Skip to main content
NCBI home page
Molecular Oncology logoLink to Molecular Oncology
. 2012 Oct 30;7(3):323–333. doi: 10.1016/j.molonc.2012.10.008

Effects of endoplasmic reticulum stressors on maturation and signaling of hemizygous and heterozygous wild‐type and mutant forms of KIT

Sabrina Brahimi-Adouane 1,2,†, Jean-Baptiste Bachet 1,3,4,†, Séverine Tabone-Eglinger 5, Frédéric Subra 6, Claude Capron 7,8,9, Jean-Yves Blay 10,11, Jean-François Emile 1,2,✉
PMCID: PMC5528484  PMID: 23146721

Abstract

Gain of function mutations of KIT are frequent in some human tumors, and are sensible to tyrosine kinase inhibitors. In most tumors, oncogenic mutations are heterozygous, however most in vitro data of KIT activation have been obtained with hemizygous mutation. This study aimed to investigate the maturation and activation of wild‐type (WT) and mutant (M) forms of KIT in hemizygous and heterozygous conditions. WT and two types of exon 11 deletions M forms of human KIT were expressed in NIH3T3 cell lines. Membrane expression of KIT was quantified by flow cytometry. Quantification of glycosylated forms of KIT and phosphorylated forms of AKT and ERK were performed by western blot. Simultaneous activation of WT KIT and treatment with endoplasmic reticulum (ER) inhibitors, tunicamycin or brefeldin A induced a complete inhibition of membrane expression of the 145 kDa form of KIT. By contrast activation or ER inhibitors alone, only partly inhibited this form. ER inhibitors also inhibited KIT activation‐dependent phosphorylation of AKT and ERK1/2. Brefeldin A induced a complete down regulation of the 145 kDa form in hemizygous M, and induced an intra‐cellular accumulation of the 125 kDa form in WT but not in hemizygous M. Heterozygous cells had glycosylation and response to ER inhibitors patterns more similar to WT than to hemizygous M. Phosphorylated AKT was reduced in hemizygous cells in comparison to WT KIT cells and heterozygous cells, and in the presence of brefeldin A in all cell lines. Effects of ER inhibitors are significantly different in hemizygous and heterozygous mutants. Differences in intra‐cellular trafficking of KIT forms result in differences in downstream signaling pathways, and activation of PI3K/AKT pathway appears to be tied to the presence of the mature 145 kDa form of KIT at the membrane surface.

Keywords: KIT receptor, Cell line, Imatinib, Glycosylation, Protein trafficking, AKT

Highlights

  • ►

    KIT mutations are heterozygous in most tumors.

  • ►

    In vitro experiments were almost always performed in hemizygous cell models.

  • ►

    Glycosylation patterns of heterozygous cells are different than of hemizygous cells.

  • ►

    Wild‐type KIT is associated with modifications in intra‐cellular trafficking.

  • ►

    Activation of downstream signaling pathways is influenced by the wild‐type allele.

Abbreviations

ER

endoplasmic reticulum

IM

imatinib mesylate

RTK

receptor tyrosine kinase

WT

wild-type

1. Introduction

Several human tumors contain somatic gain of function mutations of KIT, including gastrointestinal stromal tumors (GIST) (Hirota et al., 1998), mast cell tumors (Staser et al., 2010), melanomas (Willmore‐Payne et al., 2005), and thymomas (Buti et al., 2011). Germ line gain of function mutations of KIT is responsible for an increase incidence of GIST in humans (Bachet and Emile, 2010) and in transgenic mice (Nakai et al., 2008). The major oncogenic role of KIT mutations has been further demonstrated by the clinical benefit of treatment with tyrosine kinase receptor inhibitors, notably in human GISTs (Blanke et al., 2008). More than 150 different mutations of KIT have been reported in human. The most frequents are localized within exon 9 coding for extra‐cellular domain, exon 11 coding for intra‐cellular juxta‐membrane autoinhibitory loop or exon 13 or 17 coding for the enzymatic domain (Emile et al., 2004; Rubin et al., 2007; Staser et al., 2010). The frequency and the type of KIT mutations are highly dependent on the tumor type, exon 17 and 11 mutations being the most frequent in human adult mastocytosis and GIST, respectively.

GIST are the most frequent sarcomas in human (Cassier et al., 2010), and have KIT mutations in 71% of cases with various prognosis (Emile et al., 2012). The deletions within exon 11 of KIT (Emile et al., 2004; Rubin et al., 2007; Bachet and Emile, 2010) which are the most frequent mutations in GIST, have also the worse prognosis, the delWK557‐558 mutation in particular (Martin et al., 2005; DeMatteo et al., 2008; Bachet et al., 2009). Another factor of bad prognosis is GIST with homozygous mutations (Lasota et al., 2007; Emile et al., 2008), while most are heterozygous, with expression of both the wild‐type (WT) and the mutant allele (Théou et al., 2004).

Because of the high frequency of KIT gain of function mutations in different tumor types, and the benefit of treatment with TKR inhibitors, a lot of cellular models of KIT WT and mutant has been developed. Cells of various lineage were used, such as hematopoietic (EML, TF1), lymphoid (Baf3, FDC‐P1), mast cell (HMC‐1), fibroblasts (NIH3T3, rat2, COS‐1) or stem cells (CHO) (Kitayama et al., 1995; Tsujimura et al., 1999; Jahn et al., 2007; Xiang et al., 2007; Tabone‐Eglinger et al., 2008; Bougherara et al., 2009; Sun et al., 2009; Yang et al., 2010; Kim et al., 2011), and expressed various types of KIT mutations. To our knowledge except for one study (Kim et al., 2011), all cell models used were hemizygous and may thus not be representative of KIT biology in tumors (Kitayama et al., 1995; Tsujimura et al., 1999; Jahn et al., 2007; Xiang et al., 2007; Tabone‐Eglinger et al., 2008; Bougherara et al., 2009; Sun et al., 2009; Yang et al., 2010), although most tumors are heterozygous for KIT mutations. Therefore, we developed a cellular model with hemizygous or heterozygous KIT mutations, and analyzed the consequence on the trafficking and the activation of KIT.

2. Material and methods

2.1. Cell lines

The production of recombinant viruses and stably infection of murine fibroblast cell lines (NIH3T3) has been described previously (Tabone‐Eglinger et al., 2008). In addition to the previously described hemizygous forms (KIT WT, KIT delWK557‐558/D6, KIT delNGNNYVYIDPTQPYDHK564‐581/D54 and MigrR1 (MIGR) as control), two heterozygous cell lines were developed. They contain both, the WT and the mutant alleles, namely WT/D6 and WT/D54.

For this, 293T cells were co‐infected with plasmids containing both the WT and the mutant allele (D6 or D54) during the same experiment. The media containing virus was then harvested and used to infect NIH3T3 cells (multiplicity of infection of 0.5). WT were infected with pLNCX oncological vector (Clontech®) and selected with geneticin for 7–10 days before being isolated and amplified. KIT mutants were inserted in oncological vector MIGR1, amplified for 48 h after infection and green fluorescent protein‐positive cells were sorted. Heterozygous cells were submitted to both, geneticin and green fluorescent protein‐positive selection. Both retroviral vectors included the same cytomegalovirus promoter. The last cell line expressed an empty vector MigrR1 (MIGR) used as control.

2.2. Cells culture and proliferation study

NIH3T3‐infected cells were cultivated at 37 °C in DMEM supplemented with 10% Newborn Calf Serum (Life Technologies®), 2% penicillin/Streptomycin and 50 ng/mL of Chinese Hamster Ovary‐mouse SCF (initial gift from P. Dubreuil) so that the growths of wild‐type and mutant forms were equivalent. This mSCF‐CHO was also used for the experiments except for the proliferation assay. Cells containing WT form were grown with 0.5 mg/mL of geneticin (Calbiochem®) but removed for experiments.

To study proliferation in low serum concentrations, cells were plated at 50,000 cells per well, then serum starved for 4 h before incubation with DMEM containing 1% serum for the times indicated. At each time point, cells were harvested, suspended in a precise volume of medium, and then counted with a FACS scan (Becton Dickinson®). Results were expressed as mean of triplicate ± SE of one single experiment.

2.3. Inhibitors

Imatinib mesylate (Glivec, STI571) was provided from Novartis. Tunicamycin (from Streptomyces sp.) and brefeldin A (from Penicillium brefeldianum) were purchased from Sigma–Aldrich® (France). N‐Glycosidase F (PNGase) and endoglycosidase H (EndoH) were purchase from Roche Diagnostic® (Mannheim, Germany).

2.4. Antibodies

The following antibodies were used for western‐blotting: polyclonal rabbit anti‐KIT antibodies from Dako® (Trappes, France), rabbit anti‐KIT phosphotyrosine 703 from Clinisciences® (Montrouge, France), anti‐AKT antibody, p44/42 MAP Kinase antibody, anti‐phospho‐Akt Ser473 antibody from Cell Signaling Technology® (Danvers, USA), anti‐p‐ERK antibody from Santa‐Cruz Biotechnology® (CA, USA), Hrp‐conjugated goat anti‐rabbit IgG antibody, Hrp‐conjugated goat anti‐mouse IgG antibody and Hrp‐conjugated rabbit anti‐goat antibody from Anaspec® (CA, USA), and anti‐β‐actin from Sigma–Aldrich®.

2.5. Proteins extraction and western‐blotting

Cells were seeded in multiple six‐well plates, grown to 80% confluence in complete medium and then starved from SCF overnight. Reagents were then added at 37 °C for times indicated: SCF 50 ng/mL, imatinib mesylate (IM) 1 μM, brefeldin A 3 μg/μL, tunicamycin 5 μg/μL.

The cells were washed and lysed directly in culture plate with 200 μL/10 cm2 of lysis buffer containing 10 mM Tris p7.4, 150 mM NaCl, 1 mM EDTA, 0.5% deoxycholic acid, 0.1% sodium dodecyl sulfate, 1% Triton X‐100, 1 mM Na3VO4, 10 mM sodium fluoride, and 10% Protease inhibitor cocktail (Sigma®). Lysates were incubated for 30 min on ice and centrifuged at 10,000×g for 15 min at 4 °C.

After determination of protein concentrations with micro BCA protein assay kit® (Pierce Biotechnology, Thermo Scientific), 12–30 μg protein extracts were loaded on a 10% acrylamide SDS‐PAGE gel, then immunoblotted onto nitrocellulose membranes. Membranes were blocked in PBS‐T (0.05% Tween‐20) containing 5% non fat dry milk for non‐phosphorylated antibodies or 5% BSA for phosphorylated antibodies, incubated overnight at 4 °C with the primary antibodies, then revealed with a horseradish peroxidase‐labeled secondary antibody and visualization with ECL detection Kit (GE Healthcare, Amersham Biosciences®).

The bands of interest were detected using a luminescent image analyzer, LAS‐3000 (Fujifilm®, Tokyo, Japan). Results were quantified using MultiGauge software® (version 2.02) of the LAS‐3000. The background noise was subtracted from each samples and the control condition was the standard, set to 100%. We calculated the ratio to actin used as loading control. Three experiments or more were performed. All results are given as means ± SD.

2.6. Glycosidase treatment

After protein extraction and denaturation, extract protein of 20 μg of proteins was resuspended in 50 mM sodium citrate (pH 5.5), 0.1% SDS and 1% β‐mercapto‐ethanol. The samples were divided proportionally for untreated (control), EndoH at 10 mM, and PNGase F at 2.5 U. The proteins were treated for 3 h at 37 °C.

2.7. RNA extraction and RT‐PCR

RNA was extracted with RNeasy MiniKit® (Qiagen, France) according to the manufacture's recommendation. With a RT‐PCR of RNA extract, cDNA was obtained and a PCR was performed with exon 9–12 primer KIT (exon 9*‐12Bis) for amplification of KIT gene and have a qualitative and quantitative result. The results were analyzed on a sequencer (Applied Biosystem®) and were read on GeneMapper software (Applied Biosystem®).

2.8. Flow cytometry

Transfected cells were seeded in multiple six‐well plates, grown to 80% confluence in complete medium and then starved from SCF overnight. In order to amplify and accelerate the reagents effects, we decreased number of cell surface expression of KIT by incubating all cells with SCF during 30 min. For conditions with ER stressors, brefeldin A or tunicamycin was added to SCF in wells during pre‐treatment. Then, cells were washed with fresh medium and incubated with reagents for 3 h at 37 °C: SCF 50 ng/mL, IM 1 μM, brefeldin A 3 μg/μL and/or tunicamycin 5 μg/μL.

Cells were harvested, washed in cold buffer and stained with 3 μL of mouse anti‐human KIT (clone 104D2) phycoerythrin‐conjugated antibody or control IgG1a antibody (both from Dako®, Denmark) for 30 min at 4 °C, to finally be fixed by 5% paraformaldehyde. For analysis of total cell expression of KIT, cells were fixed in 2% PBS‐paraformaldehyde for 15 min at 4 °C before permeabilization with 0.3% saponin, and then incubated with the PE‐conjugated antibody.

Cells were analyzed by flow cytometry on a Becton Dickinson® FACS Quanto (2.0 FACS Diva version 6.1.3 software). For mutant forms, hemizygous and heterozygous, cells were selected on the presence of green fluorescent protein. The KIT expressing cell population was gated thanks to the negative control. Then, modification of KIT expression at the cell surface was expressed as a percentage of positive cells to the KIT in control condition (set to 100%).

2.9. Statistical analysis

The experiments were performed until six times for KIT ratio quantification, three times for western blot and five times for FACS. Statistical differences were identified by application of student test. Probability values <0.05 were considered as significantly different. Results were expressed as mean ± SD.

3. Results

3.1. KIT transcription and cell line proliferation

PCR was used to assay KIT mRNAs in the various cell lines. As expected, the only forms of KIT expressed in hemizygous cells were WT, D6 and D54 KIT. In heterozygous cells, the mRNAs for the mutated and WT alleles were similarly abundant in the WT/D6 cell line, and the D54 mRNA was twice as abundant as the WT mRNA in the WT/D54 cell line (Supplementary Figure 1a). Quantification by real time PCR indicated that the levels of KIT mRNA (mutated and/or WT KIT) were similar in all cell lines, both homozygous and heterozygous (Supplementary Figure 1b).

The hemizygous KIT mutants (D6 and D54) have a high proliferative capacity (Tabone‐Eglinger et al., 2008).We tested the proliferative capacity of the double‐transfected cell lines in low serum concentration medium (Figure 1). Like the D6 and D54 cells, the heterozygous WT/D6 and the WT/D54 cell lines proliferated significantly more than WT (P = 0.01) or MIGR (P = 0.01) NIH3T3 cells, after 84 h of culture.

Figure 1.

Figure 1

Open in a new tab

Proliferation rates of the NIH3T3‐infected cell lines. The proliferation of NIH3T3 control, WT, D6, D54, WT/D6 and WT/D54 cells was compared. Cells were plated at 50,000 cells per well and then starved for 4 h before incubation with DMEM supplemented 1% serum. After 12, 36, 60 and 84 h of culture, the cells were harvested and counted. The results are expressed in percentage of amplification to the 12 h time point (means of triplicate ± standard deviation).

3.2. The mature 145 kDa form of WT KIT is down‐regulated by KIT activation

Western blot analysis of cell lines confirmed the production of the KIT protein in two forms of 145 and 125 kDa (Supplementary Figure 2). Protein extracts treated with PNGase, which cleaves all N‐linked oligosaccharide residues, did not contain either of these two forms, but did contain a KIT immunoreactive band at 100 kDa. Treatment of protein extracts with EndoH, which only cleaves high‐mannose structures, eliminated the 125 kDa form but not the 145 kDa form, with the appearance of the 100 kDa form (Supplementary Figure 2).

In WT KIT cells, the 145 kDa form was 2.0 times more abundant than the 125 kDa form in the absence of SCF (Figure 2a and d). Addition of SCF did not significantly modify the amount of the 125 kDa form, but halved the amount of the 145 kDa form, such that the 145/125 ratio was 0.9 (P = 0.0001). IM restored the 145/125 ratio to 2.0, confirming that the lower concentration of the 145 kDa form in the presence of SCF was related to KIT activation and consecutive degradation (Figure 2).

Figure 2.

Figure 2

Open in a new tab

Effects of SCF, imatinib and ER stressors on KIT glycosylation and signaling pathways. Western blot analysis of WT (2a), D6 (2b) and WT/D6 (2c) cell lysates to determine total expression (KIT, AKT, ERK1,2, β‐actin) or phosphorylated fraction (pAKT, p‐ERK1,2) of proteins. Cells were treated during 5 h with SCF, imatinib (IM), brefeldin A and/or tunicamycin as indicated. Photographies are representative of two (ER stressors) and three (controls) independent experiments. p = phospho. (2d) The relative expression of the 145/125 kDa forms of KIT was quantified and compared among the different cell lines in the absence of ER stressor and in the presence or not of SCF or imatinib as indicated. The results are represented as ratio to the WT with SCF (means of 3 experiments ± standard deviation). P < 0.05; †: P < 0.005; ‡: P < 0.005.

3.3. Hemizygous but not heterozygous cells expressing mutant forms of KIT have a glycosylation pattern different from that in the WT

Both 125 and 145 kDa forms of KIT were present in cells expressing either D6 or D54 mutants (Figure 2b and d). However, the 145/125 ratio was always <1 in hemizygous mutants, with a mean value of 0.7 for D6 and 0.6 for D54 (P = 0.0001 and P = 0.0005 for the difference with WT, respectively). SCF treatment did not change the 145/125 ratio in hemizygous mutants. IM increased the amount of the 145 kDa form, such that the 145/125 ratio was 3.0 in D6 and 1.8 in D54 cells. These ratios in the presence of IM were thus 4.3 (D6) and 3.5 (D54) times higher than in the presence of SCF alone (P = 0.001 for both).

The 145/125 ratios in heterozygous cells were more similar to those in WT cells than those in hemizygous cells (Figure 2). Indeed, the 145/125 ratios were 1.9 in WT/D6 and 1.8 in WT/D54 cells in the absence of SCF. Addition of SCF to WT/D6 reduced amount of the 145 kDa form 1.9 times, such that the ratio was 1.02 (P = 0.02). With IM, the 145/125 ratio in WT/D6 cells was restored to 2.1. The IM‐induced increase of the 145/125 ratio was significantly higher in D6 than in either WT or WT/D6 cells (respectively, P = 0.03 and P = 0.04). The findings for WT/D54 cells were similar (Figure 2d).

3.4. Brefeldin A and tunicamycin affect the expression of both forms of WT KIT

Brefeldin A inhibits trafficking of proteins from the ER to the Golgi. There was 71–87% less of the 145 kDa form in cells treated with brefeldin A than in cells grown in the same conditions (no SCF, SCF alone or SCF + IM) (Figure 2a).

Tunicamycin inhibits asparagine N‐linked glycosylation of proteins in the ER. Concomitant treatment with tunicamycin and SCF was associated with the almost complete loss of the 145 kDa form of KIT (99% reduction), and in the absence of KIT activation tunicamycin reduced the abundance of the 145 kDa form to a lesser extent (Figure 2a). Unlike brefeldin A, tunicamycin treatment led to the complete loss of the 125 kDa form and appearance of a 100 kDa form that was not dependent on KIT activation.

3.5. Glycosylation of KIT in heterozygous mutants is similar to that in WT and differs from that in hemizygous mutants

Brefeldin A strongly down‐regulated the 145 kDa form in WT and mutated cell lines, in all conditions tested (−79% to −99%; Figure 2). However, in hemizygous cell lines, this down regulation was almost complete and not dependent on the presence of SCF or IM, whereas it was only partial in the absence of KIT activation in WT and heterozygous mutants. Brefeldin A up regulated the 125 kDa form weakly (+30%) in hemizygous cell lines, and more strongly (+42% to 55%) in both WT and heterozygous cell lines.

Tunicamycin treatment led to a major decrease of the 125 kDa form and the appearance of 100 kDa WT and M forms of KIT in all tested conditions. The 145 kDa form also decreased in all conditions, but was still detectable in the absence of KIT activation. In the absence of SCF, heterozygous cells had a response pattern more similar to that of WT cells than to that of hemizygous M cells (Figure 2).

3.6. Membrane expression of WT KIT depends on both maturation through the endoplasmic reticulum (ER) and degradation following activation

SCF significantly decreased the percentage of KIT‐positive cells (P = 0.008) (Figure 3a). By contrast, IM restored and increased the number of KIT‐positive cells after SCF treatment (P = 0.004) (Figure 3a).

Figure 3.

Figure 3

Open in a new tab

Modification of KIT membrane expression upon treatment with SCF, imatinib (IM), brefeldin A and/or tunicamycin. Flow cytometric analyses of WT (3a), D6 (3b), WT/D6 (3c), D54 (3e), WT/D54 (3f) after 3 h of incubation with the different culture conditions. The mean of fluorescence resulting from KIT cell surface immunostaining was expressed as percentage of the control medium condition for each cell line (±standard deviation; n≥ of at least two experiments). *: P < 0.05; **: P < 0.01; †: P < 0.005; ‡P < 0.001. (3d) Similar analysis after addition of brefeldin A or tunicamycin compared to control conditions for WT, D6 and WT/D6 cells. *: P < 0.05 statistically significant; **: P < 0.005; °: statistical difference incalculable.

The surface expression of KIT was 54% lower for WT cells treated with brefeldin A than for untreated controls (Figure 3a). Furthermore treatment with both brefeldin A and SCF significantly decreased KIT expression as compared to SCF alone (P = 0.007) (Figure 3d). IM treatment in the presence of brefeldin A reduced the membrane expression of KIT by 59% (P = 0.02), evidence of KIT activation‐related degradation (Figure 3a).

We tested the membrane expression of KIT in the absence of SCF activation was due to persistence of previously synthesized and non‐activated KIT receptor. We pretreated cells with SCF for 30 min before adding brefeldin A alone; this resulted in a large decrease of KIT membrane expression (Figure 4a).

Figure 4.

Figure 4

Open in a new tab

Brefeldin A did not affect the synthesis of KIT but altered its cell surface expression. (4a): Flow cytometry analysis of KIT cell surface expression of D6 cell line upon treatment with brefeldin A, with or without pre‐treatment of SCF. Resulting histograms obtained with the cytometer showed relative cell number in function of relative log KIT‐PE fluorescence. (4b): Total KIT cell expression was analyzed after similar treatments using permeabilization of cells. The results are representative of two independent experiments.

The effects of tunicamycin were similar to those of brefeldin A (Figure 3a).

3.7. Membrane renewal of KIT mutants is inhibited by brefeldin A

Treatment with SCF had no significant effects on the membrane expression of KIT in cells expressing KIT mutations (Figure 3). By contrast, IM strongly up regulated the membrane expression of KIT in hemizygous cells (D6, +270%, P = 0.002; D54, +340%, P = 0.005). This up regulation was weaker in heterozygous cells (WT/D6, +43%, P = 0.002; WT/D54, +16%, P = 0.09). Activation‐dependent surface expression of KIT was also quantified as the ratio: [positive cells in the presence of IM and SCF]/[positive cells in the presence of SCF]. These ratios were 1.7, 5.3 and 2.6 for WT, D6 and WT/D6 cells, respectively (P = 0.04).

Brefeldin A reduced KIT membrane expression in both D6 and WT/D6 cells more than in WT controls in the same cell culture conditions (Figure 3). We used FACS analysis of permeabilized cells to determine the total KIT content (ie. membrane and intra‐cellular). The total amount of KIT per cell was not modified by brefeldin A, indicating that it did not affect the synthesis of KIT (Figure 4b).

The effects of tunicamycin (decrease) and of IM (increase) on KIT membrane expression were significantly larger in hemizygous cells than in heterozygous cells (Figure 3).

3.8. KIT phosphorylation and activation of downstream signaling pathways

The immature form of KIT was predominant and constitutively phosphorylated in hemizygous mutants. In heterozygous cells there were similar proportions of the mature and immature forms, and both the immature and the mature forms were constitutively phosphorylated in the absence of SCF (Figure 5).

Figure 5.

Figure 5

Open in a new tab

Western blot analysis of KIT expression and phosphorylation after long term treatments. The NIH3T3 cell lines were incubated as indicated with or without SCF and/or Imatinib (IM) for 24 h before cell lysis. KIT protein expression and phosphorylation of the Tyr703 residue were analyzed on total cell lysates of NIH3T3 cell lines (CT = NIH3T3 containing empty vector).

In WT cells, activation with SCF significantly increased AKT, ERK1 and ERK2 phosphorylation (P = 0.01, P = 0.03 and P = 0.009, respectively; Figures 2a and 6a). Activation of AKT and MAP kinase pathways was inhibited by IM. Both brefeldin A and tunicamycin inhibited the SCF‐induced phosphorylation of AKT, ERK1 and ERK2, and the baseline phosphorylation of ERK2 (Figure 2a).

Figure 6.

Figure 6

Open in a new tab

Quantification of western‐blotting experiments analyzing ERK1,2 and AKT phosphorylations. Western‐blots experiments performed on WT (6a), D6 (6b) and WT/D6 (6c) cells upon indicated treatments, were quantified. Graphs represent ratio of proteins of interest to the actin; ratio of the control condition being set to 1. *: P < 0.05; **: P < 0.01.

In hemizygous cells, SCF increased AKT phosphorylation slightly (37% for D6), and did not affect the phosphorylation for ERK1 and ERK2 proteins (Figure 6b). IM was associated with a reduction of AKT, ERK1 and ERK2 phosphorylations in D6 cells compared to control, respectively of 58% (P = 0.07), of 55% (P = 0.004), and of 62% (P = 0.01).

Among theheterozygous cells, WT/D6 cells had the same profile as WT cells with SCF treatment leading to a very large increase of AKT phosphorylation (600%), and large increases of ERK1 (290%) and ERK2 (270%) phosphorylation (Figure 6c). Thus, the activities of MAP Kinase and AKT in WT/D6 cells were very similar to those in WT cells.

In D6 and WT/D6 cells, AKT phosphorylation decreased substantially following simultaneous treatment with both brefeldin A and SCF (Figures 2 and 6).

4. Discussion

Gain of function mutations of KIT is oncogenic in several tumors, and are targets for treatments with tyrosine kinase receptor inhibitors. Therefore a lot of work has been performed on in vitro models to improve knowledge of biology of WT and mutant forms of KIT. Surprisingly, although KIT mutations are heterozygous in most tumors, in vitro experiments were almost always performed in hemizygous cell models (Kitayama et al., 1995; Tsujimura et al., 1999; Jahn et al., 2007; Xiang et al., 2007; Tabone‐Eglinger et al., 2008; Bougherara et al., 2009; Sun et al., 2009; Yang et al., 2010; Kim et al., 2011). We show here that biology of heterozygous KIT mutants is more reminiscent of that of WT than of hemizygous mutant forms. This suggests that in vitro modeling of tumors should be performed with heterozygous rather hemizygous cell models. Our study was performed on two types of exon 11 KIT mutations, which is the most frequently mutated exon in human GISTs (Emile et al., 2004; Rubin et al., 2007; Bachet and Emile, 2010). Thus, our results deserve to be confirmed with other types of mutations and in other cell models.

We investigated differences of KIT glycosylation among the WT, hemizygous and heterozygous cell lines. Post‐translational modifications of proteins begin early in ER with addition of a generic N‐linked high‐mannose oligosaccharide to the new polypeptide chain when this one emerges from the translocon (Imperiali and Rickert, 1995). Then, the glycoproteins move to the Golgi where the glycan chains will be modified to produce complex N‐linked glycans and/or high‐mannose glycans. These last modifications allow the biosynthesis of mature glycoproteins which will move to their final destinations and will act with specific functions (Helenius and Aebi, 2001). To investigate differences in KIT maturation between cell lines, we used two ER stressors, the brefeldin A which targets the transport of protein from the ER to the Golgi body and the tunicamycin which inhibits the N‐glycosylation (Mehmet, 2000). Our results confirmed that the 145 kDa contains a complex glycosylation pattern and the precursor 125 kDa form of KIT contains a generic N‐linked high‐mannose oligosaccharide specific to the ER. The 100 kDa form of KIT is the native protein without the generic N‐linked high‐mannose oligosaccharide which is added to the new polypeptide chain when this one emerges from the translocon (Bougherara et al., 2009). As already published, the preponderant form of KIT was the 125 kDa form in hemizygous mutant cell lines (Xiang et al., 2007; Tabone‐Eglinger et al., 2008; Bougherara et al., 2009; Yang et al., 2010). We have shown here that the ratio of the two forms of KIT (145 kDa/125 kDa) in heterozygous mutant was more similar to the WT than to hemizygous mutants. Moreover, the response patterns of heterozygous cell lines to ER stressors were also more similar to WT than hemizygous cells. These results were observed in WT/D6 cells as in WT/D54 cells, and this despite the D54 allele was two times more abundant than WT allele in the WT/D54 cell line. In human heterozygous GISTs, the both alleles, mutated KIT and WT KIT, are expressed in similar proportion (Théou et al., 2004).

Aberrant intra‐cellular compartmentalization and activation of mutant receptor tyrosine kinase (RTK) has been previously reported with KIT (Xiang et al., 2007; Tabone‐Eglinger et al., 2008; Bougherara et al., 2009; Yang et al., 2010) and other RTKs, such as FLT3, fibroblast growth factor receptor 3 (FGFR3) and PDGFRA (Clarke and Dirks, 2003; Gibbs and Legeai‐Mallet, 2007; Choudhary et al., 2009). In WT KIT cells, the 145 kDa form is the only to be phosphorylated in the presence of SCF, and KIT activation is associated with marked decrease KIT membrane expression reflecting its internalization and degradation (Kitayama et al., 1995; Tabone‐Eglinger et al., 2008; Bougherara et al., 2009). Deletion of extra‐cellular domain of WT KIT is associated with a loss of WT KIT activity whereas it has no effect on mutant KIT showing that membrane expression of mutant KIT is not necessary for activation of the receptor (Tsujimura et al., 1999). Moreover, intra‐cellular receptor signaling of mutant KIT is sufficient to drive neoplasia in vivo (Kitayama et al., 1995; Xiang et al., 2007). However, despite the absence of membrane expression of mutant KIT, in vitro data suggest that receptor self‐association in the intra‐cellular compartment may be important for activation of downstream pathways (Tsujimura et al., 1999). In cell models, the aberrant compartmentalization and activation of mutant KIT has been reported with mutations of the juxta‐membrane domain and of the phosphotransferase domain, and treatment with inhibitors restores the membrane expression (Xiang et al., 2007; Tabone‐Eglinger et al., 2008; Bougherara et al., 2009; Yang et al., 2010). Interestingly, heterozygous cell lines harbored both the characteristics of WT KIT and mutant KIT. On the one hand, a constitutive phosphorylation of the 125 kDa and 145 kDa forms of KIT was observed in the absence of SCF stimulation, and was abolished in the presence of IM. On the other hand, as for WT KIT but not hemizygous mutants, the cell surface expression of the 145 kDa form in heterozygous mutants was not abolished by brefeldin A in the absence of SCF.

The intra‐cellular activation of the 125 kDa form of mutated KIT, as the paracrine activation of the 145 kDa form of WT KIT, seems to be able to mediate oncogenic signaling pathways and tumor proliferation (Xiang et al., 2007; Tabone‐Eglinger et al., 2008). However, despite common pro‐oncogenic capacities, WT and mutated KIT might activate different signaling pathways according to their intra‐cellular location and interactions with others proteins. KIT recruitment to lipid rafts is required for PI3K/AKT activation (Jahn et al., 2007). Difference in AKT activation has been previously described in Ba/F3 cell lines transfected with WT KIT or two mutated forms of KIT, the V559D and the D816V (Yang et al., 2010). Whereas both WT KIT and V559D KIT mutant induced AKT phosphorylation, D816V was associated with the absence or a negative regulation of this pathway. Interestingly, cell surface dimerization of the V560G KIT mutant was previously reported, whereas D816V KIT mutant was not or very few associated with receptor association in the extra‐cellular domain (Kitayama et al., 1995). Treatment of mutant KIT‐ITD (internal tandem duplication in the juxta‐membrane domain) cell lines with brefeldin A was associated with an inhibition of KIT membrane expression and of phosphorylation of AKT (Choudhary et al., 2009). In the present study, AKT was less phosphorylated in hemizygous cell lines than in WT and heterozygous cell lines in the presence of SCF. Moreover, the inhibition of KIT membrane expression with brefeldin A was associated with a major decrease of phosphorylated AKT in WT and heterozygous cell lines. These data suggest that membrane expression of KIT may be necessary to activate the PI3K/AKT pathway.

Differences between hemizygous and heterozygous cell lines may be related to receptor dimerization. Ligand binding followed by KIT dimerization is the initial step before activation of intrinsic protein kinase domain and activation of downstream signaling pathways (Roskoski, 2005). In the absence of SCF, dimerized form of the V559D mutant KIT was found in hemizygous cell lines suggesting that this mutation may yield receptor dimerization without ligand binding (Kitayama et al., 1995). Conflicting data have been reported for D816V mutant KIT. In hemizygous cell lines, dimerized form of the D816V mutant KIT was almost absent without SCF stimulation suggesting that mutation within the kinase domain may be associated with an activation of the receptor without dimerization (Kitayama et al., 1995). At the opposite, KIT D816Y and KIT‐ITG mutants formed homodimers and heterodimers with WT KIT in the absence of SCF stimulation in heterozygous cell lines (Kim et al., 2011). We did not assess the KIT homo‐ or hetero‐dimerization in the present study. However, the constitutive activation of the 145 kDa form in heterozygous mutants in the absence of SCF suggests that the WT and the mutant forms associate together.

In our proliferation study, the growth curves of D6 and WT/D6 cells were very similar while the growth curve of WT/D54 cells was superior to this of D54 cells. The D6 and D54 mutations have probably not the same biologic effects. The D54 deletions are responsible for the loss of the two tyrosines residues Tyr568 and Tyr570 which are implicated in the Rac1/JNK pathway, the Ras/MAP kinase pathway, the negative regulation and the degradation of KIT (Roskoski, 2005). Consequently, a deletion of both Tyr568 and Tyr570 could disturb not only the autoinhibition conformation of exon 11 but also alter some protein–protein interactions and signaling pathway activations. The D6 mutation through the preservation of the two tyrosine residues could be more aggressive than the D54 mutations, as it has been suggested in the literature (Martin et al., 2005). Consequently, the addition of the WT allele could less increase the proliferative capacity of the D6 cells than the D54 cells. This point should be assessed in future studies.

Although most tumors with KIT or other RTK mutations are heterozygous, almost all the in vitro data have been obtained in hemizygous models. The results we obtained concerning glycosylation, cellular localization and activation of WT and hemizygous mutant forms of KIT were similar to the data previously published, thus confirming the value of our cell model. We show here that in heterozygous conditions of KIT, the glycosylation, cellular localization and activation of KIT are more similar to WT than to hemizygous mutant condition. Thus future in vitro studies on RTK should include validation on heterozygous cell models.

Conflict of interest

The authors declare no conflict of interest.

Support/Grants

This work was supported in part by non profit organism AREP.

Supporting information

The following are the supplementary data related to this article:

Supplementary Figure 1 mRNA expression of both KIT alleles in hemizygous and heterozygous cell lines. 1a: KIT alleles were detected by length analysis for fluorescent PCR product: WT KIT was close to 356 pb, D6 KIT allele was close to 350 pb and D54 KIT allele was to 302 pb 1b: KIT alleles were detected by TaqMan analyses. The level of KIT mRNA (mutated or WT KIT) was similar in all cell lines, homozygous as heterozygous.

Click here for additional data file. (48.9KB, pdf)

Supplementary Figure 2 Analysis of KIT WT glycosylation by western blot. Proteins were denatured and treated with either PNGase or EndoH N‐glycosidase as compared to control, untreated cells. β‐actin antibody was the loading control. Immature form was sensitive to EndoH which removed mannose‐rich moieties and PNGase treatment removed all glycosylated moieties of KIT.

Click here for additional data file. (20.5KB, pdf)

Acknowledgments

The authors would like to thank Ph. Aegerter, P. Martel‐Samb and Z. Hélias‐Rodzewicz for fruitful discussions and J. Tisserand for contributing in western blot analyses.

Supplementary data 1.

1.1.

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.molonc.2012.10.008.

Brahimi-Adouane Sabrina, Bachet Jean-Baptiste, Tabone-Eglinger Séverine, Subra Frédéric, Capron Claude, Blay Jean-Yves, Emile Jean-François, (2013), Effects of endoplasmic reticulum stressors on maturation and signaling of hemizygous and heterozygous wild‐type and mutant forms of KIT, Molecular Oncology, 7, doi: 10.1016/j.molonc.2012.10.008.

References

  1. Bachet, J.B. , Emile, J.F. , 2010. Diagnostic criteria, specific mutations and genetic predisposition in gastrointestinal stromal tumors. Appl. Clin. Genet. 3, 85–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bachet, J.B. , Hostein, I. , Le Cesne, A. , 2009. Prognosis and predictive value of KIT exon 11 deletion in GISTs. Br. J. Cancer 101, 7–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Blanke, C.D. , Rankin, C. , Demetri, G.D. , 2008. Phase III randomized, intergroup trial assessing imatinib mesylate at two dose levels in patients with unresectable or metastatic gastrointestinal stromal tumors expressing the Kit receptor tyrosine kinase: S0033. J. Clin. Oncol. 26, 626–632. [DOI] [PubMed] [Google Scholar]
  4. Bougherara, H. , Subra, F. , Crépin, R. , 2009. The aberrant localization of oncogenic KIT tyrosine kinase receptor mutants is reversed on specific inhibitory treatment. Mol. Cancer Res. 7, 1525–1533. [DOI] [PubMed] [Google Scholar]
  5. Buti, S. , Donini, M. , Sergio, P. , 2011. Impressive response with imatinib in a heavily pretreated patient with metastatic c-KIT mutated thymic carcinoma. J. Clin. Oncol. 29, e803–e805. [DOI] [PubMed] [Google Scholar]
  6. Cassier, P.A. , Ducimetière, F. , Lurkin, A. , 2010. A prospective epidemiological study of new incident GISTs during two consecutive years in Rhône Alpes region: incidence and molecular distribution of GIST in a European region. Br. J. Cancer 103, 165–170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Choudhary, C. , Olsen, J.V. , Brandts, C. , 2009. Mislocalized activation of oncogenic RTKs switches downstream signaling outcomes. Mol. Cell. 36, 326–339. [DOI] [PubMed] [Google Scholar]
  8. Clarke, I.D. , Dirks, P.B. , 2003. A human brain tumor-derived PDGFR-alpha deletion mutant is transforming. Oncogene 22, 722–733. [DOI] [PubMed] [Google Scholar]
  9. DeMatteo, R.P. , Gold, J.S. , Saran, L. , 2008. Tumor mitotic rate, size, and location independently predict recurrence after resection of primary gastrointestinal stromal tumor (GIST). Cancer 112, 608–615. [DOI] [PubMed] [Google Scholar]
  10. Emile, J.F. , Théou, N. , Tabone, S. , 2004. Clinicopathologic, phenotypic, and genotypic characteristics of gastrointestinal mesenchymal tumors. Clin. Gastroenterol. Hepatol. 2, 597–605. [DOI] [PubMed] [Google Scholar]
  11. Emile, J.F. , Bachet, J.B. , Tabone-Eglinger, S. , 2008. GIST with homozygous KIT exon 11 mutations. Lab. Invest. 88, 456–457. [DOI] [PubMed] [Google Scholar]
  12. Emile, J.F. , Brahimi, S. , Coindre, J.M. , 2012. Frequencies of KIT and PDGFRA mutations in the MolecGIST prospective population-based study differ from those of advanced GISTs. Med. Oncol. 29, 1765–1772. [DOI] [PubMed] [Google Scholar]
  13. Gibbs, L. , Legeai-Mallet, L. , 2007. FGFR3 intracellular mutations induce tyrosine phosphorylation in the Golgi and defective glycosylation. Biochim. Biophys. Acta 1773, 502–512. [DOI] [PubMed] [Google Scholar]
  14. Helenius, A. , Aebi, M. , 2001. Intracellular functions of N-linked glycans. Science 291, 2364–2369. [DOI] [PubMed] [Google Scholar]
  15. Hirota, S. , Isozaki, K. , Moriyama, Y. , 1998. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 279, 577–580. [DOI] [PubMed] [Google Scholar]
  16. Imperiali, B. , Rickert, K.W. , 1995. Conformational implications of asparagine-linked glycosylation. Proc. Natl. Acad. Sci. U.S.A. 92, 97–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Jahn, T. , Leifheit, E. , Gooch, S. , 2007. Lipid rafts are required for Kit survival and proliferation signals. Blood 110, 1739–1747. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kim, S.Y. , Kang, J.J. , Lee, H.H. , 2011. Mechanism of activation of human c-KIT kinase by internal tandem duplications of the juxtamembrane domain and point mutations at aspartic acid 816. Biochem. Biophys. Res. Commun. 410, 224–228. [DOI] [PubMed] [Google Scholar]
  19. Kitayama, H. , Kanakura, Y. , Furitsu, T. , 1995. Constitutively activating mutations of c-kit receptor tyrosine kinase confer factor-independent growth and tumorigenicity of factor-dependent hematopoietic cell lines. Blood 85, 790–798. [PubMed] [Google Scholar]
  20. Lasota, J. , vel Dobosz, A.J. , Wasag, B. , 2007. Presence of homozygous KIT exon 11 mutations is strongly associated with malignant clinical behavior in gastrointestinal stromal tumors. Lab. Invest. 87, 1029–1041. [DOI] [PubMed] [Google Scholar]
  21. Martin, J. , Poveda, A. , Llombart-Bosch, A. , 2005. Deletions affecting codons 557–558 of the c-KIT gene indicate a poor prognosis in patients with completely resected gastrointestinal stromal tumors: a study by the Spanish Group for Sarcoma Research (GEIS). J. Clin. Oncol. 23, 6190–6198. [DOI] [PubMed] [Google Scholar]
  22. Mehmet, H. , 2000. Caspases find a new place to hide. Nature 403, 29–30. [DOI] [PubMed] [Google Scholar]
  23. Nakai, N. , Ishikawa, T. , Nishitani, A. , 2008. A mouse model of a human multiple GIST family with KIT-Asp820Tyr mutation generated by a knock-in strategy. J. Pathol. 214, 302–311. [DOI] [PubMed] [Google Scholar]
  24. Roskoski, R. , 2005. Structure and regulation of Kit protein-tyrosine kinase – the stem cell factor receptor. Biochem. Biophys. Res. Commun. 338, 1307–1315. [DOI] [PubMed] [Google Scholar]
  25. Rubin, B.P. , Heinrich, M.C. , Corless, C.L. , 2007. Gastrointestinal stromal tumour. Lancet 369, 1731–1741. [DOI] [PubMed] [Google Scholar]
  26. Staser, K. , Yang, F.C. , Clapp, D.W. , 2010. Mast cells and the neurofibroma microenvironment. Blood 116, 157–164. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sun, J. , Pedersen, M. , Rönnstrand, L. , 2009. The D816V mutation of c-kit circumvents a requirement for Src family kinases in c-kit signal transduction. J. Biol. Chem. 284, 11039–11047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Tabone-Eglinger, S. , Subra, F. , El Sayadi, H. , 2008. KIT mutations induce intracellular retention and activation of an immature form of the KIT protein in gastrointestinal stromal tumors. Clin. Cancer Res. 14, 2285–2294. [DOI] [PubMed] [Google Scholar]
  29. Théou, N. , Tabone, S. , Saffroy, R. , 2004. High expression of both mutant and wild-type alleles of c-kit in gastrointestinal stromal tumors. Biochim. Biophys. Acta 1688, 250–256. [DOI] [PubMed] [Google Scholar]
  30. Tsujimura, T. , Hashimoto, K. , Kitayama, H. , 1999. Activating mutation in the catalytic domain of c-kit elicits hematopoietic transformation by receptor self-association not at the ligand-induced dimerization site. Blood 93, 1319–1329. [PubMed] [Google Scholar]
  31. Willmore-Payne, C. , Holden, J.A. , Tripp, S. , 2005. Human malignant melanoma: detection of BRAF- and c-kit-activating mutations by high-resolution amplicon melting analysis. Hum. Pathol. 36, 486–493. [DOI] [PubMed] [Google Scholar]
  32. Xiang, Z. , Kreisel, F. , Cain, F. , 2007. Neoplasia driven by mutant c-KIT is mediated by intracellular, not plasma membrane, receptor signaling. Mol. Cell. Biol. 7, 267–282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Yang, Y. , Létard, S. , Borge, L. , 2010. Pediatric mastocytosis-associated KIT extracellular domain mutations exhibit different functional and signaling properties compared with KIT-phosphotransferase domain mutations. Blood 116, 1114–1123. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

The following are the supplementary data related to this article:

Supplementary Figure 1 mRNA expression of both KIT alleles in hemizygous and heterozygous cell lines. 1a: KIT alleles were detected by length analysis for fluorescent PCR product: WT KIT was close to 356 pb, D6 KIT allele was close to 350 pb and D54 KIT allele was to 302 pb 1b: KIT alleles were detected by TaqMan analyses. The level of KIT mRNA (mutated or WT KIT) was similar in all cell lines, homozygous as heterozygous.

Click here for additional data file. (48.9KB, pdf)

Supplementary Figure 2 Analysis of KIT WT glycosylation by western blot. Proteins were denatured and treated with either PNGase or EndoH N‐glycosidase as compared to control, untreated cells. β‐actin antibody was the loading control. Immature form was sensitive to EndoH which removed mannose‐rich moieties and PNGase treatment removed all glycosylated moieties of KIT.

Click here for additional data file. (20.5KB, pdf)

Articles from Molecular Oncology are provided here courtesy of Wiley

RESOURCES