Merlin-rapid analysis of dense genetic maps using sparse gene flow trees. Nat Genet 30:97-101 Cordell HJ (2004) Bias toward the null hypothesis in modelfree linkage analysis is highly dependent on the test statistic used.

BackgroundCXCR4, the receptor for the chemokine stromal-derived factor 1 (SDF-1), has been shown to mediate many of the processes essential for cancer progression such as tumor cell proliferation, metastasis, and angiogenesis. To understand the role of CXCR4 in the biology of neuroblastoma, a disease that presents with wide spread metastases in over 50% of patients, we screened ten patient derived-neuroblastoma cell-lines for basal CXCR4 expression and sought to identify characteristics that correlate with tumor cell phenotype.ResultsAll cell lines expressed CXCR4 mRNA at variable levels, that correlated well with three distinct classes of CXCR4 surface expression (low, moderate, or high) as defined by flow cytometry. Analysis of the kinetics of CXCR4 surface expression on moderate and high expressing cell lines showed a time-dependent down-regulation of the receptor that directly correlated with cell confluency, and was independent of SDF1. Cell lysates showed the presence of multiple CXCR4 isoforms with three major species of approximately 87, 67 and 55 kDa associating with high surface expression, and two distinct species of 45 and 38 kDa correlating with low to null surface expression. Western blot analysis of CXCR4 immunoprecipitates showed that the 87 and 67 kDa forms were ubiquitinated, while the others were not. Finally, treatment of cells with a proteasome inhibitor resulted in down regulation of CXCR4 surface expression.ConclusionsTaken together, these data show that regulation of CXCR4 surface expression in neuroblastoma cells can occur independently of SDF-1 contribution arguing against an autocrine mechanism. Additionally these data suggest that post-translational modifications of CXCR4, in part through direct ubiquitination, can influence trafficking of CXCR4 to the surface of neuroblastoma cells in a ligand-independent manner.

A cDNA microarray comprising 5184 different cDNAs spotted onto nylon membrane filters was developed for prostate gene expression studies. The clones used for arraying were identified by cluster analysis of > 35 000 prostate cDNA library-derived expressed sequence tags (ESTs) present in the dbEST database maintained by the National Center for Biotechnology Information. Total RNA from two cell lines, prostate line 8.4 and melanoma line UACC903, was used to make radiolabeled probe for filter hybridizations. The absolute intensity of each individual cDNA spot was determined by phosphorimager scanning and evaluated by a bioinformatics package developed specifically for analysis of cDNA microarray experimentation. Results indicated 89% of the genes showed intensity levels above background in prostate cells compared with only 28% in melanoma cells. Replicate probe preparations yielded results with correlation values ranging from r = 0.90 to 0.93 and coefficient of variation ranging from 16 to 28%. Findings indicate that among others, the keratin 5 and vimentin genes were differentially expressed between these two divergent cell lines. Follow-up northern blot analysis verified these two expression changes, thereby demonstrating the reliability of this system. We report the development of a cDNA microarray system that is sensitive and reliable, demonstrates a low degree of variability, and is capable of determining verifiable gene expression differences between two distinct human cell lines. This system will prove useful for differential gene expression analysis in prostate-derived cells and tissue.

A cDNA microarray comprising 5184 different cDNAs spotted onto nylon membrane filters was developed for prostate gene expression studies. The clones used for arraying were identified by cluster analysis of > 35 000 prostate cDNA library-derived expressed sequence tags (ESTs) present in the dbEST database maintained by the National Center for Biotechnology Information. Total RNA from two cell lines, prostate line 8.4 and melanoma line UACC903, was used to make radiolabeled probe for filter hybridizations. The absolute intensity of each individual cDNA spot was determined by phosphorimager scanning and evaluated by a bioinformatics package developed specifically for analysis of cDNA microarray experimentation. Results indicated 89% of the genes showed intensity levels above background in prostate cells compared with only 28% in melanoma cells. Replicate probe preparations yielded results with correlation values ranging from r = 0.90 to 0.93 and coefficient of variation ranging from 16 to 28%. Findings indicate that among others, the keratin 5 and vimentin genes were differentially expressed between these two divergent cell lines. Follow-up northern blot analysis verified these two expression changes, thereby demonstrating the reliability of this system. We report the development of a cDNA microarray system that is sensitive and reliable, demonstrates a low degree of variability, and is capable of determining verifiable gene expression differences between two distinct human cell lines. This system will prove useful for differential gene expression analysis in prostate-derived cells and tissue.