New molecular players in gastric carcinogenesis

Gastric cancer (GC) is a heterogeneous disease with two major histological subtypes “intestinal” (IGC) and “diffuse” (DGC). 1-3% of GCs are truly hereditary and are dominated by the Hereditary Diffuse Gastric Cancer (HDGC) syndrome, predisposing to DGC. HDGC is mainly explained by germline alterations in the CDH1 tumor suppressor gene; however, at least 50% of cases remain without a known genetic determinant. It has recently been reported that intron 2 of the CDH1 gene gives rise to a number of non-canonical transcripts, none of which have been investigated in GC patients. Moreover, the application of exome sequencing has lately unraveled germline lesions in diverse unexpected genes predisposing to the disease. Therefore, to better characterize a cohort of 30 Italian HDGC patients already proven to be CDH1 mutation-negative, I analyzed the expression patterns of the CDH1 canonical transcript and the CDH1a and CDH1j non-canonical transcripts, arising from intron 2, in the blood of HDGC patients and normal controls by reverse transcription quantitative PCR (RT-qPCR). In addition, as part of a larger collaborative study, 7 cases underwent exome sequencing of a panel of 94 cancer-related genes and I performed an in silico analysis on all resulting sub-polymorphic and rare non-synonymous variants. In all 30 cases no expression defects in CDH1, CDH1a and CDH1j could be detected, thus dismissing the involvement of CDH1 defective splicing in those patients. Regarding exome sequencing, upon combining my in silico analysis with supportive literature findings, two genes, CDKN2A and SDHC, were found to explain disease manifestation in 2 out of the 7 patients, respectively. These genes could be put forth as novel candidates for HDGC. The CDH1 gene encodes E-cadherin, a cell-cell adhesion protein. CDH1 inactivating mutations, leading to loss of protein expression, are common in DGC, while alternative mechanisms that can subtly modulate E-cadherin expression characterize the more common IGC. These mechanisms are still poorly understood and may include small regulatory RNAs (micro-RNAs). Moreover, non-canonical transcripts arising from CDH1 intron 2, which has recently emerged as a cis-modulator of E-cadherin expression, can also be involved. Among those transcripts CDH1a proved to be expressed in GC cell lines, while being absent in the normal stomach; however its presence in GC tissue is yet to be investigated. Hence, I evaluated by digital PCR (dPCR) the expression of CDH1 and CDH1a transcripts in cancer and normal tissue samples from 32 IGC patients. Following the application of the fine-tuned dPCR, I found a significant decrease in CDH1 expression in tumors compared to the normal counterparts (P=0.001), which was especially evident in 76% of cases. Furthermore, CDH1a was detected at extremely low levels in 47% of tumors, but not in normal mucosa. A trend was additionally observed of having more CDH1 in tumors lacking the CDH1a transcript. To determine whether miRNAs could have a contributing role, I applied an in silico and literature driven approach to select potential CDH1-regulating miRNAs in IGC. A list of 16 miRNAs was produced, among which miR-92a and miR-101, have been quantified thus far by RT-qPCR. While neither miRNA correlated with the expression of CDH1, miR-101 was found to be significantly lower in tumors compared to normal mucosa (P=1.565x10-05), which points towards its implication in gastric carcinogenesis in ways that surpass CDH1 regulation. On the whole, the dPCR results support the notion that abnormal isoforms and transcripts’ imbalance resulting from cryptic abnormalities along the CDH1 locus, although subtly modulating E-cadherin expression, can still contribute to the carcinogenic process of the intestinal type.