Abstract:
Neuroendocrine tumors of the small intestine (SI-NETs) have experienced a dramatic increase in incidence over the last three decades. Although defined by a small proliferative index (mostly G1 and G2 tumors), the tumors give frequently and early rise to metastases, which often exceed the size of the primary tumor and kill the patients in the end.
SI-NETs are genetically poorly characterized, the frequent loss of one chromosome 18 (Chr18) being the exception. Therefore, this doctoral thesis focused on this lesion in order to investigate potential tumor suppressors located on this chromosome (SMAD2, SMAD4, Elongin A3, CABLES, PMAIP1, and DCC).
SMAD2 and SMAD4 showed retained expression in the 14 SI-NET samples investigated (12 with loss of Chr18), leaving only haploinsuffiency as possible mechanism in tumor development and progression.
Elongin A3 and CABLES mRNAs were differentially expressed between the 1xChr18 and 2xChr18 cohort, suggesting that the loss of Chr18 has an impact on mRNA level. However, western blot analysis of 21 SI-NETs revealed preserved protein expression of Elongin A3 and CABLES. Interestingly, CABLES western blot depicted – in addition to the normal doublet-isoform – an additional isoform at ~55 kDa in the tumor samples, which was not present in the HEK293 control. Among alternative splicing, aberrant splicing of this protein is known in tumors, which could lead to the loss of the CDK-binding domain of CABLES, resulting in enhanced cell growth and tumor formation due to faster progression through the cell cycle [1].
PMAIP1 was not expressed in eight samples investigated. Since a 100% loss of a tumor suppressor is rare, the hypothesis that the lack of PMAIP1 is a normal feature of normal neuroendocrine enterochromaffin cells is favored, rather than the loss being a characteristic of neuroendocrine tumor cells.
Remarkably, the tumor suppressor protein DCC showed total loss or, at least, clearly reduced expression in nearly 30% (6/21) of the tumor samples. Abridged DCC function can result in reduced apoptosis, giving rise to tumor growth and dissemination. Alternative splicing and mutations in the intronic region of DCC [2] render this gene even more interesting.
Further investigations of our lab will focus on the transcriptome and proteome of SI-NETs, and thereby on the differential expression of gene transcripts and proteins between tumors with and without loss of Chr18; hoping to shed light on the role of DCC (and CABLES), which we found to be altered in SI-NETs.
In 2013, Banck et al. published the genomic landscape of SI-NETs with amplifications of the PI3K/AKT/mTOR pathway being the most frequent aberration [3]. Subsequently, we analyzed six genes (PDGFRα, PDGFRβ, PIK3CD, AKT1, AKT2, mTOR) involved in this pathway by FISH; revealing advanced, metastatic tumors as well as more invasive ¬¬tumors to harbor significantly more copy number (CN) alterations than tumors of early stage without metastases and less invasive tumors (UICC and T stage comparison). However, no association with protein expression or activation could be identified. One possible explanation for the discrepancy between gene and protein expression might be that epigenetic events play a role in the transcriptional control of amplified genes, thereby preventing protein overexpression.
Since especially the more aggressive tumors (defined by UICC stage IIIB and IV, as well as tumor stage 3 and 4) are lacking effective treatment, the inhibition of the PI3K/AKT/mTOR pathway could be a useful new tool in the therapy of SI-NETs. Therefore, a similar trial to the RADIANT-4 study [4] with the inclusion of functional gastrointestinal neuroendocrine tumors should enlighten the possible effect of everolimus or another inhibitor of the PI3K/AKT/mTOR pathway on the tumor progression.
Neuroendocrine tumors