Identification and characterization of CDH1 germline variants in sporadic gastric cancer patients and in individuals at risk of gastric cancer

Abstract Gastric cancer (GC) remains a leading cause of cancer death worldwide. A hallmark of this type of tumour is the poor prognosis of patients. Identification of subjects with an increased risk of developing GC and the early detection of GC, are promising approaches to reduce the morbidity and mortality of this malignancy. CDH1 gene encodes E-cadherin, a known tumour suppressor protein expressed in epithelial cells that plays a major role in cellular adhesion and tissue architecture maintenance. CDH1 gene mutations are thought to affect protein function and lead to cell-cell adhesion deregulation in epithelial tissues and to enhance metastatic tumour potential. Germline mutations in CDH1 gene are implicated in hereditary diffuse gastric cancer (HDGC) syndrome and in early onset gastric cancer (EOGC), but the role of these variants in sporadic GC and in subjects at risk to develop GC, is poorly investigated. Most studies limit the screening of CDH1 mutations to patients with HDGC, familial GC or GC patients at very early onset through genetic counselling, but only few studies investigated the CDH1 the contribution of these mutations in sporadic GC cases without familial aggregation or in subjects at risk of developing GC. Moreover, the potential functional effects of CDH1 variants are not always explored. In our study, screening of the entire coding region and all exon flanking sequences of the CDH1 gene was performed by direct sequencing in a series of random, consecutive sporadic GC cases. We extended the analysis of CDH1 germline mutations in a collection of individuals at risk for GC: 59 subjects who reported an ascertained case of GC among parents, children, siblings or offspring's, named first degree GC-Relatives (FDR), and 20 autoimmune metaplastic atrophic gastritis (AMAG) patients. In the study 52 healthy blood donors (BD) were also included in the CDH1 analyses. The aim of this PhD project was to evaluate CDH1 mutations in these individuals to identify a possible hereditable marker that might improve the early detection of GC, stratifying high risk subjects in association with other known risk factors for GC diagnosis. All the participants were voluntarily subjected to a gastrointestinal endoscopy with biopsy in our institution. This is the first study in which CDH1 screening was investigated in AMAG patients. CDH1 germline mutations found were characterized by means of structural modelling and bioinformatic tools, immunohistochemical (IHC) stains for evaluation of E-cadherin and ?-catenin proteins expression and subcellular localization, and E-cadherin mRNA level by real-time polymerase chain reaction (PCR). The functional impact of splicing was assessed for non-polymorphic intronic variants. Among the 59 GCs, the lower portion (antrum) of the stomach was the prevalent anatomic site (p< 0.05) for the tumours detected by endoscopy. No statistically differences were found comparing associations between gender, histotype, tumour location, H. Pylori infection, and TNM staging. A complex panel of germline variants was detected for CDH1 gene. In GC series, we found six different alterations in nine distinct GC patients (15.2%): 67% of the missense type (p.G274S; p.A298T; p.T470I: p.A592T) and 33% of non-missense type, one in the promoter region (5'UTR-71C>G) and one in the intronic region close to exon 13 (IVS12 c.1937-13T>C). All the missense substitutions are localized in the extracellular functional E-cadherin domain which is directly implied in cell-cell adhesion mechanisms. Of note the germline missense (p.G274S) was a new CDH1 mutation, while the p.A298T was already found only in HDGC syndrome and it was described as pathogenetic. Our molecular modelling studies highlighted that p.G274S change in mature protein is not dramatic for local structure but the serine residue represents a potential site for post translation modifications, such as phosphorylation or glycosylation. However, data from our functional in vitro assays did not support an effect of this mutation on adhesion mobility of cells. Moreover, E-cadherin was not under expressed in the tumour gastric tissue of patient harbouring the p.G274S, even if a reduction in ?-catenin expression was observed in the same sample. Therefore, the pathogenic effect of p.G274S change by our in silico and in vitro approaches seems to be unlikely, so we cannot conclude the significance of this mutation. Modelling and SIFT results confirmed that the p.A298T substitution affects protein function, while the p.T470I change is tolerated. Through the structural analysis for p.T470I, we hypothesized a protective role that may favour protein-protein interactions in the extracellular medium. The p.A592T was detected in the four studied groups, suggesting that it is a polymorphic variant (frequence >1%). Modelling suggested a non-pathogenic role as already confirmed by published in vitro and in silico studies. The C-to-G change before the start codon (-71C>G), represented the most common variant associated with GC in our series, occurring in three (5.1%) out of 59 GC patients. Due to a lack of tumour material, an evaluation or a correlation between CDH1 hypermethylation and E-cadherin expression was not performed, so the predicted pathogenicity of this promoter variant still remains to be elucidated. The intronic IVS12 c.1937-13T>C variant was found in two GC patients and in one BD, but it was already described in HDGC and EOGC subjects. Our assays demonstrated for the first time that this substitution leads to an aberrant CDH1 transcript harbouring a deletion of exon 11 with the formation of a premature stop codon in the extracellular E-cadherin domain. The translated protein will lack part of the extracellular domain, and completely loose the transmembrane domain and the cytoplasmic tail of E-cadherin, which is involved in ?-catenin binding. Moreover, IHC stains showed a reduction in the expression of E-cadherin and ?-catenin in the gastric tumour tissues of patients harbouring this intronic mutation. Evaluation by real time-PCR of E-cadherin expression from the EBV immortalized B-lymphocytes from one GC patient carrying this mutation showed a strong (60%) and significantly reduction (p<0.01) compared with the control without CDH1 alterations, and the BD with the same CDH1 alteration. In FDR high risk group, we found one substitution in the 5'UTR promoter region (-54G>C) already described as a rare variant able to decrease the transcriptional activity of CDH1, one 5' near gene variant (-176C>T) with unknown significance, and an intronic mutation close to exon 5, IVS4 c.532- 18C>T, already reported in EOGC cases. This last one showed no effects in producing CDH1 aberrant splicing, as the new germline mutation in the intronic region close to exon 1 (IVS1 c.48+7C>T) found in one AMAG patient. Other variants in the AMAGs were polymorphic, and thus we excluded a potential effect of these on E-cadherin function. In conclusion, our results show that pathogenic CDH1 germline variants (p.A298T) can also be detected in sporadic GC patients without fulfilling the strict criteria for HDGC. We also report the finding, in a sporadic GC, of a new missense germline mutation (p.G274S) with an uncertain significance. Furthermore, we demonstrate for the first time a potential deleterious effect on splicing and a decreased E-cadherin expression for the intronic IVS12 c.1937-13T>C mutation. Our detection of the same genetic alteration and splicing effect in one BD, but with a limited effect on E-cadherin mRNA level, is intriguing and deserves further studies. Performing CDH1 screening in individuals at risk for GC, we detected the IVS4 c.532-18C>T variant already reported in EOGC cases, and a new IVS1 intronic mutation in one AMAG patient, although there was no influence on CDH1 splicing in both cases. The overall percentage of the missense mutations (67%) in our series is predominantly of the non-missense type, suggesting that the North-East of Italy should be considered as a middle-to-high-risk area for the incidence of E-cadherin mutations and the risk of GC.