caspase-dependent apoptosis. The observed upregulation of Bax may be a consequence of the increment of p53 after the treatment with α-TOS. Indeed, p53 may also upregulate DR5 protein levels as previously described in malignant mesothelioma (Tomasetti et al, 2004), jurkat cells (Dalen & Neuzil, 2003) and colon cancer cells (Lim et al, 2007). Caspase 8 activation, on the other hand, may facilitate Bax translocation to the mitochondria, probably via cleavage of Bid into tBid, which then would trigger Bax-driven mitochondrial membrane potential loss. Simultaneously in the nucleus, caspase 3 would mediate cleavage of PARP-1 (Roucou et al, 2002). The present study demonstrated that combining kit depletion and α-TOS systemic treatment was sufficient to prevent tumour growth of mast cell leukaemia xenografts in mice.
The strategy described here may represent an advance in the treatment of mast cell neoplasms responding poorly to chemotherapeutic agents and resistant to imatinib mesylate. The combined treatment based on kit RNAi and α-TOS may also be an alternative therapy for certain KIT-related malignancies, such as GISTs, that respond primarily to imatinib mesylate but develop resistance in a time-dependent manner due to the appearance of new KIT mutations (McLean et al, 2005; Haller et al, 2007). Our results illustrate the critical importance of combining agents in order to improve cancer therapeutics.
This study was supported by a grant from the Fundación Mutua Madrileña Automovilística (FMMA) Madrid-Spain, and also from the Fondo de Investigaciónes Sanitarias, Ministerio de Ciencia e Innovación (before Sanidad y Consumo, PI06/0554). The Centro de Biología Molecular Severo Ochoa (CBMSO) is also a recipient of an institutional grant from the Ramon Areces Foundation, and I. Ruano was supported by a grant from the Comunidad Autónoma de Madrid, Spain. Dr David Butterfield (Mayo Clinic, Rochester, USA), Dr Luis Escribano (Hospital Virgen del Valle, Toledo, Spain) are acknowledged for their advice about human mastocytosis.