The lack of effective conventional therapies for the treatment of advanced stage melanoma has stimulated interest in the application of novel strategies for the treatment of patients with malignant melanoma. Because of its expression in a large percentage of melanoma lesions and its restricted distribution in normal tissues, the high molecular weight-melanoma-associated antigen (HMW-MAA), also known as the melanoma chondroitin sulfate proteoglycan (MCSP), has been used to implement immunotherapy of melanoma. The potential clinical relevance of HMW-MAA/MCSP has stimulated investigations to characterize its structural properties and biological function in melanoma cells. Over the last 10 years, the field of HMW-MAA/MCSP research has seen tremendous growth. Specifically, a significant amount of information has been accumulated regarding (1) the structural characteristics of the HMW-MAA/MCSP, (2) its role in the biology of melanoma cells, and (3) the potential molecular mechanisms underlying the association between HMW-MAA/MCSP-specific immunity and survival prolongation in melanoma patients immunized with HMW-MAA/MCSP mimics. In this review, we summarize the characteristics of the HMW-MAA/MCSP in terms of its structure, antigenic profile, tissue distribution, and similarities with its counterparts in other animal species. Additionally, we discuss the role the HMW-MAA/MCSP plays in melanoma cell biology with emphasis on the recently identified signal transduction pathways triggered by the HMW-MAA/MCSP. Finally, we discuss the potential molecular mechanisms underlying the beneficial effect of anti-HMW-MAA/MCSP antibodies on the clinical course of the disease in patients with melanoma.

The non-classical HLA class I antigen HLA-G is an immune modulator which inhibits the functions of T cells, NK cells, and the Dendritic cells (DC). As a result, HLA-G expression in malignant cells may provide them with a mechanism to escape the immune surveillance. In melanoma, HLA-G antigen expression has been found in 30% of surgically removed lesions but in less than 1% of established cell lines. One possible mechanism underlying the differential HLA-G expression in vivo and in vitro is that the HLA-G gene is epigenetically repressed in melanoma cells in vitro. To test this hypothesis, we treated the HLA-G negative melanoma cell line OCM-1A with the DNA methyltransferase inhibitor 5-aza-2'-deoxycytidine (5-AC) and analyzed whether HLA-G expression can be restored. Our data strongly suggest that HLA-G is silenced as a result of CpG hypermethylation within a 5' regulatory region encompassing 220 bp upstream of the start codon. After treatment, HLA-G mRNA expression was dramatically increased. Western blot and flow cytometry showed that HLA-G protein was induced. Interestingly, HLA-G cell surface expression on the 5-AC treated OCM-1A cells is much less than that on the HLA-G positive JEG-3 cells while a similar amount of total HLA-G was observed. Possible mechanisms for the difference were analyzed in the study such as cell cold-treatment, peptide loading and antigen processing machinery components (APM) as well as β 2 microglobulin (β 2 -m) expression. Data revealed that the APM component calreticulin might be involved in the lower HLA-G surface expression on OCM-1A cells. Taken together, our results indicated that DNA methylation is an important epigenetic mechanism by which HLA-G antigen expression is modulated in melanoma cells in vitro. Furthermore, to the first time, we hypothesized that the deficiency of calreticulin might be involved in the low HLA-G surface expression on the 5-AC treated OCM-1A cells.

In this study, the authors propose a new feature selection scheme, the incremental forward feature selection, which is inspired by incremental reduced support vector machines. In their method, a new feature is added into the current selected feature subset if it will bring in the most extra information. This information is measured by using the distance between the new feature vector and the column space spanned by current feature subset. The incremental forward feature selection scheme can exclude highly linear correlated features that provide redundant information and might degrade the efficiency of learning algorithms. The method is compared with the weight score approach and the 1-norm support vector machine on two well-known microarray gene expression data sets, the acute leukemia and colon cancer data sets. These two data sets have a very few observations but huge number of genes. The linear smooth support vector machine was applied to the feature subsets selected by these three schemes respectively and obtained a slightly better classification results in the 1-norm support vector machine and incremental forward feature selection. Finally, the authors claim that the rest of genes still contain some useful information. The previous selected features are iteratively removed from the data sets and the feature selection and classification steps are repeated for four rounds. The results show that there are many distinct feature subsets that can provide enough information for classification tasks in these two microarray gene expression data sets.

Anaplastic thyroid carcinoma (ATC) is almost universally fatal. Elevated keratin-8 (KRT8) protein expression is an established diagnostic cancer biomarker in several epithelial cancers (but not ATC). Several keratins, including KRT8, have been suggested to have a role in cell biology beyond that of structural cytoskeletal proteins. Here, we provide evidence that KRT8 plays a direct role in the growth of ATCs. Genomic and transcriptomic analysis of >5000 patients demonstrates that KRT8 mutation and copy number amplification are frequently evident in epithelial-derived cancers. Carcinomas arising from diverse tissues exhibit KRT8 mRNA and protein overexpression when compared to normal tissue levels. Similarly, in a panel of patient-derived ATC cell lines and patient tumors, KRT8 expression shows a similar pattern. sh-RNA-mediated KRT8 knockdown in these cell lines increases apoptosis, whereas forced overexpression of KRT8 confers resistance to apoptosis under peroxide-induced cell stress conditions. We further show that KRT8 protein binds to annexin A2, a protein known to mediate apoptosis as well as the redox pathway.

Abstract. The computational difficulties occurred when we use a conventional support vector machine with nonlinear kernels to deal with massive datasets. The reduced support vector machine (RSVM) replaces the fully dense square kernel matrix with a small rectangular kernel matrix which is used in the nonlinear SVM formulation to avoid the computational difficulties. In this paper, we propose a new algorithm, Systematic Sampling RSVM (SSRSVM) that selects the informative data points to form the reduced set while the RSVM used random selection scheme. This algorithm is inspired by the key idea of SVM, the SVM classifier can be represented by support vectors and the misclassified points are a part of support vectors. SSRSVM starts with an extremely small initial reduced set and adds a portion of misclassified points into the reduced set iteratively based on the current classifier until the validation set correctness is large enough. In our experiments, we tested SSRSVM on six public available datasets. It turns out that SSRSVM might automatically generate a smaller size of reduced set than the one by random sampling. Moreover, SSRSVM is faster than RSVM and much faster than conventional SVM under the same level of the test set correctness.