No abstract

No abstract

No abstract

The winged helix transcription factors HNF-3/FKH (forkhead homologs) activate endodermal-derived and acute-phase gene expression and control gut development in Drosophila. Trefoil factor family (TFFs) peptides are vertebrate products secreted by mucin-producing epithelial cells of the gastrointestinal tract involved in restitution and repair of the mucosa. They are positively regulated in ulcerative and neoplastic conditions. We describe a consensus sequence in human and rodent TFF promoters close to the TATAA box showing striking similarity to the binding site of the HNF-3/FKH family. In gel retardation assays, HNF-3 alpha and beta bound predominantly to the site in TFF1 (formerly pS2) and, to a lesser extent, to the sites in TFF2 or TFF3. Mutations generated in this motif severely impaired transcription of TFF1 reporter genes. Cotransfection with expression vectors of HNF-3alpha and beta, but not the related HFH 11A and B, specifically activated the wild-type TFF1 reporter genes. Activation of endogenous expression of TFF1 by HNF-3 alpha and beta gene products was more than 1000 fold in the pancreatic cell line Capan-2 and fivefold in the gastric cell line MKN-45, whereas the intestinal cell lines HUTU 80 and HT-29 displayed no effect. Thus, HNF-3/FKH factors contribute causally to cell-specific regulation of TFF genes and may explain the acute-phase response of TFF peptides.

Small peptides displaying a cysteine-rich module (termed P-domain or trefoil motif) form a recently increasing group of peptides abundantly expressed at mucosal surfaces of specific tissues and are associated with the maintenance of surface integrity. The estrogen-inducible pS2 gene (BCEI) and the human homolog to the porcine spasmolytic peptide (hSP) gene (SML1) appear synchronously expressed in healthy stomach mucosa and several carcinomas of the gastrointestinal tract. Both genes were shown to be located at 21q22.3. A new trefoil peptide from human intestinal mucosa (hITF/hP1.B) and its gene (TFF3) were described recently. By PCR analysis of a somatic cell hybrid panel and FISH using two large genomic recombinants (110 kb, 210 kb) cloned in the Bacterial Artificial Chromosome (BAC) system, we show that this gene coding for the new member of human P-domain/trefoil peptides also maps to chromosome region 21 q22.3 suggesting a physical linkage of all three trefoil peptide genes.

We studied the pattern of immunohistochemical expression of E-cadherin in a series of 50 gastric carcinomas, aiming to analyze its relationship with histotype and features of biological aggressiveness of the tumors and survival of the patients. Abnormal E-cadherin expression was significantly (p=.0007) higher in diffuse/isolated-cell type carcinomas than in intestinal/glandular carcinomas. In mixed carcinomas abnormal E-cadherin expression in the diffuse/isolated-cell-type component (94.4%) was significantly (p=.007) higher than in intestinal/glandular component (55.6%). Significant relationships were observed between abnormal E-cadherin expression and nodal metastases (p=.004) and pTNM stages (p=.05). Survival of patients with tumors displaying abnormal E-cadherin expression was worse than that of patients with tumors presenting normal expression, though not attaining the threshold of statistical significance (p=.l9). We conclude that abnormal E-cadherin expression is correlated with diffuse/isolated-cell histotype and features of biological aggressiveness of gastric carcinoma.

No abstract

Purpose: To evaluate the benefit of the additional available information present in spectral CT datasets, as compared to conventional CT datasets, when utilizing convolutional neural networks for fully automatic localisation and classification of liver lesions in CT images. Materials and Methods: Conventional and spectral CT images (iodine maps, virtual monochromatic images (VMI)) were obtained from a spectral dual-layer CT system. Patient diagnosis were known from the clinical reports and classified into healthy, cyst and hypodense metastasis. In order to compare the value of spectral versus conventional datasets when being passed as input to machine learning algorithms, we implemented a weakly-supervised convolutional neural network (CNN) that learns liver lesion localisation without pixel-level ground truth annotations. Regions-of-interest are selected automatically based on the localisation results and are used to train a second CNN for liver lesion classification (healthy, cyst, hypodense metastasis). The accuracy of lesion localisation was evaluated using the Euclidian distances between the ground truth centres of mass and the predicted centres of mass. Lesion classification was evaluated by precision, recall, accuracy and F1-Score. Results: Lesion localisation showed the best results for spectral information with distances of 8.22 ± 10.72 mm, 8.78 ± 15.21 mm and 8.29 ± 12.97 mm for iodine maps, 40 keV and 70 keV VMIs, respectively. With conventional data distances of 10.58 ± 17.65 mm were measured. For lesion classification, the 40 keV VMIs achieved the highest overall accuracy of 0.899 compared to 0.854 for conventional data. Conclusion: An enhanced localisation and classification is reported for spectral CT data, which demonstrates that combining machine-learning technology with spectral CT information may in the future improve the clinical workflow as well as the diagnostic accuracy.