Background Crizotinib, a dual anaplastic lymphoma kinase (ALK) and mesenchymal-epithelial transition (MET) tyrosine kinase inhibitor, is currently being evaluated for the treatment of neuroblastoma. Its effects are thought to be mediated mainly via its activity against ALK. Although MET genomic/protein expression status might conceivably affect crizotinib efficacy, this issue has hitherto not received attention in neuroblastomas.
Aims/Methods MET genomic and protein expression status was characterised by silver in situ hybridisation and immunohistochemistry (IHC) respectively, in a cohort of 54 neuroblastoma samples. MET splice isoforms were characterised in 15 of these samples by quantitative PCR.
Results One case (1/54; prevalence 1.85%) displayed MET genomic amplification, while another case (1/54; prevalence 1.85%) displayed strong membranous MET protein expression (IHC score 3+). Alternative exon 10-deleted and exon 14-deleted MET splice isoforms were identified.
Conclusions MET amplification and protein expression, although low in prevalence, are present in neuroblastomas. This has implications when crizotinib is employed as a therapeutic agent in neuroblastomas. Additionally, the existence of alternatively spliced MET isoforms may have clinical and biological implications in neuroblastomas.
- Molecular Pathology
- Paediatric Pathology
Statistics from Altmetric.com
Neuroblastoma represents the most common paediatric extracranial solid tumour.1 Although survival rates have improved over the past few decades, high-risk neuroblastoma patients still experience a <40% overall survival rate.1
Several novel targeted therapeutics are being evaluated for the treatment of neuroblastoma.2 Crizotinib, a dual anaplastic lymphoma kinase (ALK) and mesenchymal-epithelial transition (MET) tyrosine kinase inhibitor,3 has been employed in the clinical setting for the treatment of various tumours, including ALK-rearranged lung carcinomas,4 inflammatory myofibroblastic tumours5 and anaplastic large-cell lymphomas.6 Crizotinib also displays in vitro activity against neuroblastomas,7 and notably, a recent phase I trial reported two neuroblastoma patients with complete response to crizotinib therapy.8
The effect of crizotinib in neuroblastomas is believed to be mediated in large part by its activity against ALK, with ALK copy number and mutation status thought to be the important determinants of therapeutic response.7 ,9 However, because crizotinib also inhibits MET kinase, there is a possibility that the MET genomic/protein expression status might affect crizotinib efficacy, an issue which hitherto has not received any attention in neuroblastomas. As documented in other tumour types, such as oesophagogastric adenocarcinomas10 and a single case each of lung carcinoma11 and glioblastoma,12 MET amplification is associated with sensitivity to crizotinib.
MET, a tyrosine kinase, functions as the receptor for the hepatocyte growth factor (HGF). The HGF-MET axis promotes tumourigenesis by stimulating cell proliferation, survival, motility, invasion and angiogenesis.13 MET expression has been previously documented in human neuroblastoma tissues and cell lines.14–16 MET genomic status is not well characterised in neuroblastomas. The MET gene is located on chromosome 7q21–q31, and gain of chromosome 7q is observed in neuroblastomas,9 ,17 ,18 suggesting that MET copy number gain/amplification may be present. MET splice isoforms, known to exist in other tumour types,19–23 have not been characterised in neuroblastomas.
In this study, we characterised MET genomic status, protein expression and splice isoforms in human neuroblastoma samples by silver in situ hybridisation (SISH), immunohistochemistry (IHC) and quantitative PCR (qPCR) respectively, with a view towards understanding the potential role of MET in affecting crizotinib efficacy in neuroblastomas.
Materials and methods
Fifty-four surgically resected cases comprising 46 Schwannian stroma-poor neuroblastomas and eight ganglioneuroblastomas, were obtained from the archives of the Department of Pathology and Laboratory Medicine, KK Women's and Children's Hospital, Singapore, for incorporation into the tissue microarrays (TMAs). For this study, only Schwannian stroma-poor tumours and ganglioneuroblastomas were selected; ganglioneuromas were excluded. The 54 cases comprised 16 pretreatment, 35 post-treatment and three relapsed cases. Of the post-treatment primary neuroblastomas, eight had pretreatment small biopsies that were optimal for evaluation. Clinicopathological data including age at diagnosis, gender, tumour site, duration of follow-up, stage, metastatic sites, current status and MYCN genomic status, was obtained where available. (The MYCN genomic status was evaluated by fluorescence in situ hybridisation as detailed previously24).
Ethics approval was obtained from the Singhealth Centralised Institutional Review Board (CIRB Ref: 2012/450/B).
Tissue microarray construction
One tissue core (0.6 mm diameter) was punched from a Schwannian stroma-poor area from the donor tissue blocks and deposited into a recipient block using a manual tissue-arraying instrument (Beecher Instruments, Sun Prairie, Wisconsin, USA).
Dual-colour MET/Chr 7 silver in situ hybridisation
Sections of 4 µm thickness were stained using the Ventana BenchMark ULTRA (Ventana Medical Systems, Roche Diagnostics, Arizona, USA) using the INFORM MET DNA PROBE kit (Ventana Medical Systems). This kit was used in a previous study.25 The sections were deparaffinised, pretreated with CC2 Extended buffer (Ventana Medical Systems) at pH 6.0 at 86°C, followed by enzyme digestion of protein using ISH Protease 3 (Ventana Medical Systems) for 12 min. The dinitrophenyl (DNP)-labelled MET DNA probe (Ventana Medical Systems) was then added to the sections for hybridisation for 6 h. Bound MET DNA probe was detected using the ultraView Silver-ISH DNP detection kit (Ventana Medical Systems) and visualised as a black signal.
Subsequently, the DNP-labelled Chromosome 7 probe (Ventana Medical Systems) was added for hybridisation for 2 h. The DNP-labelled Chromosome 7 probe was detected using the ultraView Alkaline Phosphatase Red ISH detection kit (Ventana Medical Systems) and visualised as a red signal. Slides were then counterstained with Haematoxylin II (Ventana Medical Systems) and Bluing reagent (Ventana Medical Systems).
Twenty non-overlapping cell nuclei were scored for MET and Chr7 signals. MET amplification was defined as a MET/Chr7 ratio of greater than 2.2.10 As a positive control, MET-amplified xenograft tissue was obtained (Ventana Medical Systems).
IHC was performed on 4 µm-thick sections using the Ventana Benchmark ULTRA (Ventana Medical Systems) and ultraView detection kit (Ventana Medical Systems). The primary antibodies used were anti-MET rabbit monoclonal antibody (clone SP44; Ventana Medical Systems) and anti-ALK mouse monoclonal antibody (Dako Denmark A/S, Glostrup, Denmark).
The scoring criteria for MET IHC are similar to those described in another recent study of ovarian carcinomas26: negative, no discernible staining, or background-type staining; 1+, definite cytoplasmic staining, and/or equivocal discontinuous membranous staining; 2+, unequivocal membranous staining with mild to moderate intensity in at least 10% of tumour cells; 3+, strong and complete membranous staining in at least 10% of tumour cells.
RNA extraction and reverse transcription
Frozen tissue was available for 15 neuroblastoma samples for RT-qPCR analysis. The samples were homogenised in the presence of TRI-Reagent (Sigma, Madison, Wisconsin, USA) and total RNA extracted according to manufacturer's instruction. The integrity of isolated total RNA was validated by denaturing agarose gel electrophoresis, and the concentration measured by Nanodrop 2000 (Thermo Fisher Scientific, Wilmington, Delaware, USA). Two micrograms of total RNA were reverse transcribed using ImPromII reverse transcriptase (Promega) and 0.5 µg random hexamer for 60 min at 42°C according to manufacturer's instruction. The reaction was terminated by heating at 70°C for 5 min.
Design of primers and quantitative PCR
The nucleotide sequence and exon information of the longest transcript (wild type) of MET proto-oncogene (homo sapiens, accession number NM_001127500) was retrieved from NCBI Gene database http://www.ncbi.nlm.nih.gov/gene. AlleleID 7 (Premier Biosoft) was used to design primer sets (table 1) that amplify the consensus or isoform-specific regions (figure 1). All primer sets are exon spanning to avoid amplification from genomic sequences, and have been evaluated for possible false priming to known human sequences using AlleleID 7. Real-time qPCR using SYBR Green I was performed on BioRad CFX96 (Bio-Rad Laboratories, Hercules, CA) in a total volume of 20 µL in 1× XtensaMix-SGTM (BioWORKS, Singapore), containing 2.5 mM MgCl2, 200 nM of primers and 0.3 U of iTaq DNA polymerase (i-DNA Biotechnology, Singapore). Real-time qPCR was carried out after an initial denaturation for 3 min at 95°C followed by 40 cycles of 30 s denaturation at 95°C, 30 s annealing at 60°C and 30 s extension at 72°C. All real-time PCR quantification was carried out simultaneously with non-template controls, and melt curve analyses were performed at the end of reaction to verify the identity of the products. The threshold cycles (Ct) were calculated using the CFX manager v1.6 (Biorad). The expression levels of different MET isoforms were normalised to the expressions of Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in the same samples.
We identified one neuroblastoma case with MET amplification (MET/Chr7 ratio 2.42) on the TMA analysis, for which further evaluation on the full section showed a similar result (MET/Chr7 ratio 2.95) (figure 2A). This case did not display MET expression by IHC (figure 2B). Of note, this case was MYCN-amplified. The rest of the cases did not display MET amplification (MET/Chr7 ratios ranging from 0.76 to 1.70).
We also identified a separate case with MET expression by IHC (score 3+) that was MET non-amplified on TMA analysis. Further evaluation on the full section showed strong membranous staining in approximately 70% of the tumour cells, with isolated tumour cells displaying strong cytoplasmic expression (figure 2C,D). Of note, this case did not display MYCN amplification. The clinicopathological information of these two cases is shown in table 2.
The rest of the cases did not display any MET expression (score 0). The concordance between MET SISH and IHC is shown in table 3.
In addition to the two previous cases, full sections of seven more cases were evaluated for intratumoural heterogeneity. The results were similar to the TMA analyses. The matched pretreatment small biopsies showed results similar to the post-treatment surgical resections for both the SISH and IHC analyses. The primary tumours of two relapsed neuroblastomas were also analysed, and both showed results similar to the metastatic tumours (MET non-amplified, IHC 0). The primary tumour of the third relapsed case was not available for evaluation.
One case (MET non-amplified, IHC 0) showed strong ALK cytoplasmic expression (figure 3). This case was MYCN-amplified.
qPCR Analysis in the 15 samples revealed the presence of alternatively spliced exon 10-deleted and exon 14-deleted MET isoforms (figure 4A). The expression ratios relative to the wild-type isoform of the individual cases are shown in table 4. These 15 cases were all MET non-amplified, IHC 0. In general, exon 10-deleted and exon 14-deleted isoforms were more abundant than the wild-type isoforms (figure 4B).
The identification of MET amplification and expression in neuroblastomas, although low in prevalence, is important both from a clinical and biological perspective. MET amplification in oesophagogastric carcinomas,10 and a single case each of lung carcinoma11 and glioblastoma12 was associated with sensitivity to crizotinib therapy; MET amplification may similarly predict crizotinib sensitivity in neuroblastomas. Of note, although the prevalence of MET amplification in our neuroblastoma series is 1.85% (1/54), this prevalence is not too different to that reported in oesophagogastric carcinomas (2% as reported by Lennerz et al10).
To the best of our knowledge, MET expression has not been studied by IHC in a cohort of primary human neuroblastoma samples. The strong membranous expression observed in one of our cases again suggests dependency on MET signalling, and the potential for crizotinib therapy (or other MET-specific inhibitors) in this setting.
The absence of upregulated protein expression in the presence of genomic amplification, and vice versa, in our neuroblastoma cohort is not surprising, in view of the existing body of knowledge concerning MET. With regard to our MET non-amplified/IHC 3+ sample, it is known that MET expression may be upregulated in the absence of genetic aberrations.13 In various studies including gastric carcinoma,27 non-small cell lung carcinoma28 and ovarian clear-cell carcinoma,26 the percentage of amplified cases among MET-overexpressing cases ranged from 0 to approximately 50%. Similarly, MET expression may be negative despite the presence of amplification.26 ,28
The most immediate clinical implication of our findings is how the MET genomic/protein expression status might affect crizotinib therapy in neuroblastomas. Central to this question is the relationship between MET and ALK status. While we did not evaluate ALK copy number and mutational status in this study, ALK IHC revealed one positive case that was MET non-amplified/IHC 0. Given the rarity of MET amplification/expression and ALK expression, and our relatively small sample size, we were not able to undertake rigorous statistical analysis to probe the relationship between MET and ALK status. We believe, however, that our findings warrant an expanded study investigating this relationship, as well as whether crizotinib displays efficacy in ALK-negative, MET-amplified/expressing neuroblastomas. It is interesting to point out that both MET and ALK share related signalling pathways, for example, the MAPK-signalling pathway.15 ,29
The identification of alternatively spliced MET isoforms in neuroblastomas is important for several reasons. First, alternatively spliced isoforms may affect IHC results as relevant epitopes are modified as a result of splicing (eg, see refs 30, 31). In this study, we employed the anti-MET rabbit monoclonal antibody (clone SP44; Ventana Medical Systems), for which the exact epitope is unknown. We cannot rule out the possibility that the negative IHC scores are a result of these splice isoforms. Second, alternatively spliced isoforms may differ functionally from the wild-type isoforms. This is particularly true for the exon 14-deleted MET isoform, which displays attenuated degradation and prolonged, ligand-dependent cell signalling.20 Further work entails understanding the roles of these isoforms in neuroblastoma tumorigenesis.
How important is the HGF-MET axis as an oncogenic pathway in neuroblastomas? HGF promotes neuroblastoma invasion in vitro and in vivo,15 and MET silencing by siRNA results in decreased invasiveness of TrkB-overexpressing neuroblastoma cell lines.16 PHA665752, a MET-specific small molecular inhibitor,32 inhibits HGF-mediated migration and proliferation/cell survival in a MET-expressing neuroblastoma cell line.14 Together, these provide circumstantial evidence that MET participates in neuroblastoma oncogenesis.
In summary, MET amplification and protein expression, although low in prevalence, are present in neuroblastomas. This has implications for crizotinib therapy in neuroblastomas, and also raises the possibility that MET-specific targeted therapeutics might potentially be of value in MET-amplified/expressing neuroblastomas. Additionally, the existence of alternatively spliced MET isoforms may have clinical and biological implications in neuroblastomas.
Take home messages
Mesenchymal-epithelial transition (MET) genomic amplification and protein expression, although low in prevalence, are present in neuroblastomas. This has implications when crizotinib is employed as a therapeutic agent in neuroblastomas.
The presence of alternatively spliced MET isoforms in neuroblastomas has several implications, including the possibility that such isoforms might affect IHC results, as well as contribute to neuroblastoma tumorigenesis.
Interactive multiple choice questions
This JCP review article has an accompanying set of multiple choice questions (MCQs). To access the questions, click on BMJ Learning: Take this module on BMJ Learning from the content box at the top right and bottom left of the online article. For more information please go to: http://jcp.bmj.com/education. Please note: The MCQs are hosted on BMJ Learning—the best available learning website for medical professionals from the BMJ Group. If prompted, subscribers must sign into JCP with their journal's username and password. All users must also complete a one-time registration on BMJ Learning and subsequently log in (with a BMJ Learning username and password) on every visit.
The authors wish to thank Professor Too Heng-Phon for his kind advice regarding analysis of the splice isoforms.
Contributors BY, KJY, KTC and MST contributed to study design and manuscript preparation. BY, ML, LZ, CHK, MYL and KJY contributed to experimental data acquisition and analysis. LA contributed to clinical data acquisition.
Competing interests None.
Ethics approval Singhealth Centralised Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.