Role of ALK activation in the development and maintenance of the neoplastic phenotype in neuroblastoma

10008

Background: Activating mutations of the Anaplastic Lymphoma Kinase (ALK) receptor could be responsible for most familial neuroblastoma cases and for up to 15% of somatic cases. The objective of the present study was to further investigate the role of ALK activation in neuroblastoma. Methods: Tissue microarrays were constructed containing 101 primary tumors and 56 paired normal tissues. Sections were immunostained with anti-ALK or anti-P-ALK antibodies, and with antibodies directed against the ALK ligands: PTN (Pleiotrophin) or MDK (Midkine). The Wilcoxon signed rank test was applied for comparison of paired data. Associations with prognostic factors were analyzed using t-tests. Effects of the ALK inhibitor TAE684 (Novartis) on cell proliferation and signaling was evaluated in wild-type or mutated ALK neuroblastoma cell lines and xenografts. Results: ALK was expressed in about 100% of tumors and normal tissues, while phospho-ALK was detected in 5% of normal tissues and 50% of tumors. Sequencing of the kinase domain of ALK showed that its phosphorylation was largely independent of mutations and we found that MDK and PTN ligands were expressed in 66% and 50% of tumors, respectively. Interestingly, ALK, P-ALK, and MDK were expressed at higher levels in tumors as compared with paired normal tissues (p < 0.0001), while PTN showed an inverse tendency, being more expressed in normal tissues (p = 0.07). In tumors, P-ALK was associated with good-prognosis factors, including favorable stages (p = 0.01), absence of MYCN amplification (p = 0.05) and a younger age at diagnosis (p = 0.03). Inhibition of cell proliferation by TAE684 was detectible in all neuroblastoma cell lines, regardless of ALK status. However, TAE684 failed to demonstrate antitumor activity in advanced stage neuroblastoma xenografts expressing either a wild-type or a mutated ALK. Interestingly, ALK pathway activation (P-STAT3, P-AKT) was weak or barely detectible in these xenografts. Conclusions: ALK activation occurs during neuroblastoma oncogenesis, along with a concomitant switch between the expressions of PTN and MDK. However, ALK may not be a relevant therapeutic target since in vivo inhibition showed no antitumor activity.

[Table: see text]

Pleiotrophin, a candidate gene for poor tumor vasculature and in vivo neuroblastoma sensitivity to irinotecan

In vivo neuroblastoma (NB) xenograft model, resistant to the DNA-topoisomerase I inhibitor irinotecan (CPT-11), has been established to study resistance mechanisms acquired in a therapeutic setting. Common mechanisms of resistance were not involved in this resistance. Thus, we compared the gene expression profiles of sensitive, resistant, and reverted tumors using cDNA expression arrays. Expression of selected transcripts was confirmed by quantitative real-time PCR. We found that pleiotrophin (PTN), a heparin-binding growth factor, was the only gene significantly affected: PTN gene expression was downregulated in all resistant tumors (8-14-fold) as compared to sensitive tumors, and was increased (2-4-fold) in all reverted tumors as compared to resistant tumors. PTN thus appeared to be a likely candidate gene associated with resistance to CPT-11 in this in vivo model. To investigate the direct implication of PTN in NB, we transfected two NB cell lines with RNA interferences in order to silence PTN. PTN failed to demonstrate implication in resistance to CPT-11 in vitro but could influence sensitivity to CPT-11 exclusively through an in vivo mechanism. Indeed, vasculature was significantly enhanced in resistant NB xenografts compared to sensitive and reverted xenografts, and we suggest that PTN is acting in our resistant in vivo NB model as an angiostatic factor.