No abstract

Recognition of individuals with a genetic predisposition to gastric cancer (GC) enables preventive measures. However, the underlying cause of genetic susceptibility to gastric cancer remains largely unexplained. We performed germline whole-exome sequencing on leukocyte DNA of 54 patients from 53 families with genetically unexplained diffuse-type and intestinal-type GC to identify novel GC-predisposing candidate genes. As young age at diagnosis and familial clustering are hallmarks of genetic tumor susceptibility, we selected patients that were diagnosed below the age of 35, patients from families with two cases of GC at or below age 60 and patients from families with three GC cases at or below age 70. All included individuals were tested negative for germline CDH1 mutations before or during the study. Variants that were possibly deleterious according to in silico predictions were filtered using several independent approaches that were based on gene function and gene mutation burden in controls. Despite a rigorous search, no obvious candidate GC predisposition genes were identified. This negative result stresses the importance of future research studies in large, homogeneous cohorts.

Germline inactivation of the E-cadherin gene (CDH1) is associated with hereditary diffuse gastric cancer (HDGC), a rare autosomal dominant syndrome predisposing to both diffuse gastric cancer (DGC) and lobular breast cancer (LBC). We searched for CDH1 germline defects in 32 HDGC Italian probands selected according to international consensus criteria and in 5 selected relatives. We used a series of molecular methods, including: DNA sequencing, multiplex ligation-dependent probe amplification, single-nucleotide primer extension, bisulfite sequencing, reverse-transcription PCR, and bioinformatics tools. We identified pathogenic mutations in 6 out of 32 probands (19%): one truncating and two missense mutations, one large deletion, one allelic expression imbalance and one splicing defect. Three out of six CDH1 constitutive alterations were novel. Our data support the need for a multimethod approach for CDH1 genetic testing, demonstrating that both DNA and RNA analyses are required to increase the detection rate of pathogenic mutations, thus reducing the number of patients without a clear molecular diagnosis. On the whole, our results indicate that not only DGC patients, but also subjects with personal or family history of LBC might benefit from CDH1 genetic testing. Moreover, our findings support the notion that prophylactic gastrectomy should be offered to asymptomatic CDH1 mutation carriers; indeed, while endoscopic analysis with histological examination of random gastric biopsies can miss cancer foci, gastrectomy performed in these subjects always revealed foci of cancer cells.

E-cadherin is a cell-cell adhesion protein encoded by CDH1 tumor-suppressor gene. CDH1 inactivating mutations, leading to loss of protein expression, are common in gastric cancer of the diffuse histotype, while alternative mechanisms modulating E-cadherin expression characterize the more common intestinal histotype. These mechanisms are still poorly understood. CDH1 intron 2 has recently emerged as a cis-modulator of E-cadherin expression, encoding non-canonical transcripts. One in particular, CDH1a, proved to be expressed in gastric cancer cell lines, while being absent in the normal stomach. For the first time, we evaluated by digital PCR the expression of CDH1 and CDH1a transcripts in cancer and normal tissue samples from 32 patients with intestinal-type gastric cancer. We found a significant decrease in CDH1 expression in tumors compared to normal counterparts (P = 0.001), which was especially evident in 76% of cases. CDH1a was detected at extremely low levels in 47% of tumors, but not in normal mucosa. A trend was observed of having less CDH1 in tumors expressing CDH1atranscript. The majority of tumors with both a decrease in CDH1 and presence of CDH1a also showed a decrease in miR-101 expression levels. On the whole, the decrease of CDH1 transcript, corresponding to the canonical protein, and the presence of CDH1a, corresponding to an alternative isoform, are likely to perturb E-cadherin-mediated signaling and cell-cell adhesion, thus contributing to intestinal-type gastric carcinogenesis.