Activating mutations of the ALK gene have been identified in sporadic and familial cases of neuroblastoma (NB), a cancer of the peripheral nervous system, and are thought to be the primary mechanism of oncogenic activation of this receptor in this pediatric neoplasm. To address the possibility that ALK activation may occur through genomic rearrangements as detected in other cancers, we first took advantage of high-resolution array-comparative genomic hybridization to search for ALK rearrangements in NB samples. Using complementary experiments by capture/paired-end sequencing and FISH experiments, various types of rearrangements were fully characterized, including partial gains or amplifications, in several NB cell lines and primary tumors. In the CLB-Bar cell line, we described a genomic rearrangement associated with an amplification of the ALK locus, leading to the expression of a 170 kDa protein lacking part of the extracellular domain encoded by exons 4 to 11, named ALKΔ4-11. Analysis of genomic DNA from the tumor at diagnosis and relapse revealed that the ALK gene was amplified at diagnosis but that the rearranged ALK allele was observed at the relapse stage only, suggesting that it may be implicated in tumor aggressiveness. Consistently, oncogenic and tumorigenic properties of the ALKΔ4-11 variant were shown after stable expression in NIH3T3 cells. Moreover, we documented an increased constitutive kinase activity of this variant, as well as an impaired maturation and retention into intracellular compartments. These results indicate that genomic rearrangements constitute an alternative mechanism to ALK point mutations resulting in receptor activation. Cancer Res; 73(1); 195–204. ©2012 AACR.

Neuroblastoma (NB) is an embryonal cancer of the peripheral sympathetic nervous system observed in early childhood, characterized by a broad spectrum of clinical behaviors and a great genetic heterogeneity (1–3). The ALK gene, located at 2p23, encodes a receptor tyrosine kinase (RTK) that belongs to the insulin-receptor superfamily (4, 5). Activating ALK mutations, mainly affecting residues located within the kinase domain, have been identified in sporadic, familial, and syndromic cases of NB (6–11) and ALK amplifications have also been reported in a subset of sporadic cases (2, 12). Furthermore, it has been suggested that the wild-type (WT) ALK receptor may exert oncogenic activity in NB when a critical threshold of expression is achieved (13).

In addition to ALK aberrations, other recurrent somatic genetic abnormalities have been characterized in-depth in NB samples. MYCN amplification at chromosome 2p24 is observed in about one-fourth of primary tumors and is associated with advanced-stage disease and rapid tumor progression (1, 14). Losses of chromosome arms 1p, 3p, and 11q and gains of chromosome arms 1q, 2p, and 17q are recurrently observed in NB samples and are of prognostic significance (1–3, 15). The breakpoints on these various segments seem not to be tightly clustered however their precise position remains unknown in most cases.

In several human cancers, including a subset of anaplastic large cell lymphomas (ALCL), diffuse large B-cell lymphomas, inflammatory myofibroblastic tumors (IMT), and non–small cell lung (NSCL) cancers, recurrent chromosome translocations targeting the ALK gene have been extensively characterized (4, 5). The resulting fusion proteins contain invariably the tyrosine kinase domain of ALK and the various partners provide an oligomerization domain. This leads to constitutive ALK signaling that has been shown to induce cell transformation in vitro and in vivo (4, 5). The ALK receptor is, therefore, an attractive therapeutic target in an expanding number of human malignancies, both in children and adults (16).

In the present work, we fully characterized various types of rearrangements involving the ALK gene in several NB cell lines and primary tumors. We documented a rearrangement leading to a novel truncated ALK variant, exhibiting oncogenic properties and associated with tumor aggressiveness.

NB cell lines and primary tumors

NB cell lines have been described previously (17–19). The CLB-Bar cell line was established from a stage 4 NB tumor at relapse. A subset of primary NB tumors previously characterized by whole-genome array-comparative genomic hybridization (CGH) was further analyzed in this study (15).

Array-CGH analysis

A custom NimbleGen ultrahigh density oligonucleotide array was designed for this study. We used a 4-plex format with 72,000 features starting from chromosome 2 position 29,249,000 to position 30,018,000 [all base pair positions are referring to human genome build 36 (hg18)], thereby covering the entire ALK gene and extending 20 kb upstream and downstream of the gene with 1 feature every 11 bases. The precise procedure is described in supplementary methods. Twenty-six cell lines were analyzed on this array: 106C, CHP212, CLB-Ba, CLB-Bar, CLB-Br, CLB-Car, CLB-Ga, CLB-Ge, CLB-Ma, CLB-Re, CLB-Tr, GI-M-EN, IMR32, LAN-1, N206/Kelly, SJNB-1, SJNB-6, SJNB-8, SJNB-12, SK-N-AS, SK-N-BE, SK-N-BE(2C), SK-N-BE(M17), SK-N-DZ, SK-N-SH, and TR14. We also studied three primary tumors showing amplification at the ALK locus, 1 tumor with amplification near the ALK locus and 9 tumors exhibiting 2p gain, with breakpoint in the vicinity of the ALK gene.

For a subset of samples, Affymetrix Genome-Wide Human SNP 6.0 Array was used to obtain a whole genomic profile according to Affymetrix recommendations (Supplementary Methods).

Capture and paired-end sequencing

The Sure Select method (Agilent Technologies) was used to capture regions of interest. The bait library was designed using the Agilent e-array software. The complete ALK genomic locus, together with 50 kb upstream and 30 kb downstream of the gene, was covered at ×5, excluding repeated regions. Boundaries of other amplified regions on the 2p chromosome of IMR32 and NB838 samples, defined by SNP 6.0 arrays, were also included in the captured regions (Supplementary Table S1). Libraries for paired-end sequencing were constructed for 10 samples (CLB-Bar, CLB-Br, CLB-Ga, CLB-Ge, CLB-Ma, CLB-Re, IMR32, N206, NB838, and SJNB8) from genomic DNA according to the manufacturer's protocol (Illumina). Capture was conducted on the bait library according to the protocol provided by Agilent Technologies. Paired-end sequencing (2 × 50 bp) was conducted on the Illumina Genome Analyzer II. Reads were analyzed as described in Supplementary Methods.

PCR validation of identified rearrangements

Information about primers used to confirm rearrangements predicted by capture and paired-end sequencing is listed in Supplementary Table S2.

FISH experiments

For FISH experiments, we used either the LSI ALK dual color, Break Apart Rearrangement probe (Vysis) or bacterial artificial chromosome (BAC) probes or five-kb PCR products synthesized using TaKaRa ExTaq (Lonza) labeled with various fluorochromes by a nick-translation kit (Vysis). Primers used to generate the PCR fragment probes as well as BACs used for FISH experiments are listed in Supplementary Table S3. FISH experiments were conducted as previously described (19). Images were captured with Metamorph software and analyzed with ImageJ software.

Immunoblots

Total ALK was detected using Zymed 18-0266 rabbit polyclonal antibody (Invitrogen). An antibody (number 3341, Cell Signaling Technologies) directed against the phosphorylated Y1604 residue was used to reveal phosphorylated ALK. Western blot analyses were conducted using anti-ERK1/2 (9102), antiphospho-ERK1/2 (4377), anti-AKT (9272), antiphospho-AKT (4056), anti-STAT3 (9132), and antiphospho-STAT3 (9145) antibodies from Cell Signalling Technologies.

ALK immunoprecipitation and proteomic analysis

Cell extracts were prepared as previously described (10). 10 mg of proteins were incubated with a mix of rabbit polyclonal and mouse monoclonal anti-ALK antibodies (20) for 2 hours at 4°C then with protein G Sepharose beads (Sigma) for 1 hour at 4°C. The immunoprecipitates were washed twice with cold PBS supplemented with 0.1% Triton then resuspended in gel loading buffer (50 mmol/L Tris pH = 8.2, Urea 8M, Thiourea 2M) and resolved by SDS-PAGE. For proteomic analysis, label-free quantitative analysis of immunoaffinity-purified proteins was conducted as described in Supplementary Methods.

Transforming potential of ALKΔ4–11

NIH3T3 cells were cultured as previously described (21). The cDNA encoding the ALKΔ4–11 variant was generated from total RNA of the CLB-Bar cell line using high-fidelity PCR. Its sequence was checked by Sanger sequencing and cloned into the pcDNA3 expression vector. Stably transfected cells were obtained after transfection using Lipofectamine LTX reagent and selected by 400 μg/mL geneticin (Gibco).

For proliferation assays 20,000 stably transfected cells were plated in duplicates in 24-well plates. The number of living cells was counted 3 and 6 days after using a Vi-cell XR counter (Beckman Coulter). For soft agar assays, 250,000 stably transfected cells were mixed in 0.3% agar with complete medium and plated on 0.3% agarose-coated 6-well plates (in triplicate). After 17 days, colonies were manually counted. Tumor formation was evaluated 21 days after subcutaneous injections of 1 million transfected cells in Nude mice.

In vitro kinase assay

In vitro kinase assay was conducted as previously described (10) using 500 μg of NIH3T3 cell lysates.

Deglycosylation of ALK with F and H-endoglycosidases and cell surface protein biotinylation

Deglycosylation and cell surface protein biotinylation experiments were conducted on NIH3T3 cells stably expressing WT ALK or the ALKΔ4–11 variant as previously described (21).

Copy number changes within the ALK gene in NB samples identified by high-resolution array-CGH

To identify rearrangements targeting the ALK locus, we searched for intragenic transitions in copy number using an ultrahigh density array covering the entire ALK gene in 13 primary tumors selected for ALK amplification or 2p gain and in 26 NB cell lines (see Material and Methods). The whole genomic profile of cases presenting such rearrangements associated with ALK amplification was further explored using SNP 6.0 arrays. Fig. 1A shows amplification of the ALK locus as well as several other amplified regions on the 2p arm in the CLB-Bar, IMR32, and NB838 samples. Analysis of the ALK locus with our ultrahigh density array indicated that these 3 samples had amplified and rearranged ALK alleles. In the CLB-Bar cell line, most of the locus was amplified; however, a 39-kb region from 29,668,100 to 29,707,200 positions in intron 3 showed an absence of amplification (Fig. 1B). In the IMR32 cell line, array-CGH identified a partial amplification of the ALK locus, between positions 29,550,200 and 29,777,700, the corresponding breakpoints being located in introns 4 and 2, respectively. For the NB838 tumor, harboring a whole ALK amplification, we observed a breakpoint in intron 2 close to position 29,790,000. In five samples, copy number changes were observed within the ALK locus in the absence of amplification (Fig. 1B). Whereas in CLB-Br, NB176, and SJNB8, only 1 breakpoint was detected (in introns 11, 4, and exon 23, respectively), 2 breakpoints were observed in the CLB-Ga cell line (in intron 1) and the ALK locus appeared to be highly rearranged in the CLB-Re cell line. Altogether, breakpoints within the ALK locus were detected in 2 tumors and 6 cell lines.

Figure 1.

ALK amplifications and rearrangements revealed by array-CGH in NB samples. A, multiple amplicons on chromosome 2p in NB cases presenting with whole or partial amplification of the ALK gene are revealed by SNP 6.0 arrays in the CLB-Bar, IMR32, and NB838 samples. B, intragenic transitions in copy number within the ALK gene. The ultrahigh density array covers the ALK gene with a median probe interval of 11 bp. As the array contains only the ALK locus, after normalization, the value at 0 for the Log2 ratio corresponds to the mean copy number of the locus. This explains the particular profiles observed for the CLB-Bar and NB838 samples that harbor amplification of most of the ALK gene. In A and B, the 2p ideogram and chart of the exon-intron structure of the ALK gene are shown above the copy number profiles, respectively.

Figure 1.

ALK amplifications and rearrangements revealed by array-CGH in NB samples. A, multiple amplicons on chromosome 2p in NB cases presenting with whole or partial amplification of the ALK gene are revealed by SNP 6.0 arrays in the CLB-Bar, IMR32, and NB838 samples. B, intragenic transitions in copy number within the ALK gene. The ultrahigh density array covers the ALK gene with a median probe interval of 11 bp. As the array contains only the ALK locus, after normalization, the value at 0 for the Log2 ratio corresponds to the mean copy number of the locus. This explains the particular profiles observed for the CLB-Bar and NB838 samples that harbor amplification of most of the ALK gene. In A and B, the 2p ideogram and chart of the exon-intron structure of the ALK gene are shown above the copy number profiles, respectively.

Close modal

Various ALK rearrangements characterized by capture and paired-end sequencing

To further explore ALK rearrangements observed by array-CGH and document the nature of the involved partner, we conducted paired-end sequencing after DNA capture with a specific bait library. Files were processed to identify genomic rearrangements that matched with the breakpoints detected by array-CGH. For the breakpoints identified by array-CGH for which this analysis did not identify any rearrangement, a further exploration of the capture data files was conducted with less stringent filters in the vicinity of the expected regions of breakpoint.

The capture and paired-end sequencing strategy identified an inverted duplication in CLB-Br and a large duplication in SJNB8, which were experimentally validated (Supplementary Table S2). In CLB-Br, the rearrangement was shown to be somatic, as it was not detected in the lymphocytes of the patient (data not shown). This rearrangement targeted the intron 11 of the ALK gene, however the exact position of the breakpoint could not be determined because of homology between regions of the 2p chromosome involved in this duplication, i.e. sequences extending from positions 29,335,931 to 29,336,086 and from positions 29,336,256 to 29,336,425. No fusion gene is expected from this rearrangement. In SJNB8, the ALK gene broken in exon 23 is fused to intron 11 of the CLIP4 gene. These 2 genes are fused in the opposite transcriptional orientation and are therefore not expected to lead to fusion transcripts and proteins.

For the CLB-Re cell line, at least seven copy number transitions were detected within the ALK locus (Fig. 1B) and the capture approach allowed the identification of several intra and interchromosomal rearrangements that were experimentally validated. In particular three deletions were confirmed, as well as 3 interchromosomal rearrangements with chromosome 3 (Supplementary Table S2). FISH analysis revealed the presence of 2 abnormal derivatives in all examined metaphases of this tetraploid cell line (18), which is consistent with a highly rearranged ALK gene in a homogeneous cell population (Supplementary Fig. S1).

In the CLB-Ga cell line, the capture approach identified a rearrangement between the intron 1 of the ALK locus and a telomeric sequence (Supplementary Table S2). The repeated structure of the telomeric sequence precluded the design of primers for PCR confirmation. However, sequencing of larger fragments using a whole-genome mate-pair strategy (unpublished results) identified the same rearrangement and allowed PCR primers design and validation on genomic DNA. The precise nature of the telomere involved in this rearrangement remains unknown at that time.

For the NB176 primary tumor, inverse PCR (Supplementary Methods) allowed the characterization of an unbalanced translocation between the ALK gene and a region of chromosome 1 that does not contain any gene (Supplementary Table S2). The positions of the breakpoints are fully consistent with the array-CGH profile showing 1p deletion and 2p gain in this sample.

The case of the N206/Kelly cell line is different. Indeed, a translocation involving the ALK locus, detected by FISH analysis using the Break Apart Vysis probe, has been reported in this cell line (22). Interestingly, we did not observe any transition in copy number within the ALK locus, neither with our specific array, nor with the SNP 6.0 array. Nevertheless, the capture and paired-end sequencing approach allowed the identification of an interchromosomal rearrangement between the ALK gene and the CLPTM1L gene located on chromosome 5 (Supplementary Table S2). Reverse-transcriptase PCR (RT-PCR) confirmed the expression of a fusion transcript between exon 4 of the ALK gene and exon 8 of the CLPTM1L gene (data not shown). As the CLPTM1L sequence is not in frame, a fusion protein of 405 amino acids is predicted containing the first 385 amino acids of the ALK receptor rapidly followed by a STOP codon. FISH analysis with several BACs scattered over the ALK gene clearly showed that this cell line harbors 2 normal copies of the ALK gene and 1 copy disrupted between positions 29,351,163 and 29,513,137 (Supplementary Fig. 2 and Supplementary Table S2). This observation is, therefore, consistent with the absence of any transition in copy number within the ALK locus and suggests that one part of the disrupted ALK copy may be translocated to chromosome 5.

Complex patterns of 2p coamplification involving the ALK gene

For IMR32, the capture and paired-end sequencing strategy indicated that the 2 breakpoints located on each side of the ALK partial amplicon were pasted to a part of chromosome bands 2p14 and 2p16.2, also amplified in this cell line. Additional links between the other 2p amplicons were detected as indicated in Supplementary Table S2 and confirmed by PCR. Altogether, it is possible to draw a scheme representing the structure of a complex amplicon containing various parts of chromosome 2, including a piece of the ALK gene (Fig. 2A). FISH experiments confirmed that the different parts of this structure were indeed present in a common homogeneous staining region (hsr), previously identified on a derivative chromosome der1 t(1qter → 1p32:17q21 → 17qter(inshsr2)) (ref. 18; Fig. 2B).

Figure 2.

Structure of a complex 2p amplicon involving the ALK locus in IMR32 cell line. A, scheme of the potential structure of a complex amplicon involving several 2p regions including a portion of the ALK gene in the IMR32 cell line. The upper part represents the normal positions of the amplified regions, whereas the lower part deciphers the organization of these regions within the amplicon. The arrows indicate the orientation of each fragment, from the lowest to the highest genomic position. Gene names or cytobands related to amplified regions are indicated. B, FISH experiments confirmed coamplification of part of the ALK locus and five distinct 2p regions in the same hsr.

Figure 2.

Structure of a complex 2p amplicon involving the ALK locus in IMR32 cell line. A, scheme of the potential structure of a complex amplicon involving several 2p regions including a portion of the ALK gene in the IMR32 cell line. The upper part represents the normal positions of the amplified regions, whereas the lower part deciphers the organization of these regions within the amplicon. The arrows indicate the orientation of each fragment, from the lowest to the highest genomic position. Gene names or cytobands related to amplified regions are indicated. B, FISH experiments confirmed coamplification of part of the ALK locus and five distinct 2p regions in the same hsr.

Close modal

We applied the same approach for the primary tumor NB838 and the breakpoint revealed by array-CGH was precisely mapped at position 29,778,881 in the second intron of the ALK gene (Supplementary Table S2). The ALK gene was fused to repeated sequences precluding the identification of the exact partner involved in the rearrangement. A closer examination of the capture and paired-end data revealed other rearrangements within the ALK gene with various 2p amplicons, indicating a highly rearranged ALK locus in this tumor (data not shown).

Finally, for the CLB-Bar cell line, the same strategy revealed a more complex pattern of rearrangements targeting the ALK gene than first expected from the ultrahigh density array. It indeed confirmed the 2 breakpoints initially detected by array-CGH but also revealed additional breakpoints within the ALK locus (Supplementary Table S2). A more detailed examination of this array was consistent with these results, and other 2p amplified regions detected by SNP 6.0 array appeared to be involved in some rearrangements with the ALK locus (Supplementary Table S2 and Supplementary Fig. S3). Importantly, the rearrangement between positions 29,328,032 (intron 11) and 29,707,160 (intron 3) juxtaposes the region encoding exons 1 to 3 to the one encoding exons 12 to 29 of the ALK gene (Fig. 3A). This rearrangement was validated by PCR and complementary experiments revealed that a variant ALK protein is produced from this genomic rearrangement (vide infra). FISH experiments confirmed that the 2 ALK pieces involved in the rearrangement, A1 and A4, were coamplified in a common population of dmin (Fig. 3B). These dmin also contained the 2p amplicons located at around 9, 14, 20, and 22 millions, as well as the A2 and A3 regions of the ALK gene. The MYCN locus was amplified in a distinct population of dmin (Supplementary Fig. S3).

Figure 3.

ALK amplification and rearrangement in the CLB-Bar cell line and tumors at diagnosis and relapse. A, in the CLB-Bar cell line, regions A1 to A4 show amplification and are separated by nonamplified regions (upper part). An intragenic rearrangement juxtaposes the region encoding exons 1 to 3 (A4 region) to the one encoding exons 12 to 29 (A1 region; bottom). B, FISH analysis shows that the 2 parts of the ALK locus, A1 (RP11-1116M6, red) and A4 (RP11-453M4, green), are juxtaposed in a same population of dmin in the CLB-Bar cell line. C, FISH analysis with MYCN (RP11-463P22, red) and ALK (RP11-328L16, green) BAC probes on tumor appositions. Left, tumor at diagnosis; right, tumor at relapse.

Figure 3.

ALK amplification and rearrangement in the CLB-Bar cell line and tumors at diagnosis and relapse. A, in the CLB-Bar cell line, regions A1 to A4 show amplification and are separated by nonamplified regions (upper part). An intragenic rearrangement juxtaposes the region encoding exons 1 to 3 (A4 region) to the one encoding exons 12 to 29 (A1 region; bottom). B, FISH analysis shows that the 2 parts of the ALK locus, A1 (RP11-1116M6, red) and A4 (RP11-453M4, green), are juxtaposed in a same population of dmin in the CLB-Bar cell line. C, FISH analysis with MYCN (RP11-463P22, red) and ALK (RP11-328L16, green) BAC probes on tumor appositions. Left, tumor at diagnosis; right, tumor at relapse.

Close modal

The rearranged ALK locus is only observed at relapse

As the CLB-Bar cell line was derived from a NB tumor at relapse, we sought to determine when ALK amplification and rearrangement occurred. First, FISH analysis with MYCN and ALK probes on tumor appositions revealed a great heterogeneity in the amplification of both loci in the tumor at diagnosis (Fig. 3C, left). At relapse, heterogeneity was significantly reduced and most of the nuclei exhibited amplification of MYCN and ALK (Fig. 3C, right). Second, PCR analysis indicated that the rearranged ALK locus was present in the tumor at relapse, and is therefore not an event acquired during the culture process. However, this rearrangement was not detected in DNA extracted from the tumor at diagnosis (Fig. 4A). SNP 6.0 array analysis revealed that the pattern of amplicons on chromosome 2p in the CLB-Bar cell line is similar to the profile of the tumor at relapse but different from the profile of the tumor at diagnosis (Fig. 4B). Strikingly, the ALK locus was already amplified in the tumor at diagnosis (Fig. 4B and C) but, in agreement with data obtained by PCR, it did not appear rearranged in the tumor at diagnosis (Fig. 4C). Altogether, these results indicate that the ALK gene was amplified at diagnosis in a subset of tumor cells and that the rearranged allele was observed at the relapse stage only, suggesting that it may be implicated in the aggressiveness of the tumor.

Figure 4.

Molecular analysis of the CLB-Bar cell line and tumor DNA at diagnosis and relapse. A, PCR analysis confirmed that the rearranged ALK allele is present in the genomic DNA of the CLB-Bar cell line and the tumor at relapse but not in that of the tumor at diagnosis. We used primers that amplify part of the PPAPDC1A gene on chromosome 10 as a positive control. B, SNP 6.0 array genomic profiles showing ALK amplification and various 2p amplified regions in the CLB-Bar cell line, the tumor at relapse and the tumor at diagnosis. C, enlargement of the ALK locus for the three samples; the rearranged ALK allele is detected in the cell line and tumor at relapse by the absence of amplification of a 39 kb region (arrows).

Figure 4.

Molecular analysis of the CLB-Bar cell line and tumor DNA at diagnosis and relapse. A, PCR analysis confirmed that the rearranged ALK allele is present in the genomic DNA of the CLB-Bar cell line and the tumor at relapse but not in that of the tumor at diagnosis. We used primers that amplify part of the PPAPDC1A gene on chromosome 10 as a positive control. B, SNP 6.0 array genomic profiles showing ALK amplification and various 2p amplified regions in the CLB-Bar cell line, the tumor at relapse and the tumor at diagnosis. C, enlargement of the ALK locus for the three samples; the rearranged ALK allele is detected in the cell line and tumor at relapse by the absence of amplification of a 39 kb region (arrows).

Close modal

Characterization of the ALKΔ4–11 protein variant in the CLB-Bar cell line

We further showed that the amplified and rearranged ALK allele in the CLB-Bar cell line is expressed at the protein level. Indeed, Western blot analysis revealed that a variant protein of 170 kDa was the main ALK form in this cell line (Fig. 5A). Mass spectrometry analysis determined that a part of the extracellular domain was missing in this 170 kDa form (Fig. 5B). A peptide overlapping the junction of the ALK variant between amino acids 317 and 681 was detected in the proteomic analysis confirming that the variant lacks the amino acids encoded by exons 4 to 11 (Fig. 5C). We, therefore, named this variant, ALKΔ4–11. The missing exons encode part of the first and the complete second Meprin, A-5, receptor protein-tyrosine phosphatase mu (MAM) domains as well as the potential ligand-binding site (391–401) and the LDLa domain (453–471; refs. 4, 5). Interestingly, we observed that the ALKΔ4–11 variant was highly phosphorylated (Fig. 5A).

Figure 5.

Identification of the ALKΔ4–11 protein variant in the CLB-Bar cell line. A, in addition to bands at 220 and 140 kDa, a major band at 170 kDa (arrow) was detected by Western blot analysis using an ALK antibody directed against the kinase domain of the receptor. The various forms are also observed after immunoprecipitation (IP) with ALK antibodies. The 170 kDa variant strongly reacts with an antiphospho-ALK (P-ALK, Tyr 1604). WCL, whole cell lysate. B, proteomic analysis and quantification of peptide areas ratio of the 170 kDa/220 kDa forms revealed an underrepresented amino acid region in the extracellular domain of the 170 kDa form (peptides 8–14 in black, amino acids 318–672). Position of the 23 quantified peptides along the ALK receptor is shown below the histogram. C, mass spectrometry analysis identified 1 peptide overlapping the junction between amino acids 317 (R, last amino acid encoded by exon 3) and 681 (V, first amino acid encoded by exon 12).

Figure 5.

Identification of the ALKΔ4–11 protein variant in the CLB-Bar cell line. A, in addition to bands at 220 and 140 kDa, a major band at 170 kDa (arrow) was detected by Western blot analysis using an ALK antibody directed against the kinase domain of the receptor. The various forms are also observed after immunoprecipitation (IP) with ALK antibodies. The 170 kDa variant strongly reacts with an antiphospho-ALK (P-ALK, Tyr 1604). WCL, whole cell lysate. B, proteomic analysis and quantification of peptide areas ratio of the 170 kDa/220 kDa forms revealed an underrepresented amino acid region in the extracellular domain of the 170 kDa form (peptides 8–14 in black, amino acids 318–672). Position of the 23 quantified peptides along the ALK receptor is shown below the histogram. C, mass spectrometry analysis identified 1 peptide overlapping the junction between amino acids 317 (R, last amino acid encoded by exon 3) and 681 (V, first amino acid encoded by exon 12).

Close modal

Oncogenic properties and impaired maturation of ALKΔ4–11

To further characterize the ALKΔ4–11 variant, it was stably expressed in NIH3T3 cells (Fig. 6A). As compared with the WT receptor, this variant exhibited a much stronger in vitro kinase activity (Fig. 6B). Moreover, expression of the ALKΔ4–11 variant favored NIH3T3 cell proliferation compared with the WT ALK receptor (Fig. 6C) and led to a higher number of colonies in a soft agar assay (Fig. 6D and E). Finally, subcutaneous injection in nude mice showed that cells expressing this variant induced tumors with a higher frequency (5/5) compared with cells expressing the WT receptor (2/5).

Figure 6.

Oncogenic properties of ALKΔ4–11. A, Western blot analysis of NIH3T3 cells stably expressing ALK WT, ALKΔ4—11, or ALK R1275Q. B, increased kinase activity of the ALKΔ4–11 variant. 32P incorporated in ALK immunoprecipitates is visualized by autoradiography (top) before immunoblotting with an anti-ALK antibody (bottom). C, NIH3T3 cells expressing ALKΔ4–11 showed increased proliferation compared with those expressing ALK WT. The number of living cells is reported at various time points. D, increased colony formation in soft agar assay of NIH3T3 cells expressing ALKΔ4–11 as compared with cells expressing ALK WT. E, colony number quantification of soft agar experiments conducted in triplicates.

Figure 6.

Oncogenic properties of ALKΔ4–11. A, Western blot analysis of NIH3T3 cells stably expressing ALK WT, ALKΔ4—11, or ALK R1275Q. B, increased kinase activity of the ALKΔ4–11 variant. 32P incorporated in ALK immunoprecipitates is visualized by autoradiography (top) before immunoblotting with an anti-ALK antibody (bottom). C, NIH3T3 cells expressing ALKΔ4–11 showed increased proliferation compared with those expressing ALK WT. The number of living cells is reported at various time points. D, increased colony formation in soft agar assay of NIH3T3 cells expressing ALKΔ4–11 as compared with cells expressing ALK WT. E, colony number quantification of soft agar experiments conducted in triplicates.

Close modal

We also analyzed the glycosylation status and the subcellular localization of the ALKΔ4–11 receptor. Figure 7A shows that the ALKΔ4–11 variant was present as a major band sensitive to endoglycosidase H treatment, as it is the case for the lower band of the 220 kDa doublet of WT ALK. The various forms of WT ALK and the ALKΔ4–11 variant were sensitive to endoglycosidase F. These data provide evidence that the ALKΔ4–11 variant corresponds to a partially glycosylated form of ALK likely retained in the endoplasmic reticulum. Biotinylation experiments confirmed that the ALKΔ4–11 variant was indeed intracellular, as it was revealed by ALK antibodies but not by biotin, as the lower band of the 220 kDa doublet of WT ALK (Fig. 7B). The analysis of a subset of potential downstream targets of the ALK receptor did not reveal any significant differences between the ALKΔ4–11 form and the WT receptor (Fig. 7C).

Figure 7.

Impaired maturation of ALKΔ4–11 and analysis of downstream effectors. A, cell lysates from NIH3T3 cells expressing ALK WT or ALKΔ4-11 were untreated (C), treated with endoglycosidase F (F) or H (H), then analyzed by Western blot analysis using polyclonal anti-ALK. ALKΔ4–11 is endoglycosidase H sensitive, indicating an impaired protein maturation. B, biotinylation experiment showing total ALK protein detected with polyclonal anti-ALK in red and biotinylated proteins detected with streptavidin in green. The ALKΔ4–11 variant is not stained by biotin, showing its intracellular retention. C, cell extracts from NIH3T3 cells expressing ALK WT or ALKΔ4–11 were analyzed for pERK, pAKT, and pSTAT3.

Figure 7.

Impaired maturation of ALKΔ4–11 and analysis of downstream effectors. A, cell lysates from NIH3T3 cells expressing ALK WT or ALKΔ4-11 were untreated (C), treated with endoglycosidase F (F) or H (H), then analyzed by Western blot analysis using polyclonal anti-ALK. ALKΔ4–11 is endoglycosidase H sensitive, indicating an impaired protein maturation. B, biotinylation experiment showing total ALK protein detected with polyclonal anti-ALK in red and biotinylated proteins detected with streptavidin in green. The ALKΔ4–11 variant is not stained by biotin, showing its intracellular retention. C, cell extracts from NIH3T3 cells expressing ALK WT or ALKΔ4–11 were analyzed for pERK, pAKT, and pSTAT3.

Close modal

The present study reports that the ALK locus is targeted by rearrangements at a high frequency (23%) in NB cell lines and that rearrangements are also observed in a subset of NB tumors. ALK amplification and some intragenic rearrangements have been previously reported in NB but not characterized in details (6, 10, 11, 23, 24). Our results show that, in contrast to ALCL, IMT, or NSCLC, neither the breakpoints observed within the ALK locus nor the involved partners were recurrent in the studied NB samples. It is noteworthy to mention that the ALK gene is highly expressed in most NB, which may render the locus especially prone to rearrangements. Such a link between transcriptional regulation and genesis of genomic aberrations has recently been reported in prostate tumors presenting with a TMPRSS2-ERG fusion, in which rearrangements breakpoints were enriched near open chromatin (25).

This work identified 9 samples, including 7 cell lines and 2 primary tumors, in which the ALK gene is targeted by rearrangements. In 3 cases, we could document that abnormal transcripts are expressed from these rearrangements. In the N206 sample, we identified a chimeric transcript between ALK and CLPTM1L that was not in frame and the biologic function of which remains to be defined. We observed a highly rearranged ALK locus in the CLB-Re sample. Recently, a whole-genome mate-pair analysis of this sample revealed an extensive number of rearrangements between 2 regions of chromosomes 2p and 3p (Boeva, Jouannet and colleagues, submitted for publication). This pattern of rearrangements potentially corresponds to the phenomenon of chromothripsis (26) that may provide an explanation for the multiple breakpoints observed in the ALK gene. RNA-Seq analysis using a paired-end strategy confirmed the chimera predicted from our capture data between exon 11 of ALK and exon 2 of FHIT (data not shown). Subsequent RT-PCR indicated that the fusion was not in frame. It remains to be determined whether the truncated ALK formed from this rearrangement harbors biologic activity.

In the CLB-Bar cell line, the genomic rearrangement targeting the ALK gene leads to the expression of a novel variant form of the ALK receptor. We have previously shown that this variant was highly phosphorylated as revealed by an antiphospho Y1586 ALK antibody (10). Using NIH3T3 cells, we have now documented an increased kinase activity of this variant compared with the WT receptor as well as its impaired maturation and intracellular retention. We also showed that the ALKΔ4–11 form exhibits oncogenic and tumorigenic properties. Such properties are similar to the ones reported for another short ALK form, ALKdel2–3, also truncated in its extracellular domain (27). Whereas Okubo and colleagues observed an increased phosphorylation of STAT3 in cells expressing the ALKdel2–3 form, expression of the ALKΔ4–11 variant was not associated with increased phospho-STAT3. This suggests that further characterization of signaling pathways triggered by the ALKΔ4–11 variant will therefore be required. Our results indicate that although rare, genomic rearrangements associated with amplification may lead to ALK activation in a subset of NB samples through deletions of part of the extracellular domain of the receptor. These observations are reminiscent of the case of 2 others RTK, the PDGFRAΔ8,9 (PDGR receptor α) and the delta epidermal growth factor receptor (ΔEGFR)/EGFRvIII for which in-frame deletion mutants of the extracellular domain have been described as potent oncogenes in various human cancers (28–31). These variant receptors have notably been shown to be activated in a ligand-independent manner, which could be the case for the ALKΔ4–11 as it lacks the putative ligand-binding site and allows proliferation of NIH3T3 cells under low serum conditions.

The CLB-Bar cell line has been derived from a stage IV tumor at relapse and both samples harbor the same rearranged and amplified ALK locus, showing that this event was not linked to the in vitro culture. Strikingly, in the tumor at diagnosis, we documented an amplification of the whole ALK locus, but we could not detect the rearrangement. Our model proposes that ALK amplification occurred early in the oncogenic process. Importantly, FISH analysis documented a great heterogeneity in the amplification of the ALK locus in the tumor at diagnosis. We can, therefore, speculate that the ALK rearrangement occurred in a cell without amplification and that this allele was further amplified. Cells exhibiting this rearranged and amplified allele were likely selected during tumor progression and account for most, if not the entire, tumor at relapse. Thus, 2 independent amplification events may be suggested, an early one involving the WT ALK and a later one, involving the ALKΔ4-11 allele. Altogether, our results suggest that amplification and rearrangement would cooperate to activate the ALK receptor, resulting in an increased aggressiveness of the tumor.

In conclusion, we document a novel ALK truncated form bearing oncogenic properties and associated with tumor aggressiveness. Our data suggest that it may be useful to investigate ALK amplification and rearrangement at the time of relapse in NB patients with refractory tumors using high-resolution SNP profiling or RT-PCR experiments. Patients presenting with tumors exhibiting ALK amplification and/or deletion of the extracellular domain may potentially benefit from ALK-targeted therapy.

No potential conflicts of interest were disclosed.

Conception and design: O. Delattre, I. Janoueix-Lerosey

Development of methodology: G. Schleiermacher, S. Jouannet

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A. Cazes, C. Louis-Brennetot, P. Mazot, F. Dingli, B. Lombard, J. Cappo, V. Combaret, G. Schleiermacher, S. Ferrand, G. Pierron, M. Vigny

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Cazes, C. Louis-Brennetot, V. Boeva, J. Cappo, S. Ferrand, G. Pierron, E. Barillot, D. Loew, O. Delattre, I. Janoueix-Lerosey

Writing, review, and/or revision of the manuscript: A. Cazes, C. Louis-Brennetot, G. Schleiermacher, G. Pierron, O. Delattre, I. Janoueix-Lerosey

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): P. Mazot, R. Daveau, V. Combaret, G. Schleiermacher, O. Delattre

Study supervision: I. Janoueix-Lerosey

The authors are grateful to Marie-Claude Boutterin, Sarah Cohen-Gogo, Virginie Perrin, and Didier Surdez for their critical help and technical support. We thank the PICT-BDD of Institut Curie for imaging facilities and D. Gentien and his team from the translational department of Institut Curie for SNP 6.0 array experiments.

The U830 Inserm laboratory is supported by grants from the Institut National du Cancer, the Ligue Nationale contre le Cancer (Equipe labellisée), the Association Hubert Gouin, Les Bagouz à Manon, les amis de Claire and Enfance et Santé. The Laboratory of Proteomic Mass Spectrometry is supported by the Cancéropôle Ile-de-France and the Institut National du Cancer. A. Cazes is the recipient of a fellowship of the Cancéropôle Ile-de-France. The U900 Inserm laboratory is supported by the Ligue Nationale Contre le Cancer (Equipe labellisée).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Brodeur
GM
. 
Neuroblastoma: biological insights into a clinical enigma
.
Nat Rev Cancer
2003
;
3
:
203
16
.
2.
Janoueix-Lerosey
I
,
Schleiermacher
G
,
Delattre
O
. 
Molecular pathogenesis of peripheral neuroblastic tumors
.
Oncogene
2010
;
29
:
1566
79
.
3.
Maris
JM
. 
Recent advances in neuroblastoma
.
N Engl J Med
2010
;
362
:
2202
11
.
4.
Chiarle
R
,
Voena
C
,
Ambrogio
C
,
Piva
R
,
Inghirami
G
. 
The anaplastic lymphoma kinase in the pathogenesis of cancer
.
Nat Rev Cancer
2008
;
8
:
11
23
.
5.
Palmer
RH
,
Vernersson
E
,
Grabbe
C
,
Hallberg
B
. 
Anaplastic lymphoma kinase: signalling in development and disease
.
Biochem J
2009
;
420
:
345
61
.
6.
Caren
H
,
Abel
F
,
Kogner
P
,
Martinsson
T
. 
High incidence of DNA mutations and gene amplifications of the ALK gene in advanced sporadic neuroblastoma tumours
.
Biochem J
2008
;
416
:
153
9
.
7.
Chen
Y
,
Takita
J
,
Choi
YL
,
Kato
M
,
Ohira
M
,
Sanada
M
, et al
Oncogenic mutations of ALK kinase in neuroblastoma
.
Nature
2008
;
455
:
971
4
.
8.
de Pontual
L
,
Kettaneh
D
,
Gordon
CT
,
Oufadem
M
,
Boddaert
N
,
Lees
M
, et al
Germline gain-of-function mutations of ALK disrupt central nervous system development
.
Human Mutation
2011
;
32
:
272
6
.
9.
George
RE
,
Sanda
T
,
Hanna
M
,
Frohling
S
,
Luther
W
 II
,
Zhang
J
, et al
Activating mutations in ALK provide a therapeutic target in neuroblastoma
.
Nature
2008
;
455
:
975
8
.
10.
Janoueix-Lerosey
I
,
Lequin
D
,
Brugieres
L
,
Ribeiro
A
,
de Pontual
L
,
Combaret
V
, et al
Somatic and germline activating mutations of the ALK kinase receptor in neuroblastoma
.
Nature
2008
;
455
:
967
70
.
11.
Mosse
YP
,
Laudenslager
M
,
Longo
L
,
Cole
KA
,
Wood
A
,
Attiyeh
EF
, et al
Identification of ALK as a major familial neuroblastoma predisposition gene
.
Nature
2008
;
455
:
930
5
.
12.
De Brouwer
S
,
De Preter
K
,
Kumps
C
,
Zabrocki
P
,
Porcu
M
,
Westerhout
EM
, et al
Meta-analysis of neuroblastomas reveals a skewed ALK mutation spectrum in tumors with MYCN amplification
.
Clin Cancer Res
2010
;
16
:
4353
62
.
13.
Passoni
L
,
Longo
L
,
Collini
P
,
Coluccia
AM
,
Bozzi
F
,
Podda
M
, et al
Mutation-independent anaplastic lymphoma kinase overexpression in poor prognosis neuroblastoma patients
.
Cancer Res
2009
;
69
:
7338
46
.
14.
Seeger
RC
,
Brodeur
GM
,
Sather
H
,
Dalton
A
,
Siegel
SE
,
Wong
KY
, et al
Association of multiple copies of the N-myc oncogene with rapid progression of neuroblastomas
.
N Engl J Med
1985
;
313
:
1111
6
.
15.
Janoueix-Lerosey
I
,
Schleiermacher
G
,
Michels
E
,
Mosseri
V
,
Ribeiro
A
,
Lequin
D
, et al
Overall genomic pattern is a predictor of outcome in neuroblastoma
.
J Clin Oncol
2009
;
27
:
1026
33
.
16.
Webb
TR
,
Slavish
J
,
George
RE
,
Look
AT
,
Xue
L
,
Jiang
Q
, et al
Anaplastic lymphoma kinase: role in cancer pathogenesis and small-molecule inhibitor development for therapy
.
Expert Rev Anticancer Ther
2009
;
9
:
331
56
.
17.
Schleiermacher
G
,
Bourdeaut
F
,
Combaret
V
,
Picrron
G
,
Raynal
V
,
Aurias
A
, et al
Stepwise occurrence of a complex unbalanced translocation in neuroblastoma leading to insertion of a telomere sequence and late chromosome 17q gain
.
Oncogene
2005
;
24
:
3377
84
.
18.
Schleiermacher
G
,
Janoueix-Lerosey
I
,
Combaret
V
,
Derre
J
,
Couturier
J
,
Aurias
A
, et al
Combined 24-color karyotyping and comparative genomic hybridization analysis indicates predominant rearrangements of early replicating chromosome regions in neuroblastoma
.
Cancer Genet Cytogenet
2003
;
141
:
32
42
.
19.
Schleiermacher
G
,
Raynal
V
,
Janoueix-Lerosey
I
,
Combaret
V
,
Aurias
A
,
Delattre
O
. 
Variety and complexity of chromosome 17 translocations in neuroblastoma
.
Genes Chromosomes Cancer
2004
;
39
:
143
50
.
20.
Moog-Lutz
C
,
Degoutin
J
,
Gouzi
JY
,
Frobert
Y
,
Brunet-de Carvalho
N
,
Bureau
J
, et al
Activation and inhibition of anaplastic lymphoma kinase receptor tyrosine kinase by monoclonal antibodies and absence of agonist activity of pleiotrophin
.
J Biol Chem
2005
;
280
:
26039
48
.
21.
Mazot
P
,
Cazes
A
,
Boutterin
MC
,
Figueiredo
A
,
Raynal
V
,
Combaret
V
, et al
The constitutive activity of the ALK mutated at positions F1174 or R1275 impairs receptor trafficking
.
Oncogene
2011
;
30
:
2017
25
.
22.
McDermott
U
,
Iafrate
AJ
,
Gray
NS
,
Shioda
T
,
Classon
M
,
Maheswaran
S
, et al
Genomic alterations of anaplastic lymphoma kinase may sensitize tumors to anaplastic lymphoma kinase inhibitors
.
Cancer Res
2008
;
68
:
3389
95
.
23.
Caren
H
,
Erichsen
J
,
Olsson
L
,
Enerback
C
,
Sjoberg
RM
,
Abrahamsson
J
, et al
High-resolution array copy number analyses for detection of deletion, gain, amplification and copy-neutral LOH in primary neuroblastoma tumors: four cases of homozygous deletions of the CDKN2A gene
.
BMC Genomics
2008
;
9
:
353
.
24.
Fix
A
,
Lucchesi
C
,
Ribeiro
A
,
Lequin
D
,
Pierron
G
,
Schleiermacher
G
, et al
Characterization of amplicons in neuroblastoma: high-resolution mapping using DNA microarrays, relationship with outcome, and identification of overexpressed genes
.
Genes Chromosomes Cancer
2008
;
47
:
819
34
.
25.
Berger
MF
,
Lawrence
MS
,
Demichelis
F
,
Drier
Y
,
Cibulskis
K
,
Sivachenko
AY
, et al
The genomic complexity of primary human prostate cancer
.
Nature
2011
;
470
:
214
20
.
26.
Stephens
PJ
,
Greenman
CD
,
Fu
B
,
Yang
F
,
Bignell
GR
,
Mudie
LJ
, et al
Massive genomic rearrangement acquired in a single catastrophic event during cancer development
.
Cell
2011
;
144
:
27
40
.
27.
Okubo
J
,
Takita
J
,
Chen
Y
,
Oki
K
,
Nishimura
R
,
Kato
M
, et al
Aberrant activation of ALK kinase by a novel truncated form ALK protein in neuroblastoma
.
Oncogene
2012
Jan 16.
[Epub ahead of print]
.
28.
Clarke
ID
,
Dirks
PB
. 
A human brain tumor-derived PDGFR-alpha deletion mutant is transforming
.
Oncogene
2003
;
22
:
722
33
.
29.
Ekstrand
AJ
,
Sugawa
N
,
James
CD
,
Collins
VP
. 
Amplified and rearranged epidermal growth factor receptor genes in human glioblastomas reveal deletions of sequences encoding portions of the N- and/or C-terminal tails
.
Proc Natl Acad Sci U S A
1992
;
89
:
4309
13
.
30.
Moscatello
DK
,
Holgado-Madruga
M
,
Godwin
AK
,
Ramirez
G
,
Gunn
G
,
Zoltick
PW
, et al
Frequent expression of a mutant epidermal growth factor receptor in multiple human tumors
.
Cancer Res
1995
;
55
:
5536
9
.
31.
Ozawa
T
,
Brennan
CW
,
Wang
L
,
Squatrito
M
,
Sasayama
T
,
Nakada
M
, et al
PDGFRA gene rearrangements are frequent genetic events in PDGFRA-amplified glioblastomas
.
Genes Dev
2010
;
24
:
2205
18
.