Background
The adenocarcinoma is the most common form of lung cancer worldwide, different subsets having specific genetic backgrounds of great importance for molecular-targeted therapy. For example, somatic mutations of the epidermal growth factor receptor (
EGFR) are especially prevalent in adenocarcinomas among never smokers, females, and those with Asian ethnicity [
1]. On the other hand,
KRAS mutations are associated with the smoking habit [
2] and the two tend to be mutually exclusive. Recently, Soda et al. found a novel fusion gene,
EML4-ALK, arising from an inversion on the short arm of chromosome 2 in non-small cell lung carcinomas [
3].
ALK fusion is a unique example of tyrosine kinase activation by structural chromosome rearrangement [
4].
EML4-ALK fusion is a powerful driving molecular event by itself. The chimeric protein permits ligand-independent dimerization and constitutive activation of
ALK, resulting in dominant oncogenic activity. Multiple fusion variants of
EML4-ALK and notable clinicopathological characteristics of fusion positive tumours have been revealed [
5‐
9]. Since the tyrosine kinase is involved and activated by gene fusion, this type of malignancy has emerged as a target for anti-tyrosine kinase therapy [
4,
10‐
12].
We have revealed that
ALK fusion-positive tumours constituted a particular subset in lung adenocarcinomas in terms of clinical characteristics, histology and etiology, as well as molecular changes [
7,
8]. It is of great interest to assess global genomic alterations to provide deep insight into their genesis, especially considering these tumours arise in non- or light smokers. Single nucleotide polymorphism (SNP) microarray analysis enables precise high-throughput detection of genomic copy number alterations, gains and losses in the genome contributing to carcinogenesis [
13] with gene expression varying consistently with DNA copy number changes [
14,
15]. We therefore conducted of the present genomic profiling of lung adenocarcinomas with and without
ALK fusion.
Discussion
Recurrent chromosome translocation has been accepted to play an important role in the pathogenesis of hematological malignancies, but not of solid tumours. Recently, however, chromosome rearrangements in solid tumours such as prostate cancer and non-small cell lung cancer have been reported [
32].
ALK fusion was originally described in anaplastic large-cell lymphoma as a chimeric protein
NPM-ALK resulting from a translocation. More recently, evidence has accumulated that the
EML4-ALK fusion gene defines a novel subclass of lung adenocarcinomas with distinct clinicopathological features [
7‐
9], so that it has emerged as a target for therapy. We focused here for the first time on allelic imbalance of tumours with
ALK fusion with a novel technique which has already shown the involvement of loss of A20 function in the pathogenesis of a subset of B-cell lymphomas [
33] and gain of function of C-CBL tumour suppressor in myeloid neoplasms [
34]. Applying this methodology, we demonstrated that lung adenocarcnomas with
ALK fusion feature less amplification of loci with oncogenes and fewer deletions of loci related to tumour suppressor genes, although global chromosome aberrations were similar between tumours with and without
ALK fusion. suggesting that the fusion gene is a driver mutation, not just a passenger mutation.
Genetic instability was here categorized into two groups for simplicity, at the chromosomal level and at the nucleotide level. We earlier found the former to play a more important role in lung carcinogenesis, the frequency of LOH (loss of heterozygosity) being higher in less-differentiated tumours [
35].
ALK fusion positive tumours are more common among non-smokers and the younger population, similar to those with
EGFR mutations. We had expected fewer chromosome aberrations in
ALK fusion-positive tumours because tumours arising in such people usually harbor less LOH and a lower
TP53 mutation rate than smokers [
36‐
38]. Contrary to our expectation, the global copy number changes at the chromosomal arm level did not differ between the two groups, although significant differences of alteration frequency at the individual chromosomal arms were seen. In addition, only
ALK fusion-negative tumours showed an increase of the frequency of chromosome arm gain with the advancement of disease stage. Furthermore, at the smaller-genomic scale level,
ALK fusion-positive tumours were less amplified at the loci containing
EGFR family genes, 7p11.2 (
EGFR), 17q12 (
ERBB2) and other loci, 1p34.3 (
MYCL), 7p21.1, 8q24.21 (
MYC), 16p13.3 and 17q25.1.
EGFR and
ERBB2 play important roles by dimerizing when their ligands binds to produce downward growth signals to the tumour cells. Mutations and activation of these genes may drive carcinogenesis [
39], and increased expression is associated with a poor prognosis in NSCLCs [
40‐
43].
ALK fusion positive tumours are speculated to be less dependent on the actions of oncogenes and tumour-suppressor genes induced by copy number changes. Our results may also indicate that there is common and frequent chromosome abnormality in lung adenocarcinomas independent of
ALK fusion, such as the 5p15.33 region, including
TERT.
As for genomic loss, 9p21.3 (
CDKN2A), 9p23-p24.1 (
PTPRD) and 13q14.2 (
RB1) were significantly less frequently deleted in
ALK fusion-positive tumours. Homozygous deletion was seen only at 9p21.3 including
CDKN2A and limited to
EGFR-mutated tumours among
ALK fusion-negative neoplasms as reported in the literature [
44] and also seen in
ALK-fusion positive ones
. That deletion of 9p23-24.1 and 13q14.2 including tumour suppressor genes was rare in
ALK fusion-positive tumours suggests that they can grow even if the functions of these suppressor genes are retained.
Of all the selected loci, 5p15.33 containing
TERT (telomerase reverse transcriptase isoform 2) showed the highest frequency of recurring gain regardless of
ALK fusion. The enzyme is important for telomere regeneration and maintenance resulting in a growth advantage and Zhang et al. reported that the locus is a frequent target of amplification during tumourigenesis [
45]. Copy number gain of this locus significantly correlates with telomerase activity [
46] and is one of the most consistent alterations in the early stages of non-small cell lung cancer [
47]. In addition, increased susceptibility to lung cancer development associated with a SNP polymorphism of this locus has been reported [
48,
49]. The fact that most human tumour cells have telomerase activity indicates that its acquisition is vital for carcinogenesis and cell immortalization, and it might explain the reason why lung adenocarcinomas with or without
ALK fusion shows similar frequency of copy number gain of this locus.
Our results have some therapeutic relevance. The fact that there are less involvement of other oncogenes and tumor suppressor genes may be related to dramatic responses to targeted drugs because of intact cellular processes including apoptosis pathways. In this regard, there is an interesting paper by Camidge et al. [
50], demonstrating the inverse relationship between fused and isolated red copy number on FISH might suggest the
ALK fusion positive tumor was a “near-diploid” subtype of non-small cell lung cancer. Comparing closely, however, between their and our results, our study clearly revealed the overall frequency of chromosome aberrations are similar between
ALK fusion positive and negative tumors, suggesting not “near-diploid”. But, certainly, we need more investigations on genomic instability of
ALK fusion positive tumors.
It is well known that smoking causes genomic changes with allelic imbalance [
20]. As shown in Table
1, smokers dominate never smokers in the group without fusion whereas the fusion-positive group has more never smokers than smokers. Since the tumors without
ALK fusion include
EGFR-mutated tumors, most of which are from never smokers, the
ALK fusion-negative group is certainly heterogeneous. In due course, a study that describes comparisons of allelotypes of non-smoker’s tumors between with
ALK fusion and with
EGFR mutation should be warranted.
Competing interest
The authors have no potential conflicts of interest.
Authors’ contributions
HN, MK, SO, HM and YI designed the study. HN, KT, KI, NM, HM and YI performed pathological and/or genomic diagnosis of tumors. HN, MK, MS and SO obtained microarray data and carried out bioinformatics analysis. HN and KN analyzed mutations. YS, SO and YI collected samples and/or provided detailed clinical data of patients. HN and YI drafted the manuscript. All authors read and approved the final manuscript.